BookBub Ads Expert – Resources

Welcome to the bonus resources page for purchasers of BookBub Ads Expert. If you haven’t yet purchased your own copy, well, congratulations on getting passed security – pick up your copy of BookBub Ads Expert here. (You’ll need it!)

For everyone else, we’ve got a gallery of winning images, followed by a step-by-step guide to ad creation – with some handy screenshots, I’ve got some up-to-the-minute information on the Also Bought situation at Amazon – with some useful workarounds for those affected, tons of detailed optimization advice – with some talk through my own process, links to a bunch more tools and resources – including some great case studies, a place where you can ask questions and get help, and ways that you can spread the love about BookBub Ads Expert!

Latest News:

I have just added a VIDEO GUIDE to ad creation below, which was much requested.

Also check out the section below titled BOOKBUB CHANGES which will get you up to speed on two recent additions to the interface: Readers v Followers, and CTR-by-Author. Both are great changes, and I have some best practice tips for you below.

Winning Images

Step By Step Ad Creation

Also Boughts and Comp Authors

BookBub Changes

Optimization Advice

Tips & Tools

Have Questions?

Spread The Love!

I couldn’t include images in BookBub Ads Expert for all sorts of reasons. Chief among them: Amazon deducts file size costs from our royalties and images really bulk those files, eating into our precious moolah.

Instead, I created this handy reference page instead for all the purchasers of the book – which will also be easier for me to update. If you haven’t checked the book out, and stumbled in here somehow, this page will be of little use to you. It’s not a complete guide to using BookBub Ads by any stretch and you will miss all the good stuff that will help you build truly profitable campaigns, including my soon-to-be-indispensable ninja tricks and killer moves. I strongly recommend looking at the description for BookBub Ads Expert to see what you’re missing. (Spoiler Alert: everything!)

For the glamorous and discerning people who have already purchased the book, all the resources I mentioned there are below. This page is pretty long but you can click whatever you’re looking for from the above list and navigate directly there easily via techno-tomfoolery.

Winning Images

Winning Images header

If you have read BookBub Ads Expert you will have heard me go through the elements which make a winning BookBub Ad in exhaustive detail. I also explained how to create an ad which embodies those principles, and how to brief your designer along those lines if you are outsourcing which I strongly recommend in most cases.

Quite frankly, when someone is struggling with BookBub Ads even after reading my book, sub-par images is the #1 reason. Either spend the small amount to outsource, or dig deeper and push yourself harder when creating these ad images.

As a reminder, a winning the BookBub beauty contest requires pro design, a clear offer, communicating the niche clearly, and putting the book cover front-and-center.

Below is a gallery of images which embody these principles – some of which I commissioned, others I designed myself, and then some are just ones which I spotted and liked in my own daily email.

One last thing on images: these “rules” aren’t cast-iron Laws of the Universe or anything. Data always beats the most convincing theory. If you want to test a different approach, feel free! And if it works, it works.

To underline the point, here’s an ad I saw in my own daily BookBub email which breaks my “rules,” but works brilliantly.

(Apologies to the artist – I snapped this on the move and the resolution isn’t great, especially when blowing it up for illustrative purposes. Sorry for butchering your wonderful work!)

Anyway, if you click through from that ad above to the Amazon page for this book, you’ll see that it was breaking the letter of the law rather than the spirit – the ad art matches the cover art perfectly. And the illustration is good enough to make this book an autobuy for me and many other fans of the genre, I’m sure. A great use of great art.

Also Boughts

Also Boughts header

I talk about Also Boughts in the book, particularly how they are often a fruitful source of potential comp authors for your first test ads. They are also important and useful in lots of other ways too, of course, which causes a lot of grumbling when Amazon fiddles with them, as it routinely does, often testing various iterations of our product pages. More on that phenomenon here:

also boughts and amazon recommendations

I’m not worried about Also Boughts permanently disappearing because I think they are too popular with readers, but it can mess with your testing if that information source is closed off temporarily, right when you need it. There are workarounds though.

If you are currently seeing Also Boughts in the US Kindle Store, then don’t worry. All I would suggest is making a note of your Also Boughts, particularly the authors and books which feature near the top, and those that make multiple appearances, as it’s likely that Also Boughts will disappear for you (temporarily) at some point in the future.

But if you are not currently seeing Also Boughts in the US Kindle Store (or elsewhere), first of all, don’t panic. They will likely return at some point; they always do. Here are some workarounds in the meantime:

Check your book pages internationally – i.e. in the Canadian, Australian, and UK Kindle Stores. At least a couple of those should have Also Boughts at any given time, and while they won’t be identical to your US Also Boughts, it’s still useful information and is often a close approximation.

Check your author page on Amazon. Those Also Boughts not only have never been messed with by Amazon, they are also usually stronger potential comp authors too, as they are aggregate Also Boughts across all your books. (Multi-genre authors must take care here though for obvious reasons).

Explore the other sources I outlined in the book for generating your comp authors too. Sometimes you have to lean on this very heavily anyway, especially if many of your comp authors are all-in with Kindle Unlimited, or otherwise haven’t played in BookBub’s sandpit very much. You can get a very strong list of comp authors without using Also Boughts whatsoever.

Overall, don’t sweat it. Amazon regularly cycles through different iterations of our book pages, sometimes moving Also Boughts to the bottom of the page, sometimes ditching them altogether. They always come back eventually.

One last tip: sometimes you can see a different version of your book page by using a different device or browser. Worth a try.

Ad Creation

Ad Creation header

I have just created a brand new video guide on ad creation, which will have lots of tips for you. Check it out on YouTube here:

For those who prefer old-fangled words – tarted up with some screenshots – keep reading.

This is just a simple guide to creating a test ad so you have something you can easily refer to in another tab while making your own. For the full breakdown on what makes a successful ad, refer to Part II of BookBub Ads Expert. Ensure you have also gone through the process outlined in Part III before running your first test ads. That will save you a lot of pain.

Assuming you already have a BookBub Partner account — if not, set one up as soon as possible — then log in, click on the Ads tab from the navbar at the top of the screen, and then select the green button in the top-right of the interface marked Create New Ad.

Note that I have used the same headlines for each step below as should appear on your BookBub Ads interface for ease of reference. You can thank me via statues.

Step 1: Choose A Book (Optional)

This should be obvious, but sometimes the system misses your book. Add it here if you can, but if you still can’t select it — even after adding your book — don’t worry. I find this to be a little glitchy at times, but it doesn’t ultimately matter. The only function this serves, aside from tracking some data for you to analyze later (which you can do manually) is auto-populate the links for your ad, which you can also do manually yourself. It’s wise to add your book anyway if it’s not already on BookBub, but don’t worry about this otherwise.

To be absolutely clear: there is zero effect on the performance of your ads and this step can be comfortably skipped.

Step 2: Ad Creative

You should already have a pro graphic to drop in here in the section named Upload Existing Creative. (Or you will shortly.) Whatever you do, don’t use BookBub’s tool here to build your image — which appears when you click Build New Creative, if you are curious. While this tool doesn’t do a bad job per se, you can do a lot better than that yourself — and little differences can be crucial on this platform. Don’t get lazy here! Put in the effort to make your image amazing and you will reap the rewards in terms of improved CTR, lower CPC, and better conversion.

If you associated a book with your ad in Step 1, you will see that one or more links have auto-populated in this section. We will be restricting ourselves to Amazon US for the purposes of these tests so remove any others the system as inserted by clicking the on/off button to the left of the link. However, if you didn’t associate a book with your ad in Step 1, then simply insert the link manually yourself in the indicated spot. The system will automatically recognize which retailer and territory it relates to using cyber-sorcery.

Whether your link was inserted by the system or by you, triple-check that link. As in, actually click on it, don’t just glance at it. This is a great habit to get into, by the way. While mistakes are rare and usually on the user side, to be honest — getting this right is important enough to devote a whole click into checking whether your ad is pointing at the right place. From personal experience, mistakes are most likely when your eyes glaze over after doing lots of test ads, or complicated campaigns pointing at different places. Be especially careful in those situations.

There also a known bug involving links that you have switched off at this stage turning back on again on the confirmation page — but I’ll return to that when we get to that confirmation screen as you can’t do anything about it right now. Consider the problem flagged, yo.

Step 4: Audience Targeting

As you already have your possible author comps from the process outlined earlier, the only wrangling here is deciding how many to lump together. Obviously, you will get the cleanest read from testing one author at a time, but some of your potential comp authors might only have 500 or 1000 BookBub followers. In such cases, throw a few similar authors together. Aim for a total of a few thousand followers to get enough serves (you are shooting for at least 1,000 impressions here). Keep in mind that some of those following a given author will be based outside of America or shop somewhere other than Amazon; others won’t open their email that particular day — it varies quite a lot.

This sounds a little hazy. But don’t worry, there is a handy visual indicator in the form of a colored dial beside the Book Category field. Don’t aim for the yellow zone on the right — that’s an indication your targeting is too broad. You really want it up the middle, with the system saying “Defined,” ideally. If you are in the red zone, that means your audience is too narrow for the ad to serve and you need to add more authors. And possibly coffee.

Just make sure to also filter by genre (“Refine by Book Category” in the parlance of the interface). Remember, many authors write in more than one genre, and it’s important to narrow your audience or your CTR will drop — and BookBub is a platform where small margins can make a very big difference, which I will repeat until this mantra is tattooed on your brain by a drumming of woodpeckers, which actually is the respective collective noun, I’m reliably informed.

Step 5: Schedule & Budget

It doesn’t really matter which dates you pick as you will be pushing to serve the ad as soon as possible, but it’s always wise to put an end date in of tomorrow or some such. Without wanting to pack you all off to the monastery, this is another good habit to get into. While it hasn’t happened to me on BookBub, I have started an experimental test campaign on Facebook without putting in an end date, then promptly fell sick for a few weeks, and then wept uncontrollably when seeing what my card was charged at the end of that month — with very little to show for it in terms of sales. I can’t even bring myself to type the amount; I may need sustained therapy before I’m able to take that step. Or perhaps my old friend Mr. Whiskey could take care of those brain cells for me and we’ll call it even.

As for the budget, just drop in $10 or $15. You might not spend all of it and you definitely don’t need to spend more — that will get you around 1,000 impressions which is more than enough to know whether an ad will work. Select the option to “Fulfill as quickly as possible” as you don’t want to spread out the serving and drag out the testing process; you want results now. (Although I often choose this option even when not in the testing phase, as I prefer to micro-manage the budget — more on that later.)

Step 6: Bid

We are testing, and we want results quickly. This means bidding aggressively and bullying everyone else out of the auction. You might bid a little more competitively in the future (a lot more competitively if you are doing a drip campaign on a permafree, for example), but now is not the time for niceties—there is precious data to be harvested.

Choose CPM ads, as they will serve quickest and most reliably, and will also keep your test ads away from the BookBub website, where you definitely don’t want them as the CTR will plummet —a nd CTR is the metric with which will determine the relative strength of a comp author.

Enter a bid that easily exceeds the higher end of the range BookBub is suggesting for your genre/author targets (i.e. $12 or $13 – it varies based on demand and targeting).

Check Your Work

All that’s left on this ad creation screen is to name your ad — e.g. “Get Rich Quick Scheme #5186” — and hit continue. You’ll then come to a new screen where you will review all your information before your ad goes live, and absolutely do that: check your image, targeting, links, the lot.

Sometimes you make a mistake, but sometimes the system switches back on a link to Amazon UK or Apple Australia that you had previously switched off. Other bugs have popped up previously — like the image associated with the ad disappearing. In short, if something looks funny on this confirmation screen, don’t ignore it. Usually this is a sign that something has gone awry. Simply click the button below to go back and fix it. Make sure to thoroughly check everything at this point because the ad will serve very quickly and spend your money just as fast too.

BookBub Changes

The beauty of having a resource page like this is I can both keep you informed of any changes BookBub makes, and also best practices surrounding them too. There are only two changes to discuss: Readers v Followers, and CTR-by-Author. And the short version is: don’t panic. Neither require any major retooling of the approach or anything. In fact, both changes are pretty good and could help you a lot.

Readers v Followers

BookBub has made a change to the ad creation interface. It happened just as I was proofing the book, and I was able to make minor adjustments to the ad creation guide just before formatting. I’d like to talk about these changes in a little more detail though as they are causing a little confusion.

In short though, this is not a very big change and you don’t need to worry too much. There is a much bigger change which you may not have noticed, and it’s a pretty positive one and will help with testing, but I’ll talk about that below (underneath the headline CTR-by-Author).

Back to Readers v Followers. You might remember from the book that I talk a lot about how many followers an author has, but also stress that your ad gets shown to a lot more people than that. If you target an author, you will get shown to their followers AND anyone who has clicked on their Featured Deals or Ads in the past. All that BookBub has done now is bring that total number to the surface so we can actually see it when creating ads.

Which is very handy, if a little confusing to some!

It also proves that the difference between the number of Followers and the number of Readers – as BookBub is calling the latter group – can be quite considerable.

For example, one author I sometimes target for my writer books is Gary Vaynerchuk. With just 596 Followers, you could reasonably think he might not be worth the trouble of targeting. However, I have been targeting him recently as he had a Featured Deal not too long ago. And as you can see in the below screen shot, he actually has 7,422 Readers you can potentially target.

This is all quite new, of course, so we’re still figuring out best practices, but my suggestion is to proceed as normal, follow the process in the book, but if you have someone with a low Follower account like Gary Vaynerchuk that you really would like to target as they are a strong comp author on Amazon or Facebook or elsewhere, then definitely check their Reader count in the ad creation interface to see if they would be a fruitful target, as in this case.

And if you want very general guidelines for a useful range of Readers to target, I’ve combined results from my own tests with some of your reports, and a range of 5,000 Readers to 50,000 Readers seems to be producing best results. Lower than that is tricky to get any kind of serving. And then you can get some joy in the 50,000 – 100,000 Readers range too. Above that gets tricky for all the reasons outlined in the book about authors with high Follower counts.

Just watch out for authors with a really spectacular disparity between Follower and Reader numbers, as this may indicate an audience which has solely been built up via free books and Featured Deals and they may be extremely price sensitive or otherwise tricky targets.

One last thing to note: don’t expect to be able to hit all those Readers at once. Not only will some of them be based in different countries and shop at different retailers which you may not be targeting with that campaign, remember that readers are not like us. They don’t necessarily check email every day. And they don’t necessarily open all their emails on the days they do check. I strongly suspect you’ll reach less than half those Readers on any given day, even if you target all retailers and countries.

Factor that into your calculations.


Unlike Readers v Followers, which sounds like a big change but is really a small one, this sounds like a small change but has quite a few major knock-on effects which should help us a lot.

It has been in beta testing since April or May – and I was playing with it then, and has subsequently rolled out to all partners now on a phased basis over the summer. You should have received an email from BookBub regarding this change.

In short, BookBub are just giving us more data. But that little change in the data we see has many useful applications.

Here’s the change: BookBub now shows you a breakdown of how each author performed when targeting a group of authors – assuming each author got enough impressions to generate meaningful data. I’m not exactly sure where that threshold is, exactly, but it appears to be around 1,000 impressions – although I’ve seen data generate for a touch less too. In any event, you want to see 1,000 impressions for any given target before you can judge it properly, so that threshold is fine and you should be shooting to exceed it anyway.

BookBub now tells us the CTR each author garnered, the effective CPM, and the effective CPC too. I think the latter two will be more useful with long-term campaigns – such as those pushing a permafree – so you can see if one particular author is starting to get expensive.

For most campaigns though, seeing CTR by author is the big gamechanger. Here’s why.

As you will know from the book, and your own experience no doubt, the testing phase can be tricky. And frustrating, until you crack it. One thing that is time-consuming is having to test each author individually. Previously, that was the only way to get a solid read on the CTR of that target.

Now you can test in batches, which might speed things up for you a little. I recommend being a little conservative with this, though. Pick maybe three to four authors with similar amounts of Readers and test them together rather than testing you entire list with one test ad. I definitely don’t recommend doing that for testing.

The above precaution should minimize the risk of a bigger author in the set gobbling up all the serving. You really don’t want that because smaller authors won’t get enough impressions for you to judge them properly.

Keep an eye on that serving! There have been reports of the serving being skewed to one author or another even when they have similar sized audiences, so don’t just slap up five test ads with four authors each and then walk away. Try one set first and see if that works better for you than individual ads.

For authors who have already completed testing: this is where it gets really useful. If you have your tested and proven list of comp authors, and you are running campaigns targeting them as a set, then you can check your performance data (even for some historical campaigns) on the Aggregate Stats page for the respective campaign.

You will see the full breakdown of CTR-by-author, and also the effective CPM and CPC and impression count for each author in the set.

This is so handy! Now you can see if one greedy guts gobbled up all the serving and you didn’t get very many impressions to a few of your smaller targets. You can pull them out and start a new campaign targeting them separately instead.

It will also flag any issues to you over time, such as one author getting very expensive, or another ballooning in Reader count (perhaps because of running some Featured Deals), and becoming a less useful target, one with poor CTR that is dragging down the whole set. You can then bump them down to your Silver set of authors, or excise them completely.

I hope you enjoy this new feature!

Optimization Advice

Optimization header

Were your first results crap? Well, you’re in good company because it’s the same for almost everyone. Even if you did stumble onto a good CTR, you’ll probably struggle to replicate it and you’ll certainly strike out when you try to scale. We need to do some optimization and analysis first.

Waaaaaaait, that sounds boring. Let’s talk about MONEY.

BookBub Ads are the perfect Get Rich Quick Scheme… for Mr. BookBub. And if you close your eyes you can almost hear him chuckling over the sound of his golden helicopter powered by author rejections.

The responsiveness of the platform is a definite plus, but it also means you can lose your money in an instant if you aren’t paying attention or if you don’t spend enough time/effort making sure your image is top quality and one that aims squarely at your target audience. Your targeting better be spot on too, because one bad author-egg is enough to ruin the entire Omelet of Success.

As I hope I’ve stressed repeatedly, margins for error on BookBub are tight. But that cuts both ways! One simple tweak can change a losing ad into a winner, so make sure you are testing everything in $10/$15 clumps on Amazon US-only — that is, until you are satisfied it is a good performer that can be ramped up and rolled out everywhere.

Don’t test an ad at Amazon Australia or Kobo Canada and declare yourself a genius when the first few hundred impressions are getting a CTR of 6%. Many half decent ads will garner that in those deal-starved markets. Stick to the tougher proving grounds of Amazon US until you have nailed everything down.

Enough preamble, let’s shake this money tree.

I’ve been running large-scale BookBub campaigns for well over a year involving thousands of dollars in ad spend. At this stage I have a solid idea of comp authors and the images are top quality, perfectly tickling the target audience in all its ticklish bits.

And yet, I still test every ad.

Before I roll out a new campaign, I still go through the testing process, because BookBub’s audience can be a bit of a moving target and an ad image you might think is perfect might need another tweak. So limit the budget to $10 or $15 or $20. Give it a quick spin at a higher CPM to make sure it serves fast, and you will get a fast read on whether it’s a loser or not. But what makes a winner?

There are tons of variables here when calculating ROI: the price of the advertised book, the royalty you are getting, whether it’s in a series or not (please say it is), and whether you are in KU and likely to get reads off the back of this promo too.

And then your tolerance of that ROI will be dependent on your goals: do you need immediate ROI? Are you happy to make it back in sellthrough to Books 2 and 3? Are you comfortable being more aggressive and going into the red this week to make it back (and then some) next week in KU page reads? Are you experienced enough to work that high-stakes system? Or are you allowing a small loss to get a new release moving and Also Boughts attached?

Everyone will have different answers to those questions depending on their situation, and those answers will likely change over time too, so I’m not going to do anything more than urge you to explore those questions in a systematic way before each campaign you run.

In very general terms, though, I can make an ad with a CTR of 2% or better work for me. I’ll need a higher CTR on a freebie, of course, and I’ll tolerate a lower CTR on a $2.99+ book, perhaps. However, when I’m doing the first test of a new ad pushing a different deal to a fresh crowd, I’ll be concerned if the CTR is as low as 2%, because that means the ad won’t have much shelf-life. I really want to be seeing 3+% on those tests to have confidence that I can run the ad for a decent stretch.

Word of warning: don’t judge an ad until it’s had around 1000 impressions. This is probably down to clicks and impressions being reported at different speeds, but sometimes an ad can have an artificially high CTR after a couple of hundred impressions, and you might think that ad is a solid gold winner, and then dump money into it.


You really need 1000 impressions before you can judge it properly. I should point out that the opposite can happen too. Other times, impressions seem to be reported before clicks, and you can deem an ad a loser, only to come back an hour or two later and see the stats have updated and it was quite a good ad after all.

So, wait for 1000 impressions, and then wait a little longer to be sure you have the final data. At that point you can decide to dump money in, or yank and tweak (or quit and wine).

This is a bad ad. That CTR of 0.88% means you are burning through your audience and interesting very few of them, and those clicks are coming in at $1.33 each. If you consider that even a good BookBub ad might need three or four clickers to get a sale, this is a way to go broke fast. This ad needs a lot of tweaking, probably both the image and the targeting, but sometimes changing one is enough. And often I prefer just to change one thing for each test, so I can hone in on the problem.

Don’t worry if your performance is even worse than that example above! I have had test ads with lower than 0.5% CTR. And sometimes it doesn’t mean ripping it up and starting again; a couple of simple changes can have a big effect.

In this particular case, I made a change to the image and that CTR improved dramatically.

Still not good enough. That 1.63% CTR is a big improvement, but I’m still burning through precious audience quickly at that rate, and those CPCs are too damn high. At this stage, one small little tweak can turn a nearly there ad into a sales printing machine.

This is a lot more like it. One tweak to the targeting (taking out an iffy comp author) dramatically improved the performance. I’m not burning through my audience pointlessly – something you don’t want to do when the effective audience is often limited. And that CPC is down at a nice $0.30 which is very workable indeed.

When your testing indicates positive outcomes, like in that last example, I will add some more money to the ad, but I’ll keep monitoring it — how closely will largely depend on the level it is being served at. And that’s generally related to the time of day.

If you remember from way back in Part 2 of this series, one of the critical differences with this ad platform is the way ads are served to readers. There is just one ad slot in each email, and the impression will only be served to that reader if they open the email. Like with any email send, most opens are clustered around the time of the send. Often you’ll see something like a quarter of the list open in the first hour, then half that again in the next hour, and so on akin to some kind of radioactive half-life yoke I’ll pretend to know about.

The time of day that BookBub sends out its emails varies, but the main push seems to often be between 1pm – 2pm GMT (8am – 9am EST). So that’s when ads will serve super fast. Midnight EST or 9am GMT, not so much. Keep this in mind.

I’ll watch the performance of an ad to see if it starts dipping. If it heads towards 2%, I’ll yank or tweak, or start a new campaign targeting my silver comps instead of my gold ones. Or stop advertising altogether, depending. Like this ad, for example, which has reached the end of its effectiveness.

It was a good ad, with reasonable click costs, that converted very well, but the CTR has just started to collapse which is spiking those CPCs. You need to zap ads like this before they eat up all your winnings (I got to this one a little later than I would like). Don’t fall prey to the marketer’s version of Gambler’s Fallacy. You need to ghost that fool and get with your hot new thing.

Complication Alert! It’s not always as simple as eyeballing the dashboard though, especially not if the ad is running at a high level of serving over a few days. That number on your dash is an average, and could be masking a recent dip (another reason to keep a close eye on things when more money is at stake).

For example, check out this ad. Great performance. Delivered a lot of clicks and converted excellently. So why did I snip it?


Normally, an ad’s performance will degrade over time, meaning the CTR will be best on Day 1, drop on Day 2, and drop further on Day 3, and so on. That nice-looking average on the dashboard may be masking a Day 2 performance which is below the redline. Or a Day 3 dip, as in this case. See:

So, yeah, that ad should be yanked, and ideally should have been done earlier in the day. But the dashboard isn’t really telling you that, unless you bother to click through and view the day-by-day stats, which you need to do periodically, especially on the higher spending ads. Just something to keep an eye on.

I hope all of the above suggestions helps you test and tweak and optimize your own ads, and generally monitor campaign performance. I absolutely recommend starting with smaller budgets and a large bottle of patience. It took me several goes before I getting half-decent results, and quite a few months of experimentation before I was getting genuinely good results.

But when you start hitting those good CTRs, you can ramp up your spend with confidence, as this is an audience of deal-hungry book buyers who convert very well. And that’s an incredible tool to have at your fingertips.

Tips & Tools

Tips & Tools header

I thought it would be handy for you guys if I collected all the resources I mentioned in the book in one place for you.

I can’t stress enough that the above guide is merely a convenience for those that have worked their way through BookBub Ads Expert already. If you try and run campaigns just based on the information above, you will probably lose a lot of money. There are many other steps you need to take first, such as identifying a BookBub-specific set of comp authors to give just one such example. If you don’t read the book, you won’t understand why I use CPM Bidding 95% of the time, and why that beats the pants off CPC Bidding. And you’ll never be able to scale up BookBub Ads to the point where it can deliver you hundreds of sales in just a few days. You can get the book here:

BookBub Ads Expert

The BookBub Partners blog which is a really excellent resource — particularly now that they have revamped it and made it a lot easier to explore it all by category — e.g. you can just view the posts and guides related to BookBub Ads only, if you wish. Dive in:

BookBub Partners Blog

Canva is a FREE browser-based image wrangler and promo graphic maker extraordinaire. I have a bunch of blog posts about it here on this website too, but check it out here and see why I love it so:


If you haven’t downloaded Amazon Decoded already, I strongly recommend you do that while it is still free. Whether you are exclusive to Amazon or not, you need to know how the Kindle Store works and what changes you need to make to your launches and promos to get the blessing of the ever-capricious algorithms. It’s free when you sign up to my weekly newsletter:

Amazon Decoded

There are lots of free 3D cover generators around, but this seems to be the best at the moment. Remember to save as PNG and then upload that to Canva. A JPG file won’t have the transparent background you need.

3D Cover Generator

Nicholas Erik has a very handy list of promo sites on his website which is both curated and kept up to date. I pretty much agree with all his assessments too:

Nicholas Erik’s Promo Site List

There are a variety of ways you can calculate ROI, if you are so inclined. BookBub has a handy spreadsheet-based calculator, and Nicholas Erik wrote a lengthy blog post on this here site detailing his own method. Check them both out here, if that’s how you roll:

BookBub’s ROI Calculator

Nicholas Erik’s ROI Calculator: How To Advertise And Sell More Books


I have opened up little space here now where you can ask questions and get help – just pop your question in the comment section below. I’ll answer when I can!

Spread The Love

Spread The Love header

If you enjoyed BookBub Ads Expert I would love a review from you where you purchased the book or at Goodreads.

Retailer links


And if you want to spread the love on Facebook and Twitter, here are some posts for easy sharing:

? NEW RELEASE ?BookBub Ads Expert is here ??? And you can also pick up my other writer books at a huge discount. Level up your author business today!? Download now (all retailers):

Posted by David Gaughran on Friday, March 1, 2019

166 Replies to “BookBub Ads Expert – Resources”

  1. Hi David! Thanks so much for your excellent book on BookBub ads and all the resources you’ve provided. I have a question on budget and bidding. I’ve just set up my first test ads for my comp authors and added a budget of $15 but the suggested BookBub bid range is $7-$15. If I want to do a maximum CPM bid of $17 (i.e. bid aggressively above the range), do I need to increase my budget to, say, $18 so it’s above my maximum bid?

    Thanks for your help!

  2. Hi Dave,
    Thanks for all the work you’ve done for the community.
    After buying your Book Bub book, I’ve one question that’s not addressed. Does time of day matter when you launch your campaign? I get my Book Bub email in the morning (I”m in the US.) And do my tests for the U.S. But don’t understand how I get impressions if I launch a test for email for Amazon USA in the afternoon or the evening.
    I’m asking because my results have been dismal and am looking for an excuse. 🙂

    1. Good timing! I was going to reply that I’ve launched books, run Facebook and BookBub ad campaigns, and done all sorts of promos on every day of the week and never noticed a difference (and generally feel authors pay too much attention to macro factors which are almost always completely overshadowed by anything you are doing/not doing on a micro level).

      But I don’t need to waffle on like that because BookBub themselves published some data just yesterday confirming exactly that. How serendipitous!

      1. Good information.
        Does it matter what time of day I run a test? i.e. Will a higher rate see it in the morning when they get the email versus if I run a test in the evening?

    2. Good timing! I was going to reply that I’ve launched books, run Facebook and BookBub ad campaigns, and done all sorts of promos on every day of the week and never noticed a difference (and generally feel authors pay too much attention to macro factors which are almost always completely overshadowed by anything you are doing/not doing on a micro level).

      But I don’t need to waffle on like that because BookBub themselves published some data just yesterday confirming exactly that. How serendipitous!

  3. Hey David! First, I want you to know that I’ve bought all of your non-fiction books and have enjoyed them all. I’m reading the BookBub Ad book now, and I have a question for you:

    I understand the TESTING phase and why I should test $10-15 per author comp. What I don’t understand (and maybe I’m just missing something) is the “image test” you talk about.

    So, I test Author A for $10-15 to see how my ad performs.

    Then, if the CTR is good (well over 2% since I’m going wide and offering a FREE Book 1), then I test another ad. This time tweaking the image.

    Should the second test also be another $10-15 (1000 impression) ad?

    I’m thinking it should be based on the idea behind getting enough data. But I don’t want to waste money if I don’t have to.

    Can you please clarify?

    Thanks, I appreciate your help and insight!

    P.S. I also bought and read Tammie’s Newsletter Ninja book based on your recommendation. Good book. I would have liked to see Book 2 in that series come to fruition. I saw it on Amazon but read that it was shelved.

    P.P.S. I tried to submit this question to your Contact Page but an error occured. It’s probably nothing, but I wanted you to know in case there’s a glitch on your site.

  4. Thanks for the encouragement.

    I’m perfectly willing to continue testing, but I just don’t feel like I have any more decent ideas to test. I’ve already listed out all my also-boughts as well every writer I know who it seems to me has a similar reader base and I’ve tested every one of those who fell into an acceptable range of reader numbers (which got me down to about 20 since a good many possible targets had very high reader numbers). I’ve also constructed a pile of different ads with identical elements to those ads you flagged as working well. Anything I do from here would only be random changes without any rationale at all to support them such as picking writers I don’t think are a fit to my market and making arbitrary design change.

    In choosing targets, I stuck strictly to your suggested ranges, albeit perhaps paying more attention to readers than to followers. About half of the targets on my test list were below 50,000 readers and half in the 50-100,000 range. I also tried a couple of targets that were well above your suggested range at around 150,000 just for the hell of it. There was no real difference in result among the lot of them. All of them produced essentially the same CTR’s and stayed in a consistent and it seems to me remarkably tight range of .3%-.6%.

    You suggested I post an ad image or two and I’d be happy to do that, but I don’t see a mechanism here for attaching images. Am I missing something obvious?

  5. I appreciate all the advice, and it strikes me as good advice, but I’m still getting the same crap result I did before I read it and now I’m really baffled.

    I carefully applied all your suggestions and have run 30 test ads for a $.99 book in the mystery-thriller genre that has done well for me on its own. I used all my also-boughts plus a few others that were within your suggested range of reader counts as my targets, and I used five different ad designs (two of which were nearly exact copies of ads you pointed out as effective). I got an awful result. Worse, the result is exactly the result I was getting before when I was blundering around on my own without following your advice.

    The best CTR I could muster was .68% and all the others were in the range of .25-.45%. What flummoxes me isn’t just the awful result, but the absolute consistency of the awful result no matter what ad design and what target I used.

    Please don’t read me wrong here. I’m not criticizing your advice. But it does seem to me that the consistency of my awful results with BB ads — both when blundering around on my own and when carefully following your advice — has to be saying SOMETHING. Maybe your solid non-fiction results with BB ads are difficult if not impossible to achieve in a competitive fiction genre like mystery-thrillers. Maybe BB ads just don’t work well in some niches. I certainly can’t make them work for my mystery-thrillers no matter what I do…

    1. I got pretty good results not so long ago with thrillers, and I know plenty of others who do well in that niche also. I have run huge campaigns in romance, SF, HF, NF, mysteries, crime, and humour and gotten comparable results in all of them.

      If anything, fiction tends to be much easier than non-fiction.

      If you want to post an ad image, maybe we can help. Also let me know what size targets you were aiming at (in terms of Readers and Followers both).

      You might also want to check out this video on some recent changes over at BookBub (I will be updating this page but I currently can’t make updates to my site because of some technical issues right now!):

      Finally, the frustration you are feeling right now is not that unusual. Testing is the hardest part of BookBub Ads – and things get much easier once you get over that hump. Some authors get lucky and sail through the testing process with minimal difficulty (in fact, I know a thriller author who just did that). However, most will have a frustrating time… until it clicks.

      Unlike other platforms, most of the difficulty is right at the start with BookBub Ads – but I guess that also does a bit of gatekeeping as well and prevents the kind of flooding of the platform we have seen with things like Amazon Ads. But if you persist, and get through testing – frustrating as it can be – you should be rewarded.

  6. Hi David, I’m making my way through all of your books! Like Amazon Decoded this one was great. Now I’m preparing to test BB this weekend. I was wondering if you had any recommendations for tracking beyond clicks. I don’t want to use my Amazon affiliate code, because I fear losing my affiliate account. Do you have any suggestions to get an idea on the sales from BB ads? Thanks and have a great weekend.

  7. I’ve been basing my BookBub ads solely on Reader, not Follower numbers, from the get-go, and meeting with excellent results. Now that I’ve gotten my ads in order, it’s all rolling! Thanks for all your excellent guidance, David. I’m singing your praises to all.
    Off to write those reviews now…

    1. Well, I don’t usually like doing pre-orders so I might be the wrong person to ask but my general feeling is that not all readers like pre-orders so you are automatically cutting down your audience and then trying to make the numbers work – which is a tall order, unless you are happy to go into the red for strategic reasons.

      I would also worry that you aren’t going to get a clear read on your targets if you use a pre-order for testing. Pre-orders can struggle with conversion (because some readers just won’t pre-order books – I’m one of them, and I can tell you I can count on one hand the amount of times I’ve done it). And also CTR will take a hit too because some readers will opt out of clicking also. I think it’s a bad idea to introduce a big variable with a question mark into the testing process.

      I personally prefer not doing pre-orders and concentrating my firepower during launch week for maximum impact, but I have seen people using ads and pre-orders in interesting ways. One author uses pre-orders and ads to test variations on her covers and ad graphics before the book goes live. I think that’s a very interesting idea, but just note that to make this kind of thing work for you, *relative CTR* should be the only metric you judge things on – ignore conversion completely.

      1. Thanks!

        Based on this, I picked a DIFFERENT title that’s already out…and have started comp author searching (after bumping the price down to $0.99).

        Actually KNOWING where to look in the metrics and having some sense of what’s happening (after finishing Bookbub Ads Expert in 48 hours nevertheless) really makes me stoked about BB.

    1. I haven’t tested the audiobook ads yet personally – soon! – but I’m hearing that discounted audio does best of all, which makes sense as it’s such a deal-hungry crowd generally. I think full price audio will struggle with BookBub Ads, but I’ll have to get hands on it before I can really see. Great to have the option though!

  8. Hey David – Ran my first two comp author tests this weekend with 1.76% and 3.06% CTRs on a free reader magnet, with a total conversion rate of 18.6% (actual signups). What’s a favorable CTR for pushing a reader magnet? And what would you consider a good conversion rate on something like this?

    1. Using BookBub Ads to generate leads/sign-ups is not permitted – so if you have an running to your reader magnet, I recommend axeing it.

  9. What would be a decent conversion rate on a bookbub ad? I’m running an ad at roughly 3% CTR on a Freebie book one with four follow on novels in a countdown deal. I’m seeing roughly 15% in purchases vs clicks but a huge spike in the freebie. I figure there’s a few KU readers planning on ignoring the deal and just reading the rest through the KU route. That’s better for me, since the page reads pay better than the .99 discount royalties.
    I suppose any way I slice this, I won’t know how to calculate conversion until the dust settles…
    Anyway, your book was an amazing intro into Bookbub ads and it forced me to discipline myself and do proper split testing for once.
    Kind of getting a dopamine hit every time I check a new iteration and see a half percentage lift in CTR!
    Thanks for sharing your knowledge, David!

    1. Conversion can very wildly. Even leaving aside the quality of the traffic (are you targeting the right audience? is your image resonating with the right readers?), there are many conversion elements on your page, from cover to reviews to blurb to sample to formatting. If we assume all that is optimized, and I’m running good images to quality targets then I can see conversion reach even as high as 25%, or more. Conversion can be really great with BookBub Ads.

      That’s for a 99c book, because with BookBub, often a massive conversion element is price. You can get great CTR on a full price book… if you don’t mention the price. But your conversion could be as low as 5% meaning only super cheap clicks are going to work for you. Conversion can be higher on a freebie, and it’s often lower than 25% on a 99c book – in my experience. But when an ad is really cooking, I can hit that.

      1. For now, the sales increment is paying for the ad so I’m inclined to keep iterating the creative. I’m pretty sure this will also give a respectable bump in reads but that will take days or weeks to pan out.

      2. Not quite sure how to treat numbers for conversion, now that I think about it.
        I have 76 sales from 145 clicks but that’s because of your ninja trick – putting an entire series on sale at once. Folks take book one for free and the following four are at 99 cents for a couple more days. I was dividing the sales by four for conversion, which I think is the more accurate way of looking at things. The sales number keeps incrementing by four at a time (yes, I’ve been refreshing the view often enough to know because I’m as nervous as a ferret in cardboard underpants…)
        Adding the number of days remaining for the sale has bumped the CTR by 0.42

    1. Hi Eric – that change was made just as I was publishing the book and I made a minor reference to it in Chapter 10 – but there is much more information (regularly updated) above on that. The second, bigger number is called “Readers” by BookBub and it includes all your Followers + everyone that clicked on your ads or deals previously. More above…

      1. Coolness! Thanks again!
        Hey would you like to jump on a podcast with me?
        Would love to introduce you to my audience at!
        This is GREAT stuff, and YOU would look absolutely smashing above Steve Scott and Dave Chesson 🙂

      2. Sorry for the slow reply and thanks for the invite. I’m not doing any interviews/publicity at the moment as I’m catching up on overdue WIPs but maybe in the new year?

  10. Hi David,

    I’m unable to see the “recommended authors” section on the BookBub site that you reference on page 25 (chapter 3)–did they do away with it?

    Or is it an I’m-in-Canada thing? (I’ve tried to change my location in my settings like someone mentioned in the comments above, but I don’t see a way to do that.)

    Do I need to be in a “reader” profile to do/see these things? I’ve been using my Partners account.

    I’ve been trying other author targets, but thought the BB generated “alike” authors might save me some time and money. 😉


    P.S. I’ve been finding that $8.00/ad test at a bid of $11.04 in contemporary romance can usually get me 1000 impressions in case that helps anyone. I’ve also been finding that it’s easier to go the old way and keep the authors separate–so 1 author per test–rather than trying one ad where I lump several into one ad and wait for BB to give me the data on each author. (I have to spend more in the lump-’em way, it seems, and fiddle with turning the gobbler authors off so I can get enough delivery to the other authors so I get stats on them to see if they are any good.) Also, David’s numbers on Followers/Readers seems about dead-on for me with testing. Now… to figure out how to turn this money nozzle all the way on so it rains money on my face. 😉


  11. Hi David, thanks for all the help you’ve provided! I wanted to get your insight on inconsistent CTR. For example, I’ve targeted an author and gotten great CTR (2% or more), but only for a day or two. Sometimes the CTR will crash over the weekend. When that happens, do you usually remove that author from your gold list, or do you chalk it up to a “bad weekend”?
    Also, I’ve talked to people who swear that different ads work for different authors, but if I understood your book correctly, you typically try to keep the ad the same from author to author. What are your thoughts?

    1. When I’m testing, I keep the image consistent as I iterate the authors, that’s the only way to do a true split test – otherwise you don’t know if it was the change in image or author which changed the result.

      When I’m running a proper campaign, I throw all my best authors together, so they will all get the same image.

      I don’t see the kind of variance you are talking about so much – at least compared to other ad platforms, BB is a lot more consistent for me than the others. My gut says that you might see more such variance when one element or another isn’t fully nailed down, perhaps, or maybe when dealing with trickier multi-genre authors.

  12. Hi David,
    I’m working through your book and am starting some testing tomorrow. But I am running into an issue with my targeting. I’m being “forced” to stack many authors in my targeting in get a defined audience, even when choosing authors with over 20K readers. This is not like your advice and examples at all and I’m wondering how it will affect my results. Is this the result of a change at Bookbub?

    1. This definitely sounds weird. I’m doing some testing right now and not seeing this at all. Are you running ads just to the UK or just to a smaller retailer or something? This shouldn’t happen for Amazon US. Or have you not put in your link at all yet? If you give me an example of a couple of the authors, I’ll see if I can replicate it in my own interface and see what’s going on…

  13. David, I just took your survey and responded, “Short hair don’t care.” Translated, that means, “I look for your emails no matter when you send them.” They are all packed with useful information. I’ll admit though that one I’m particularly anxious to see is the next installment on AMS ads. Using your previous guidance on creating tons of keywords built around authors who share an audience was an immense help that I followed with positive results. Now I’m hoping to amplify that progress with your next set of wisdome. Thanks for sharing what you do.

      1. Got it. I’ve followed your advice as well as that of others (yours has performed best) and was building nicely, and then yesterday, impressions dropped precipitously. There were new ads, old ads, optimized keywords, new keywords, keywords from WordMerchant, RocketPublisher, from hard work tracking down comp authors and using keyword expander — 181 in all. I had some that did nothing other than generate impressions, but many that performed well. That came to a screeching halt yesterday as impressions plummeted with no detectable reason. I’m at a loss.

    1. I know some swear by it but I haven’t properly tried it. I grabbed a free trial, but then forgot I’d signed up as I had some Stuff to deal with which required all my attention. So I never really gave it a proper shot, but I will say this for their customer service: when I didn’t realize I had transitioned to paid and emailed them to cancel, they PROACTIVELY offered to refund the monthly fee which was just great customer service. So top marks there.

  14. David, when you get a chance, would you please discuss/amplify what the CTR rating under the author’s name when seeing them on the BB ads when creating the ad. Does that refer to the rate at which an author is clicked on when seen when others use him/her as a comp author or the rate at which that author is clicked on when seen advertising his/her own books?

    1. I asked BB about that, it’s the CTR of your own personal historical aggregated efforts for targeting that author. They were talking about limiting it to 30 days but it seems to go back further. Don’t know if there is any recency weighting or not.

  15. David (or anyone else who knows), I just opened an account with MailerLite. As you know there or questions to be answered before the account may be opened. One of them implied that subscribers must re-submit permission to send them correspondence. Been there, done that. Don’t want to do it again. Besides which, sorting out who is in the EU and who is stateside would be tedious. Any insights?

    1. No, you don’t need to ask everyone to resubscribe. If you really want to cover yourself legally, I think a small amendment to your privacy policy noting Mailerlite is your provider now, and then briefly mentioning you have switched providers in your next email is more than sufficient.

      I think we as authors have to parse a lot of this advice quite severely because often they are speaking to companies who may have been super aggressive about putting people on their list who didn’t explicitly consent to receiving a newsletter – such as people who gave their email to get a digital product delivered. We are fine, as we generally aren’t very aggressive with sign-ups, we usually have informed consent, no bait-and-switches, use double opt-in, and so on.

      (Not a lawyer, etc.)

  16. For anyone following the comments, I’ve just updated the Resources Page with some more information on the two changes since BookBub Ads Expert was published: Readers v Authors, and CTR-by-Author stats.

  17. Hi David,
    Apologies if I missed the answer to this question above. Just tell me its there and I’ll look more carefully.
    I have been having great success (for me) advertising my new release to audiences other than Amazon. After initial tests on the US site, I turned off amazon and put it out to Apple, Kobo, Google Play and B&N readers in US, UK, Canada & Australia. I’ve segregated the reporting by vendor, not country. For the past month, my CTR for this group has been steadily climbing and I now have most vendors CTR between 2.5% and 7.4% on any given day. My problem is that I seem to be unable to scale this ad up. Over time i have increased my daily bid from an inital $5 per day to $15, and I have set the bid maximum bid twice and it is now close to the max suggested. The best I can get Bookbub to do is spend maybe $3 or $4 per day, nowhere near my daily ceiling. I am getting steady sales at all of those vendors that I’ve not seen without advertising, so my conclusion is that it is working and people are buying. The sales are starting to tail off a bit too.
    Any ideas what is happening here? (My genre is Sci-Fi and I am advertising the book I released last month at full price.)

    1. How many Readers do your target authors have? Just on the face of it, it sounds like you are maxing out the audience – because there are far less users outside of Amazon and outside of America. What you are probably seeing the next day then is the ad showing to that audience for a second time. BookBub will show your ad up to four times to the same person (actually, someone told me recently this had changed to five). So, again, just guessing here, but your ad probably kept that pattern up for a few days then mostly stopped dead. The solution is to target more authors or add in America or Amazon (and if performance isn’t good enough there you probably have a deeper problem with your targets and/or your image that you need to work out with testing – otherwise your audience will always be quite limited).

  18. I’m finding the BookBub Ads course extremely helpful! Just wanted to confirm that the number we should be looking at to target authors who have between 500-20K of these are the “followers” we see if we sign into (as if we were readers), NOT the “readers who are interested in this author” we see when we are on the BookBub Partners site and are adding an author to our promotion …? For example, Emily Kimelman shows 11,459 Followers, but 3,214 “readers who are interested in” her. Thank you!

    1. Sorry to be an idiot–finally found the answer to this above, and also realized that the first typo correction I had entered was just delayed in posting. I vow to do better in the future. <: )

  19. Hi David,

    Just finished your book, which is great and so informative. I’ve started testing out comp authors for my superhero series, which didn’t yield great results (the best CTR was 1.04%). Then I tested out a new ad for my space opera series, which yielded a CTR of 0.36%. So far I’ve been creating the ads myself, but I’m thinking maybe I should commission these instead for better results. are there any artist you can refer that create BB Ad images? Thanks in advance?


    1. Hi Charles, any graphic designer will be able to do it – it’s a pretty basic job for them, and many will charge in the region of $20 or $30. Your cover designer might be the best choice, if available, as they will have all the layered files and will be able to generate ad assets for you in a matter of minutes. One of the best approaches in the medium term is to commission ads and other marketing assets – like Facebook Cover Photos, Facebook ad images, BookBub ad images etc. – while you are ordering the cover. Often the designer will bundle them for free or a nominal fee. Longer term, you can keep playing around with Canva/Photoshop yourself and trying to ape the techniques of your designer. Maybe get them to give you the “building blocks” of the ad – price tags, 3d covers, backgrounds etc. – and then you can just drag and drop them in as needed. This is a great way to learn actually, but at the start you won’t be good at things like text and composition, so commissioning your ads and learning in the background yourself at the same time can be a great approach.

  20. Hi David, Thanks for your incredible book and resources! I’ve made my image, I have a list of potential authors, and I’m currently trying to find the number of Followers for each author on Bookbub. I can’t find a search function anywhere on the Bookbub website (and the author pages are pretty useless with no links to any of their books). I am in Canada, so I was mindful of your advice that Americans might see a better version, so I logged in using a proxy for the US, but still nothing. I can (laboriously) search the Bookbub site using Google and book titles, but that only gets me maybe half of the authors. I can find reader numbers (when I set up ads), so I suppose I could use those numbers in lieu. Any advice for this step, or ideas of where I can look for advice? Google is coming up empty.

  21. Hi David,

    I’m setting out my strategy to launch a first-time series. To prepare for that, I’ve been reading the advice in the various forums and books like yours on BookBub (BB) ads. I have a regular account at BB, reviewed some books, and have followed several authors in my genre – YA Sci-fi. I subscribed to three of the BB email lists – Young Adult, Science Fiction, and Thrillers so I could watch how the BB ads worked.

    Several times now, they have served me the exact same ad for four days in a row. Worse, it’s an ad that I would never click on anyway because they are never relevant to me. The last round was a fantasy book. If that ad’s purchaser is paying CPM, and many of their ads are equally mis-targeted to the wrong recipients, that would go a long way toward explaining why it’s so difficult to get a decent CTR at BB for their CPM ads. They keep serving the same ineffective ad over and over!

    Can you ask BB why they do that? Keep sending the same ad to the same person over and over? I know you said they serve the same ad to the same person four times, but this seems like a terrible waste for CPM ad buyers. If I buy their CPM ads, I’d want them to serve it only once to any person before moving on to the next. Between the three authors I want to target, there is a potential reach of 100,000 people. But if they serve my ads to the same people four times, won’t that reduce the reach and the CTR by 75%? Why are they doing that?

    Thanks, and keep up the good fight!


    1. Hi Tom, I’m pretty sure I talk about this topic a good bit in the book – ad fatigue and frequency caps. Maybe in the very last chapter?

      Anyway, there is a frequency cap of 4 for each campaign. It used to be 3 and they nudged it up to 4. I’m totally fine with that frequency cap, as the structural nature of BookBub – ads at the end of an email – means that I’m sure that not all impressions were actually viewed by users. Not everyone scrolls to the bottom of the email or otherwise notices the ad, but as soon as that email loads and the pixel fires that counts as an impression on your dash.

      So 4 is totally fine.

      Also keep in mind that – this system mandate frequency cap aside – this could be advertiser driven. They could be testing, they could be starting new campaigns every day, they could be burning through cash or experimenting or otherwise maximizing for reach for whatever reason. BookBub doesn’t tell you how to spend your money.

      Finally, and this could particularly be a factor in your case, it could just be personal to you, to an extent. You said “I subscribed to three of the BB email lists – Young Adult, Science Fiction, and Thrillers so I could watch how the BB ads worked” which indicates you are a new BookBub user. I presume this also means you don’t follow very many authors, if any at all. The whole ad system is based on readers Following authors (and subbing to genres) and the whole targeting system is based on advertisers wanting to hit the eyeballs of someone who follows Author X.

      So my guess here is that you either have an unsophisticated/scattergun advertiser just targeting by genre, or that your eyeballs are super cheap to reach because you don’t follow many authors, and someone is just running low bid ads targeting the genres you follow as a whole and seeing who they can scoop up. Or to put it another way, the system only has a crude idea of what you like right now so you are only seeing cruder ads.

      Nothing to worry about either way.

      1. Nope.

        I follow a tight list of authors:

        And I reviewed a tight list of books:

        Both so I could ‘train’ BookBub about my preferences so I could watch how their ads perform. They should know how to serve me relevant ads. If the ads I’m getting are CPC, then fine, the 4 times served doesn’t cost the advertiser anything. But if they are CPM, then the BookBub ad system is only 25% effective because it’s wasting 3 out of 4 ads.

      2. There isn’t any machine learning or extrapolation or Also Bought-like expanding going on with the targeting. It’s pretty “dumb” in one sense in that it’s very straightforward. You will see an ad if:

        (a) you Follow that author (and the advertiser is targeting that author);
        (b) you clicked on that author’s ads or Featured Deals before (and the advertiser is targeting that author); and/or
        (c) you subscribe to that genre (and the advertiser is doing very broad targeting and simply targeting the whole category).

        I don’t see anything to worry about here.

        As for the frequency cap, I actually think it’s pitched right, and actually conservative if anything. When I’m running Facebook campaigns, I will switch them off or expand the audience when I start hitting ad fatigue, and for me that’s usually between 2 and 3 servings.

        I’m comfortable stretching that until 4 with BookBub as – and this appears to be the crucial point you are missing so I’m stressing it here – not all of those impressions are actual ad views, and they are much less likely to be ad views than on somewhere like Facebook where the impression counts when the ad is served directly in news feed etc. (Audience network aside, just simplifying here.)

        With BookBub, a typical user could have 5, 6, 7, or 8 books in their email. THey might not scroll down the whole way. They might get to the first or second book and buy something and not return to the email. Because of the ad’s position at the very bottom of the email, and because of the way impressions are tracked, the impression counts as soon as the email is opened – even if the ad is off screen and never viewed. So the “true” impression count is likely a lot lower.

        This isn’t just theoretical. Extensive testing has shown me that some audiences will tolerate a lot higher – 8 viewings, even up to 12 sometimes (although an image refresh beforce copying and restarting a campaign often helps – in case anyone tries it).

        Short version: you have nothing to worry about here either.

      3. Just to be clear: there is NO relevancy in the auction. It’s not a system like AdWords (or Facebook or AMS to a teeny extent) where some kind of ad/campaign/advertiser quality score gets factored into the auction or bids. It’s a straight money auction.

        I think the system would be very much improved with relevancy factored in, and I make the case for it when I can with BookBub – and they aren’t against the idea, I got the impression it has been thought about – but right now it’s a straight bid auction.

        So the only penalty for being a bad advertiser is the crude one of losing your shirt and running out of money and your ads stopping. That won’t stop someone with more money than sense, of course, or those seeking out pockets of value by essentially spamming broad CPC campaigns at low bids to see what little bits will serve where and when and really trying a scattergun approach to get cheap clicks, rather than the approach I recommend.

        But that’s a meta concern about the long term sustainability of the system and user trust in the ads. It has no impact on how your ads are served.

  22. Hi David, hope you’re enjoying vacation – no rush on a response!

    About to dip back into ads. As discussed above, there were some hiccups moving from individual authors who tested well (CTR 3.5%) to when grouped, the CTR plummeted. Any fresh ideas on that? If not, is your advice to not group them?

    Note purely fyi: on those authors I did group – the dashboard was unable to give me splits, saying, “not enough data on this author.” Not sure how much data they need to provide those splits.

    Thanks as always! HN

    1. You’ll see in the comments above that this appears to be an issue cropping up for people recently. I would recommend contacting BookBub with all the relevant info – they are curious to hear more about it, so they can determine if there really is an issue here.

      One possible explanation might be ad fatigue – i.e. if you have tested Author X a lot, and she doesn’t have many Followers/Readers, the audience might be tapped out when you hit Author X as a group.

      But there could be some technical issues going on with targeting authors as a group. I’ve noticed some curious results myself, and I’m not sure how even the serving is right now.

  23. Still a little stymied about the transition from testing to working ads. Lately, when I run test ads on one of my “best” comp authors on Amazon US, I have gotten 3.45% CTR for a new release advertised as being $2 off regular price. When I take the ad live across vendors, his CTR plummets to 1%. Aside from the additional vendors, I only added my own name as an author, but I have a hard time imagining my meagre 5000 readers would influence things that much. Any ideas what is happening?

    1. I’m on vacation so a little slow to respond ATM, but can you tell me the Follower/Reader numbers on the author you are targeting (aside from yourself)?

      Personally, I’ve seen some strangeness lately when targeting groups of authors – I’m not sure the serving is being done very evenly. Might be related to the rollout of some new features, I don’t know for certain.

    2. But, either way, I’d recommend running another test on that author target on its own and just get a read on it. Might have been caused by adding your own name, and the serving going askew, and your own name (or the other) having a dip in performance due to saturation. Or it could be something else – that’s just my quick take based on limited info.

  24. Is there an issue seeing an author’s reader numbers from the UK? I’ve read the book twice (made 6 pages of notes!) and started on a list of comp authors today. I’ve got a reader BookBub account so I’m in that BUT for some reason I can’t pull up any author’s profile. I get the three dots of doom on a loop. If an author is running a promo when I click their book promo page on BookBub I can see their follower numbers in the box lower down on the page, but clicking on their name still does not take me to their profile. So I logged out and back in under incognito mode (Chrome) and this time I CAN see their author profile page BUT still no follower numbers.

    I can get reader numbers if I go into partner dashboard and pretend to make an ad.

    Obviously I would rather get the follower numbers to begin testing.

    Is this – a known glitch – something I can get around with another browser or a VPN – something to ignore and try again tomorrow – just my bad luck – or a new thing which means I’ll have to use reader numbers?

    1. Thought I’d pop back with an update. A VPN did not help me to see follower numbers, nor did switching browsers – I can still only see the larger and less accurate/helpful reader numbers. However I’ve been noting down the follower numbers that I can see when the author is running a current promo so I’ve been gathering intel slowly that way.

      I also just wanted to tell you that the advice in your book about images was very helpful.

      I’ve just run my first four tests which blew through $15 for each author. CTRs of 2.24, 1.82, 2.23 and … 4.41%. I reckon I know which author I’ll be targeting to test images now 🙂

      1. Sorry for the delay in replying – I’m on vacation and only have intermittent internet.

        I think it’s a good idea to ask BookBub what they suggest here, as that will give you the opportunity to provide them with the strong feedback that international users really need the full version of the site. I had assumed the rollout to them was imminent 6 months ago, but that obviously hasn’t happened yet for whatever reason.

        In the interim, you have two choices. You can simply forget about Follower numbers and just go by Readers – we’ll all probably be transitioning to that metric anyway, so you’ll be ahead of the curve instead of behind it for once! Check the stuff I wrote above on Readers (under the heading BookBub Changes) and also the comments here – there is discussion on what numbers you should be shooting for. That could be enough to get you over this hump.

        Alternatively, if you wish to stick to reckoning by Followers, you do still have some options. I think. This is going off my memory of what the site was like for international users back when I had that older version too, so I can’t be sure they haven’t tweaked it since or that my memory isn’t faulty. But there is – or was – a way you can still see Follower numbers. It was kind of convoluted, but it worked.

        …actually, let me message my friend in Canada as I don’t want to get the steps wrong, it’s a bit fiddly.

      2. I wrote Bookbub about this, and got an answer:

        Thanks for reaching out! It looks like you’re writing to us from Canada. At the moment, we haven’t fully rolled out Author Profiles in our non-US regions, which means that the number of followers an author has isn’t available — I’m sorry for any inconvenience here!

        If you’d like, you can switch the region that your account is in through your account settings. This will allow you to see all the details on an author’s profile. I will note, however, that doing so will affect any promotional emails you receive through your BookBub account; not all of our deals are sent to every region, so some of the books you see in your Featured Deals email may not be discounted in your region.

        I switched to US, and it worked great! Easy and quick.

      3. OK so there is another workaround that may (or may not) suit. It’s pretty nifty if it does suit though: as per Emma below, you can email BookBub and request that your account is switched to a USA account, which means you will see the full, new site when logged in, with Follower numbers and everything else they now have going on. The only downside is that all your daily deals will switch to US daily deals, but that might be minor enough, depending.

  25. Hi there! I’ve been listening to podcasts and reading the Reedsy course and now I’m diving into your book. One practical issue I’m having – I cannot get a crisp, clear cover when it’s reduced down to such tiny dimensions. Any suggestions on how that’s done or where to go? (I don’t have Adobe so I’m hoping you aren’t going to tell me to use that! )

    1. Hi Sophie, you shouldn’t lose any crispness whatsoever when reducing a cover – normally that only happens when increasing something beyond its orignal size. I recommend using a free tool like Canva to make your ad images – where you can easily adjust the size of the cover etc. It’s all drag and drop. Here’s my tutorial:

      But if Canva is too much for you, then you can check out BookBrush which sounds a little simpler:

    1. Still trying to get a good handle on where the lines should be, but the numbers are much higher in both cases. I checked out about 100 writers and the variance between Followers and Readers is huge. At the lower end of the scale, some authors will have 2x Readers over Followers. At the higher end of the scale, some will have as much as 20x Readers over Followers.

      Which makes it incredibly difficult to give a range, but we need something to work with, so I’d probably suggest in the region of 5,000 Readers to maybe 60,000. That latter number could stretch up to 100,000 – it’s hard to say without more testing – but I think things generally start getting choppier north of 40,000 for most targets, and that ramps up from 60,000/70,000 upwards.

      Although I’d love to hear anyone else’s experiences on that, as there could be some genre variance at play here. For example, in romance – which often plays by its own rules anyway – I see a much greater disparity between Followers and Readers. Probably because those authors built their BB audience off the back of multiple, huge free Featured Deals. Which can skew things considerably…

  26. Hi David,

    I’m doing a free run for book 1 in my series so I’ve used that to do a bit more testing – I’m still struggling to get the targeting right for my ads. If I get results of around 1.5% CTR on a free book ad, does it still make sense to label that author as a ‘maybe’, looking to tweak the graphic until I get a higher return, or should the ‘maybe’ threshold be higher when testing on a free book?

  27. Hello David,

    First off awesome resource and amazing book. I’m halfway through it and learning a lot.

    I’ve been doing some research/homework and for one of my genres…well’s let just say that the authors in said genre, myself included, don’t really exist on book bub, as in none of them seem to even have triple digit followers.

    In such a case should I even consider doing book bub ads? Or is it better to just write that off and save it only for genres where there are authors with followers in the goldilocks zone you mention in the book.



  28. David, I know you would addressed this somewhere, but I can’t remember where. When we are creating ads to target authors, the add interface shows the number of readers rather than the number of followers. Where do we go to see the number of followers rather than readers? Did it used to be within the ad interface, or am I just imagining that? Thank you.

      1. Thanks, Edward. I wasn’t sure if it might have once been in the ad interface (or if I was just ridiculously unobservant). I do recall the “discover authors” page being discussed, so I guess we just have take that extra step when creating ads.

  29. Hi David,

    I went ahead and did the BB ads for my free Kobo promo. The authors I chose along with the category made it necessary to use three authors for each ad, otherwise the ads wouldn’t have run. I’m running the ads at $14.11 CPM, and the graphic says “FREE”.

    In about 24 hours, my CTR for the first ad is Kobo US- 2.4, UK- 6.45, AU- 6.45, CA- 5.23, overall- 621 impression, 28 clicks, CTR 4.51.
    the second ad, started at the same time: Kobo US- 3.16, UK- 5.26, AU- 9.09, CA- 10.11, overall- 225 impressions, 15 clicks, CTR 6.67.

    Would you consider these successful? I realize the odds are higher that I’d get better CTRs with an ad for a free book. With these numbers, would you keep them going? Should I choose three or six more authors and do one or two more ads? I’m hoping to make the money back in sell-through to the second book.

    1. Those CTRs look very good to me, but how is conversion looking? While I regularly get conversion rates of 25%+ on Amazon it can vary a lot more on the other retailers, so I’d try and get some sense of that… if possible. But, that question aside, these look good to me.

      1. It’s impossible to tell how effective they are since my book is one of only four highlighted on Kobo’s “Free” page.
        I wanted to mention that my sub-category, Gothic fiction, wasn’t available as a category. So I searched other authors with that sub-category and made sure I targeted them, then I put the word “Gothic” in my ad. I think that might have something to do with my success.

        Thanks, David!

  30. Hi David, I really was at the end of my mind when I last wrote: I referred to you as Steve despite that we had already corresponded. Sorry.
    I wanted to let you know that I managed to clear my “Also Boughts” of the books that didn’t there. I culled out words in AMZ Ads, halted those not producing that might have contained offending keywords, and of course advertised on BB to comp authors. I also went back and reviewed much of your material and narrowed the comps down more specifically to my genre. I’m greatly relieved. I was about to unpublish the book, wait a couple of days, and re-publish it, but now I won’t have to, which is a relief since the launch for the next book in the series is due out in a few weeks.

    I also managed to cobble together a stable of authors that are doing well for me in all markets. I have then on drip campaigns now, preferring to keep my powder dry for the next launch since it is so close. The overall effect is that all three books in the series were lifted up. Thanks.

  31. Hi David. Great book, thanks! What are your thoughts on using the built in ‘A/B Split Testing’ feature for assessing ad performance during the optimisation phase? As opposed to two separate ads? Thanks.

      1. Sorry, my bad. I was getting confused with the FB ad platform. I thought BB had a built in split test function, but it doesn’t.

        Do you run split test ads on separate days, or can they be run at the same time (or would that mean that they’re effectively competing against each other?).

  32. So I ran a 7 day 99 cent countdown deal and tested a bunch of authors. At first I was trying to get a decent graphic and try #3 seemed to work. When the dust settled I had spent around $200 and identified 4 authors between 2.1% and 2.7% and one closet to 4%. Three of the five are small enough that they had to run a full day with a $15 bid to get me my 1000 impressions.

    Because I had decent and increasing sales for three days in a row (from 7 up to 14) and found myself on the last day of my countdown, I decided to use those five authors on that day in a single ad and see if I could get over 14 sales and maybe tickle those 4 day KU algorithms. Obviously with the price going back up after that it wouldn’t be ideal, plus with those numbers my rank never got above 13K and that might not be quite enough to do it. But it seemed worth spending a small amount to see what happened.

    What happened is the ad tanked – worse than 1.0% CTR. I killed it and removed the one author who had a significantly larger readership and tried again. Still a terrible CTR. With the small amounts of impressions per author I had done during testing, it doesn’t seem likely that it was anywhere near saturated. During my testing I had 20K impressions total, while the reader (not follower) numbers for the five authors are many times that when combined.

    This has me scratching my head. It reminds me of AMS where you pull good keywords out of a performing ad and into a new one and they don’t work. Others have described something similar in the comments where. Might BB be changing things up? Or is it just bad luck? Note – I’m not saying I couldn’t improve my graphic at all, just that I have one that was working pretty well so that shouldn’t be the issue.

    1. A few different people are reporting the same thing – enough to make me wonder if there is some weird issue here, because it doesn’t make any sense on the face of it. I’m going to talk to BookBub about it.

    2. Can you share the Follower and Reader counts for the five authors in question? I wonder if saturation is indeed a factor here.

      1. Here are the counts:
        -61 followers, 5069 readers, initial CTR 2.52
        -887 followers, 9945 readers, initial CTR 2.52
        -249 followers, 14495 readers, initial CTR 4.20
        -547 followers, 14702 readers, initial CTR 2.01
        -1939 followers, 71791 readers, initial CTR 2.18

        In reviewing, I realized that the first one only got 700 impressions during the test. Also, the last one I had tested with a previous graphic as well. But overall this was a good base to try scaling up a notch. Even if we assume there is some overlap between readers, 5700 impressions should not be enough to saturate this group for a $25 budget test. Even removing the last one as I tried also, there shouldn’t be saturation. The combined 5 ad was 1.1% on 1300 impressions before I killed it and the combined 4 was 1.2 on 700 impressions before I killed it.

      2. Yeah you could be facing saturation issues quite easily – that’s why I recommended going after authors with much more Followers. The total number of Readers in the first example is just over 5k, but how many of those are on Amazon USA only? 3k? Less? And what is the typical open rate of a BookBub email? How many of those 3k (or less) readers had opened their email in the short timeframe you were testing in? That number could easily be less than 1000. Possibly even significantly less than 1000.

        You need to be targeting authors with more Readers for another reason: you’ve no “up” to scale up to even if the tests work out.

      3. So first, let me say that I appreciate all your input on this. I don’t expect you to have all the answers!

        That said, I’m confused on the numbers. Bookbub targets Readers, right? Not just Followers? Even if we just go with the last author, I would think there are at least 25,000 Amazon USA readers in the list. What’s the mechanism by which saturation occur? Does Bookbub keep sending it to the same 1000 people? I guess we could assume followers are more likely to open, BUT I’m not sure that theory holds up. Followers have actively chosen to follow authors but may not do a lot of Bookbub buying beyond those authors. I fit into this category myself. Readers on the other hand means they have actually clicked on a book by the author somewhere in the BB sandbox – I am pretty sure this is the definition of Reader because I have 9 followers myself but 2000 readers. I have had exactly one BB ad and I had 900 sales and likely some borrows and 2000 clicks feels about right for that.

        If we assume that saturation did occur, I’m trying to figure out where to go next. I have tested about 20 authors and all the ones with more followers/readers were under 2%. I have a huge list of possible comp authors, but I picked the most likely 20. Do I now go to the next 10 most likely? They’re all pure action thriller authors but are slightly less direct correlation to terrorism thrillers – maybe more conspiracy or military based thrillers.

        Or do I try another graphic? The problem with another graphic is that is a rabbit hole. Between different backgrounds, alternating between yellow and red text highlight, and different short hooks, I could spend $500 more without even blinking given that I also have to be trying new authors now. I was very relieved when I got a graphic that seemed solid on try #4. I get that this is the most time consuming and frustrating step, I just need to make sure I am doing it as effectively as I can.

        BTW, Here’s the ad that got me those CTRs:

      4. Let me answer the factual questions quickly and I’ll get back to you later on the images when I’ve more time.

        Frequency cap: BookBub shows a campaign to a Reader up to four times. This has been the frequency cap in place for quite some time and they have it pitched just right IMO. Keep in mind that many users won’t scroll all the way down to the bottom of their email and physically see your ad, so a higher frequency works with BookBub. (Whereas with something like FB I’d be snipping an ad once it shows twice to a user.)

        Note that this frequency cap is *per campaign* so if you copy a campaign the clock starts again. Or if you run a test ad to Author X, and then run a full on campaign the next day, the clock starts again.

        As for Readers v Followers, “Readers” = Followers + every user that has interacted with your books on the platform, i.e. those who have clicked on Ads or Deals for that author previously. And I would agree that means it is quite likely that the non-Follower segment of the Readers is probably more engaged and a better target for ads in many cases – except perhaps where an author has done a lot of Free featured deals etc., in which case I’d flip that around.

      5. Thanks, David. That’s in line with how I thought it worked. So the question remains how could I saturate all those tens of thousands of readers (over 10K, though there could be a bit of overlap and the actual number might be as low as 80K) with under 6000 impressions?

        One more data point – I had started Book 2 on countdown deal two days ago to continue to do testing. I tried the best CTR user from my list (249 followers, 14495 readers, initial CTR 4.20) with the same ad except the cover for book 2. I got 3.5% CTR on 1000 impressions again yesterday. It was spent by mid-afternoon. I added another $10 today to the ad and the CTR dropped substantially, it’s now 2.5% on the whole ad. It may go up as the day progresses and clicks come in, but it’s still going to be worse by a bit.

        Between what I mentioned above and this latest, it’s VERY consistent with Bookbub continuing to show it to the same users and that not working well.

  33. I see a non-fiction indie author that writes about mindset and thinking launch books to the top 1000 of Amazon on a near monthly basis. I know BookBub ads are a core part of his strategy, but it wasn’t until reading your book did I realize how he’s doing it.

    I figure he’s spending $200-$300 per day, but I do wonder who he’s targeting.

    Anyways, thanks for writing this book… lots of great tips.

  34. A lovely bonus: in the new “More items to consider” (replacing “Also Boughts” ??) I’m seeing recommended authors against which I tested! Nice quick, way to clean up your Also Boughts. Thanks, David!

  35. I’ve just begun the testing phase – thank you for really clear guidance and a generally great book!

    My first ad (as you predicted) returned a poor CTR – 0.88%, for a 99¢ book (1352 impressions, 11 clicks). I’m testing other authors for now, and I know you said in the book that any authors bringing in a CTR below 1% in the test phase should be dropped.

    BUT. This author is ALL over my book 2 also boughts on Amazon. She has 4 titles on my first AB page – it’s basically my other two books, four of her books, and then one more author on page 1. She has 22 titles overall just in my book 2 also boughts – she properly dominates my also boughts (and she’s prolific so she has a ton of books). She has also recommended my books to her (extensive) mailing list without me asking for a swap and left a review on my book’s bookbub page saying people who like her stuff would probably like my series.

    So I’m reluctant to let her go as a target as all signs point to her being a very good fit for me. She also has over 12k followers on BB.
    Is BB so different from Amazon that all this is irrelevant? Should I stick hard and fast to the 1% and above CTR rule, or is it worth trying to target her again once I’ve had good results with an ad graphic?


    1. I would say that, yes, BB is so different to Amazon (or FB) that you have to learn to ditch targets that otherwise look nailed-on elsewhere. Maybe stick this one in the Maybe pile though. And when your testing is nearly done and your images are super slick perhaps you can re-test this author and compare results again. Be prepared to kick this author to the kerb though!

      Also keep in mind that – aside from the above – Also Boughts can sometimes lead us astray because people are targeting *us* on Amazon Ads who mightn’t be the best comp authors. People can bully their way to a good position in Also Boughts by spending heavily.

      This is a factor which is more pronounced when you are more successful/visible. The very biggest authors often have Also Boughts from authors which are nominally in the same genre but have totally different audiences as a result.

      1. Well I’ve seen firsthand how right you are about BB being so different from Amazon: I’ve done 8 test ads, mostly on single authors, a couple with combined authors when the reader numbers have been smaller, and they’ve all returned proper stinky CTRs – that 0.88% looks pretty damn good by comparison. All on authors taken from my also boughts for books 2 or 3, and three of those are also on my author page also boughts, as you suggested in the book.

        I have to admit I paused 3 of those ads prior to reaching 1000 impressions, but with 1 click for about 600 impressions for each one, I can’t see how they would recover to suddenly deliver more than 1% CTR. I preferred to stop the bleeding and get back to the drawing board.

        But I’m stumped as to how I go forward from here. If I can’t use data from Amazon, how else can I identify potential target authors on BB? Test ads that fail badly work out really expensive, so I don’t want to just throw more money out of the window without some kind of plan.

        Have you got any suggestions as to how to identify potential author comps going forward? I’ve thought of clicking the genre ‘tag’ on my bookbub profile and trying the first couple of authors that pop up there, but that seems veeeery scattergun, and again potentially a very expensive way to get to know my BB audience.

  36. Hi David,
    I’m half-way through your book and reading as fast as I can. Basically I suppose what I need is assurance (or the opposite) that my plan is a good one.

    You see, I have an Editor’s Pick Free promo on Kobo for April 22-28, and I’m thinking it may be a good idea to boost the sale with a BB ad. Thing is, I haven’t tested yet. I have what I think is a good image, with my cover (FFS) both complete and partial as the background image. The text in my image is straight off what BB themselves wrote when I had my Featured Deal, and I have FREE in a nice bright box as well. I would only target Kobo readers, as I don’t plan to drop the price on Zon myself. I realize if they get wind of the sale, they might drop my price anyway.

    So, my questions are 1) am I wasting my money if I don’t test on Kobo for a free book ad? 2) is it even worth testing a Kobo ad, considering your advice in the book to test on Kindle US because Kobo is easy anyway? 3) I’m still not sure what to do about the follower vs. reader thing–I, myself, have 82 followers and over 25K readers.

    In other news, I’ve been targeting myself in slow-drip ads for my second book with positive ROI. Good news is I don’t have to bid much, because no one else is likely targeting me. 😀

    I’m probably forgetting something. Looking forward to your answer, though. 🙂

    1. I definitely caution in the book that BookBub Ads aren’t usually the cheapest place to get clicks on a freebie. I often still use them to some extent to back a free promo, but I’m usually looking to get ROI via sellthrough or page reads.

      I’m not clear if you have done testing of any kind already. If you haven’t this will be quite a gamble. If you have, and got good results on Amazon US, then you can expect results at least as good at Kobo. Also not sure who you are planning to target with this promotion. Yourself? Comp authors? Have the latter been tested?

      Normally, I would say that you can get far cheaper clicks to back a free promo at places like ENT and Robin Reads etc. However, this doesn’t apply in your case as you are solely looking to boost a Kobo promo. You can try it, just remember that click costs mightn’t be the cheapest, and that free downloaders are often hoarders who don’t buy Book 2 right away. A price drop on Book 2 can help with that, but that doesn’t sound like it would fit with your plan.

      I’d be tempted to just see what the Kobo promo can do on its own, and I’d only dabble a little in BB ads, if at all.

      BB ads, as with any book advertising, will only generally return a decent profit if they are part of a larger, overall plan. That’s the way of the world in 2019.

      1. My plan with this would be to target untested comp authors, unless I test as I go, running simultaneous tests on each author, leaving the ads with the higher CTR running and stopping the non-performing ones. As with my free Featured Deal last July, I’m counting on read-through for ROI. My focus is getting my ranking up on Kobo to prolong the exposure. I only have two books out in the series so far (the promo’s on Book 1), but I also plan to put Book 3 up for preorder this week, while the promo is running.

        The thing I don’t understand, and maybe I just haven’t arrived there in the book, is how is it possible to test ads when your book is full-price? I don’t feel that testing would be accurate when you’re not able to state the price in the ad, as you suggest to do. Is there something I’m missing?

        Thanks so much for your reply. Hope you’re having a lovely holiday!

  37. Hi David,

    Thanks so much for all the great advice. I had tried some BookBub ads last year and got very little return for the effort. After reading your book, I’m already leagues ahead with this attempt. So here is my question–which you might have answered already, but with a search through your book and this page, I didn’t see it–I’m a wide author. Two of my best comp authors (got 3-4% in testing) are KU authors. I’m wondering if targeting them, though they give me a good CTR, is a a good return on my investment in terms of sell through. Thoughts/opinions/advice? Thank you in advance!

    1. I would target them without question. Just because someone is a KU author, it doesn’t mean all their readers only borrow and don’t purchase. Many of their readers will be Amazon customers who aren’t KU subscribers. If you are worried about this, you can benchmark conversion. Run a mini campaign to those two authors and see how many sales it generates.

      1. Hmm. I’m wondering if you’ve run into this–I’m trying to use a drip campaign to advertise the first in series free book. My test ad gave me a 6.09% CTR. I copied the ad, chose to run continuously, then picked to spend a certain amount per day rather than as quickly as possible. My CTR dropped by half. Is this simply the difference between a fast and slow campaign, and will it even out?

  38. David, could you shed some light on how one chooses comp authors. I am about to release a rebranded series in the fall of the year and I have had great difficulty with exactly that in my FB explorations to date. Love the books.

    1. There’s a whole section in the book on this (Chapter 8, I’m pretty sure). In short, Also Boughts are usually fertile ground for comp authors. Author Also Boughts (on your Author Page) are great too. And then the reader side of BookBub is helpful also.

    2. To add to what David said, it’s my belief that it is critical to have a deep understanding of your genre. Even if you wind up paying someone to do advertising, understanding the tropes of your genre is important when you are writing. To do this, you need to read books in your genre. Unless you’re in a genre with very little KU presence, get a KU membership for a month and sample 15-20 top authors in your genre. You don’t have to read the whole books, but read enough to get a feel for what they do. If you have a very narrow genre, you may need to expend beyond that narrow area to get 15-20 authors, but this can be valuable just in seeing where differences lie as you start expanding outwards.

      To be honest, you should really have a deep understanding of genre before you even start writing. Not necessarily before you have an idea for the book, but before you really spend significant time on it. I’m not saying don’t write some genre-bending mash-up, but at least do so with your eyes wide open that there is no ready made target market.

  39. Hi Steve, I have a question about also boughts. I followed your advice as close to the letter as possible given late nights, a brain on the brink of addleship, and just trying to figure all this out. My first two books came out with great also boughts. Not so much with the most recent one, and that’s the one where I tried to follow your advice. I’ve thought long and hard about what was different (and btw, it would have been far worse but for following your advice. The main difference was that I did a lot of word searches and posted them to ads on Amazon Ads. I did that before discovering your books and blog. I’m guessing that people searched using words that some search function I used found to be connected in some way to my new book, bought on that basis, and so now I have scifi readers, romance readers, etc. causing my list to be populated with books that are a far fling from what I write. I did not encourage family and friends on FB to buy, etc. The explanation I gave is the only one I’ve thought of that makes sense. Does that make sense to you? Lee

  40. Just a heads up for everyone, there seems to be a bug at BookBub which is affecting new ads and is garbling the links. Not sure if everyone is affected. It seemed to hit some accounts last night, but not mine. And then today myself and some friends seem to have it also. It looks like this, and there’s no workaround I know of yet, so you’ll have to hold off on creating new ads for the moment – if you are affected – until they get it fixed (presumably that will happen later today):

    1. Update: this seems to be fixed now – at least it is for me. If you saw a big drop in serving today, this bug is probably why. My Featured Deals email didn’t even have any book ads in it. They probably took the system offline for a bit to fix the bug.

  41. Not quite. A book priced at 99¢ in the US sells at CDN$1.32 in Canada and AU$1.39 in Australia. An ad that shows 99¢ in the US would not be localized if served to Canada and Australia. I guess I could put US$0.99 in the ad. Does that seem too cumbersome? Thanks, Lee

    1. Ah, I get you now. I think customers are used to paying VAT/sales tax on top of quoted price – at least I’ve never had any comments or complaints when running big “99c” sales in CA/AU via BB and FB. And some of those FB posts would attract a lot of comments asking about paperbacks and audio editions and other series books etc. Never one about the price in the ad being inaccurate.

      Not quite the same, but as someone in Ireland, we regularly see ads for UK products and pretty much expect there will be a little extra on top when we go to buy it, whether that’s from our higher rate of sales tax, higher business costs in Ireland, or currency differences/fluctuations. I could see CA/AU being similar.

  42. Another question related to Brenda Hiatt’s above: you mentioned that a good tactic was to do the US and UK market separately and lump Canada and Australia together. I understand the reasons and agree. With regard to Canada and Australia: with the conversion rates, 99¢ is not what’s reflected in the pricing for either country. How do you handle that visually?

  43. If I didn’t say so earlier, add my praise for a great, easy-to-follow book!

    So. Now that I’ve done a fair bit of testing and have more or less locked down my image and 7 (so far) good targets, I’m trying a drip campaign, as this is for a permafree first-in-series book. Do you recommend doing separate ads to separate vendors/countries? Or all together? I started with everything EXCEPT Amazon US (figuring that would suck up all my clicks). Then I started turning off the ones getting me less than a 2% CTR or higher than $.60 CPC. Wondering if that’s the way you’d do this at this point?


    1. Thanks Brenda! There’s a couple of factors, depending. And this is just how I do it as it suits me, but there are options.

      *If you’re doing a free promo, you don’t have to worry about prices or currency symbols being different anywhere so you can run everything together in theory, but you do have to worry about bigger markets gobbling up all the serving. As such, I often separate out the US, and then either run UK/CA/AU together, or split out the UK also, depending on how much I want to ensure money gets spent in CA/AU (i.e. if I’m more aggressively building up in those territories). I usually run CA/AU together – at minimum – as the ads can struggle to serve otherwise. And then, if it’s a wide promo, I usually run all retailers in the same ads for each territory for the same reason. But I guess if building up at Apple US or somewhere was an immediate goal, you could separate out that market to ensure a certain amount gets spent, but that’s just a little fiddle for me personally and will probably only work with bigger targets anyway.

      *If you’re doing a 99c promo, often the currency symbol is the only difference – and I strongly recommend separating out the UK at minimum. CTR will be affected if you don’t localize the price tag. However, I often separate out UK and then CA/AU together for the same reasons as above – to ensure a certain level of serving there.

      *If it’s a higher priced promo, often currency differences mean you have to separate them out anyway, as a $2.99 price in the US might be $3.99 in CA/AU.

      Normally I have three campaigns set up to target USA (all retailers), UK (all retailers), and then CA/AU together (all retailers), but I can fiddle with the set up sometimes.

      And, yes, I’ll micromanage each retailer and switch them off when they dip below 2%. Sometimes I let them dip a little first – the lower bids needed to serve outside Amazon/USA can mean the clicks are still coming in sub 50c for a bit.

  44. Hi David, I agree with HN above, what a great book. Despite the mystery of this morning (which continues), I see a way ahead.

    I took out Legendary–and things got worse! I ran the combined group in the UK and had results similar to what I’d had on the individual runs–ended with 2.74% CTR. It went down. It also got worse when I took him out in the US. Meanwhile, I tested it with another candidate for the group that I had not previously used. I used the same graphic, etc. He came in at 3.10% CTR.

    I think you might have put your finger on the issue when you mentioned the new “Readers” data point now available to us. I used your suggestion of including those with a low following who have a large readership. I’m thinking that, as you said, some of those might have a large segment of free-readers. I’m thinking that the closer the follower number is to the reader number, the more reliable the figures. I’m going to test Legendary again and look for another culprit. Fortunately, I’m in KU & Lending Library, my conversion rate is pretty good, and I think I’ll recoup through re-through.

    I can’t say enough good things about the quality of your advice. There are a lot of purveyors of advice. I’m learning where to pay attention and where to be skeptical (I was raised in a British school in Morocco, but can’t remember the British spelling for that word–sceptical?

    Anyway, my hat’s off to you.

    1. Lee –

      That’s a really interesting suggestion on having reader numbers and follower numbers close together. I’m still in early stages and I think the first background I used (an image from my cover) probably wasn’t good as I tried different text to no avail. So I’ve had CTRs really poor. The latest ad with a new background seems to be doing better – it’s a guy with 400 followers and 8000 readers so it isn’t hitting impressions quickly (waiting for this morning’s emails to go out and maybe get me to 1000). But I’m going to try looking for others with Reader and Follower numbers closer together than 1:20.

    2. Lee – any luck with additional experiments around the combined author ad? I’m tempted to run separate ads per authors. (But I see BB now shows you splits per author for a combined ad.) Thanks! HN

      1. This was one reason I was holding off!

        As people are now starting to see in their accounts – from reports I’m getting – BookBub is starting to roll out a very cool new change where you will get a CTR breakdown by author.

        This will be very handy in diagnosing the issue in weird cases like this where a group of road-tested comps thrown together displays a weirdly low result. This should indicate the problematic author that needs to be re-tested or perhaps excised.

        It will also make testing a little easier too. I’ll be discussing the changes in more detail once it rolls out to everyone – presumably quite soon, it appears this change is being phased out in batches.

        I’ve been testing it for a week or two so should be able to give some tips.

  45. David,

    First: what a great book! Thank you! So witty, so easy to follow, so optimistic! Just what us Indies need. (Worthy 5-star has been posted.)

    Background from this weekend:

    1) Tested per your methodology against roughly 20 comp authors with an ad for a free novella feeder. All hit over 1000 impressions quickly. Results: Four above 1.59% CTR, four btwn 1.50% and 1.29%, 3 ‘small guys’ with few followers I combined and together came in at 1.71% (!!), and the rest individually came in lower than 1.19%. (And yes, the big names came in very low – you are spot on about not wasting $ there.)

    2) Combined the top four and hit them up with two improved ads – for split testing. Ad A = 2.10% and Ad B (with fewer words) = 2.31% Clearly fewer words better.

    3) Ran same split test against the combined small guys. Ad A = 1.71% and Ad B (fewer words) = 3.57% !!! (ha ha, those followers of the small guys sure are selective.)

    4) Turned on drip campaign for free series funnel to top four + combined small guys. Unfortunately, with over 2K impressions now getting 1.62%. Ugh.


    1) Perhaps this pool is oversaturated now?
    2) Turned on drip ad Sunday night – maybe it’s a timing issue?
    3) Back to drawing board – but on which part?


  46. Thanks, David. I tooled around a bit and discovered the downloadable CSV file that shows the specifics for each traunch of that ad. It shows that the first time I ran it, it generated 950 impressions and 15 clicks. The second one generated 757 impressions and 3 clicks.

    FYI, I have the exact same ad with the modification for “pence” running to a UK audience now (including Legendary). Preliminary results are promising: 178 impressions, 12 clicks, 6.74% CTR and 4 sales from the UK have already shown up since I started it. We’ll see how it holds up the rest of the way.

  47. Sorry, I neglected to answer your question about numbers of impressions. I re-ran the test this morning by upping the budget for the combined ad by $10 (the same amount I had run it yesterday). Unfortunately, I did not keep the numbers for the first iteration, but the 2d was even more dismal: it came out to 3 clicks and 1.05% CTR across 1717 impressions. Ugh. Should I run it minus the legendary author? Neither his following nor readers fell outside your parameters, but maybe he’s sucking them all up.

    1. That would definitely be the next thing I’d test. Just drop $10-$15 on a test ad featuring the author set, minus this author. If there is one author with more Readers than the rest (whether it’s this legendary author or another in your set), they could be the culprit here. This kind of result makes me think that one author dominated the serving for whatever reason and is making the whole set look bad.

  48. Hi David. Thanks for your response. I tested 12 authors and ended up with 6 that met your criteria in terms of followers and readers. I found, as you predicted, that the ones with larger followings and readers did not do as well. They ranged from a low of 58 followers with 5,069 followers to a high of 6,116 followers with 41,560 readers. I tested them with the same book, same graphic, nothing changed between them. That’s the way I arrived at the six that went into the combined pot. I used the same book and graphic for that. The offer was for a 99 cent new book on launch. Only when I combined the authors into a single pool and ran a test ad in the US (all of them had been run only in the US), did I see those dismal results.

  49. Hi Lee. First of all, those are some promising test results. Let me try and understand the problem fully. So the good results above were from testing, but the 1.58% was from when you lumped them all together in a group and targeted them as one? How many impressions led to that CTR? How many followers do each author have? Are the graphics the same as used in the test and as used when targeting the whole author set? Was the book the same? The offer? Trying to tease out what could have caused the variance here.

    Without more info, I can only guess, but what I suspect here is that there is one author – one with a larger number of Readers/Followers – which is dragging down the performance and gobbling up all the serving (and you might have just got a fluke when testing that individually, which is unfortunate but can happen).

  50. OK, David, now I’m dismayed, frustrated, ready to kick something. Having gone through several days of testing, I came up with 8 authors with the following CTRs: 2.43%; 2.63%; 2.36%; 5.48%; 2.36%; 4.04%. One of them was a legend I’d read since escaping college and I was sure he’d get kicked out of the group, but he came back with healthy numbers and I was oh so humbled. I had done a wholesale change of background before the round of testing that yielded those authors, and I was pleased with the result. So I ran a test using those same authors, and the CTR came back 1.58%. What do i do?

  51. David, what is the effect of time of week that tests and ads are run. Are some days typically better than other days? Are there days that are just not good (on the weekend, maybe?). If the latter case, I guess it makes sense to start a 5-day launch so that it ends on a weekday? Seems like that would indicated starting on Sunday since Friday often starts the weekend. How does any of that play in? Thanks.

    1. I don’t think it matters so much. I tend to find CPM costs drop at the weekend, as less people (especially publishers, I’m guessing) are running ads, but that’s about it. Time of day might be a bigger influencer in terms of serving peaking in the mornings and CPM prices sometimes dropping towards the end of the day when others are out of budget.

      So let other factors determine what days you run your campaign – BookBub doesn’t matter so much on that front, in my experience.

  52. I’m getting ready to pull my hair out! After a LOT of tests, I located about 5 potential comp authors who returned a CTR greater than 1% (one as high as 2.25%), but I keep falling on my face as far as finding an ad graphic that can boost those numbers. Rather than blow my brains out in a potentially never ending quest, I elected to try running a split test on my newsletter readers. I split them into six groups of around 300 and asked the first group to click on the one that rang a bell. The winner then went out to the next group along with its challenger…and so on, until I had a final winner. I must admit, it was not one I would have guessed as one to garner all the votes, but true to David’s prescient advice, using a background selection from my book cover and the 3D book graphic were the keys that my test group liked (along with some other tweaks about price placement and text). Now, sure that I had found something that would improve my BB ad stats, I ran a quick test on one of my “almost ” comps (with an initial CTR of 0.98%). While the end result did not get above 1.8% for him, I was pleased that the CTR had almost doubled. Emboldened, I put my top 5 comp authors into a “Gold” group and ran the ad to them this morning. The ad performed dismally. I completely out of ammo and am at a loss of how to next proceed. Help, please?

    1. What CTRs were you getting for these authors separately? And what were you getting as a group? What Follower/Reader count do these authors have? Often this scenario is caused by one author with a large amount of Followers/Readers dragging down the performance of the whole ad as they eat up most of the serving, but I’ll need a little more info here.

      1. Thanks. The follower, reader & individual numbers CTR% are:(1325, 45449, 1.15%), (847, 11184, 1.36%), (487,11115,2.24%), (420, 14785, 1.71%),(410, 12992, 1.22%). Collectively they gave a CTR% of 1.69% with my best ad, but the one my newsletter readers picked through the split testing was only 1.09%. Obviously my newsletter readers are steering me the wrong way.

  53. I meant “shown on a tablet screen” not “table screen” in my last entry. Update: The third test came in worse than the second one, so now I’ve taken the first test (with the 2.3% CTR) and run it against another author. I’m at a loss on how to improve it. Have you found that putting in a 3D book instead of a 3D tablet makes any difference? The next thing to try, I guess, is a complete change of background, but that risks coming in with a worse result than the original 2.33%.

  54. Hi David, I ran a test ad this afternoon. It came in at 2.33% CTR. I ran it again with no changes except that I added a line that said, “Free with KindleUnlimted.” That dropped off a full point of CTR to 1.38%. I’m running it one more time with the background color and font changed to yellow and black respectively, but I suspect that the downward jolt resulted from turning off those people that don’t use kindle or are not part of Prime. That’s probably a good thing since I won’t have non-kindle users clicking in, but I’m left at the drawing board to get the CTR back up. The background is off the book cover. Depending on what comes back on this latest trial, I guess I could try changing up the background color a bit. I had already done that to provide contrast to the book cover (shown on a table screen). If you have any suggestions, I’m all ears. Lee

    1. I’ve answered your comments further down, but just FYI, if you are targeting Amazon USA with your ad, that will only get shown to Kindle owners. You never have to worry about that side of things. (Of course, that doesn’t necessarily mean they are Kindle Unlimited subscribers, and I personally haven’t found that line to be a real draw.)

  55. Hi David

    A couple of questions, if I may:

    1.Is it possible to use an affiliate link as the target URL? Presumably this gets into shaky ground as far as Amazon TOS, I’m just curious if you’ve tried it or know people who have.

    2. Bookbub doesn’t tell you anything about how many readers or followers are in the US vs other countries, correct? Beyond that, you mention that targeting outside Amazon US sometimes gets high CTR. If so, does that mean that Bookbub looks at your link and determines where it is (say Amazon US or Amazon UK) and targets subscribers in the correct countries? It must, right, or else you’d get a truly abysmal conversion on anything except US linked ads.

    1. Sorry, I thought I responded to this already.

      1. You can theoretically use an affiliate link. It’s not against BookBub’s TOS and nothing is stopping you on that front. However, there is the issue of Amazon’s TOS. It’s not clear to me if it’s permitted or not. Amazon recently brought in much requested clarity with regard to the use of affiliate links in email (most definitely not permitted). They had the opportunity to do so for ads, but chose not to for reasons only known to them. From a moral standpoint, I don’t think there’s an issue at all as you are most certainly driving traffic to Amazon and should get the affiliate income. And it’s not like Amazon is advertising our books on BookBub and we could be accused of driving up their ad costs (the original reason this rule was brought in many years ago – I was working at Google AdWords at the time and remember it). But it’s an open question about the TOS for the affiliate program. Of course, what we really want/need is the data, so hopefully some solution will present itself. Although affiliate data gets more and more incomplete the more mobile traffic takes over too.

      2. BookBub does not give that breakdown. I can only say that about 75% of my own Followers are US-based, but can’t say if that’s typical. And yes, BB’s system reads your link and sees where you are pointing to: Amazon US, Kobo CA, etc. And then it will direct the reader to the store/territory that they indicated in their prefences when signing up. That side all works pretty seamlessly.

      1. Thanks, David. The reason for my question on #2 was also thinking about directing somewhere other than directly to a book-seller. I have no plans to do so myself because I don’t want the extra friction BUT I have on occasion seen a landing page that was so good that it almost certainly resulted in better conversion than going straight to Amazon. But in a case like that, I don’t know how Bookbub would know which reader locations to serve an ad to.

      2. I used Amazon affiliate links for months until Amazon decided to terminate my affiliate account a couple of weeks ago. I’m not sure why exactly but I suppose it’s because I used Amazon affiliate links for my Bookbub ads.
        I did it mainly to track conversion but the extra money was nice too.

        Because I violated their TOS they told me they will not pay me my affiliate commissions (close to $500). It’s how they thank me for bringing them customers!

        So, if you choose to use Amazon affiliate links, do it at your own risk.

  56. If it were possible for me to dog-ear an ebook, yours would be ratty around the edges by now. Having absorbed all that lovely knowledge, I’ve embarked on a testing mission. So far, so good, with a few missing details.

    1) When you test ads, do you try a different one each day? Do you ignore the prescheduling function and manually stagger them throughout the day as results come in? Or do you test them simultaneously? And if so, how many do you test simultaneously?

    2) You suggest starting up ads no later than 9AM EST. But if you don’t do it manually and use the prescheduling function, is that even in your hands? If you preschedule an ad for a specific day, does it start serving at 12:01AM and possibly peter out by 3AM? Or are you only suggesting it’s not worthwhile to schedule an ad after 9AM on the same day?

    Thank you for your experimentation and your generosity in sharing the results! (Also, I’m enjoying “Liberty Boy.”)

    1. Glynnis: this is where you brutally expose both my laziness and my general lack of planning.

      I’ve never actually used the scheduler.

      Maybe it’s because I have the luxury of a few extra hours before that US email goes out (I’m in Dublin, on GMT), but I’ve never really felt the need to use it – so I don’t actually know what time those ads start serving! It would be good to know though, so I’ll make a note to ask BookBub.

      I have used it to turn campaigns off, and I’m pretty sure they clock off at midnight, so it would seem to make sense that the scheduler would start campaigns at just after midnight. Which may not actually be ideal for you – so it would be good to establish that more exactly. I’m such a micro-manager though that I tend to swoop in an knock things off manually anyway though.

      As for testing, again, I do those manually.

      The danger with testing is overlap and audience fatigue. If you do too much testing too quickly, and there is a lot of overlap between your target authors – which is quite likely – you could overserve to lots of your audience which will have a dampening effect on CTR.

      So, ideally, a nice, slow testing process is optimal. However, that’s not always practical or desirable.

      I’ve done extremely fast testing with one author who had various reasons why he needed results ASAP. That will probably have a dampening effect on CTR overall (just for the testing phase – that shouldn’t be an issue when running bigger campaigns where you lump all your winning authors together). So factor that in to results when judging the effectiveness of your targets etc.

      If I had to give a general rule of thumb, maybe four or so in a day is a nice pace. But you can go faster if you need to, of course.

      1. Hi David,
        Maybe this is premature, since I haven’t done my first test ad yet, but I may as well ask it. When tweaking an ad, are you saying to send it using the same author(s) you used in the first one? If so, wouldn’t the results be worse since you already targeted that author? I think you said early in the book that you shouldn’t target an author more than once every six months or so. Am I confused? Most likely I am.

      2. Yes, the only way you can do a true split test is by only changing one element in the ad. So if you are testing a new ad image you should, ideally, target the same author so you can compare the response.

        Ad fatigue like you describe is a thing, for sure, but the level of ad fatigue will totally depend on how many Followers (Readers, really) any given author has. One of my targets only has a couple of thousands Readers. After two test ads, or a couple of thousand impressions in a regular campaign, that audience can be tapped out. It’s very different for an author with 10,000 Followers (or 50,000 Readers). That’s one of many reasons why I recommend not going to low on the Follower (or Reader) count. You burn through them too fast, whether that’s testing or regular campaigns.

        All that said as I said in (I think) the final chapter, ad fatigue happens more slowly on BookBub than other venues, in my experience. Part of the reason for this is down to the ad placement – not all readers will scroll down and actually see the ad, but they will open the email, and that will count as an impression. Another factor here is the method of serving: not all readers check their email every day or open their BookBub email every day. So you can hit a lot of fresh eyeballs on Day 2 of a campaign, for example. (BookBub will serve a campaign to a given reader up to four times – and I think they have that pegged about right.)

      3. Hmmm…more dilemmas. I’m on PST in Los Angeles, so bring me coffee when I’m up in the wee hours, testing ads!

        I think while the slow testing process works great for perma-frees, it’s difficult with books that are on sale for a short time. I just put a book on sale for 99¢ for a one-week period, and I’ve spent four days testing (one test per day, changing the ad but not the three target authors, which now I know may be contributing to audience fatigue–sorry, audience). My results are still less than ideal, but by the time I figure it all out, my sale will be over. Next time I’ll try your several tests per day model. And I guess the target author should be changed with each test to avoid overlap.

        Lots to learn, but thank you for doing the preliminary testing!

      4. I can’t say more at the moment, but there is something coming sooner rather than later which will help A LOT with this part of the process. I’ll be testing it so I will be able to share some initial best practices as soon as the change actually comes onstream (can’t say when that is but I’d imagine it’s sooner rather than later as I said).

        Which doesn’t help you right now (although I think you’re moving in the right direction by switching to a few tests a day) but you can at least take comfort in the fact that once you get through this squall, it should be plain sailing after that. The testing phase is the most frustrating part – by a country mile.

      5. I actually was thinking the same thing (I’m just starting on my image, haven’t started tests yet). I have concluded that I will use my countdown deal on Book one to start. After that I have several options:
        -do countdown deals on book 2 and 3 and use those for further tests. I don’t love this for multiple reasons
        -do a countdown deal on my 3 book boxed set. Better than doing book 2 and 3 but still a different graphic and a better deal so maybe better CTR
        -just lower the price to 99 cents on Book 1 for the remainder of my testing. It will mean the net cost to me is a bit more since I only make 33 cents instead of 60, but overall I think the cost of doing this is not huge for the consistency it buys me. I think it will mean that I have to wait 30 days after putting the price back up before I could run “real” ads with the best authors from testing, which is not great.

    2. P.S. Glad you are enjoying Liberty Boy! That’s my favorite book. Looking forward to getting back to that series later this year, finally.

  57. Hi David,

    We’re struggling with trying to make money on two author businesses at home, but we really like your books and track record.
    Howevs, we’re now bogged down in the craziness of ad testing and while we believe you completely about CPM we’re *soooooo* skint, before your magic map guides us to the fabled fire hydrant of cash(TM), can we spend less than $10 for a CPM campaign?
    So terrified, but on the basis that we’ve done pretty much all the free stuff we can do, the BookBub Ads seem the cheapest for reasonable results. If we can get it right of course.

    1. Hey Paul. The $10-$15 figure isn’t arbitrary. It’s the range that will give you around 1000 impressions – which is generally considered in online marketing to be a sufficient sample size to make a judgment on an ad. Anything less and you are in danger of calling it early – one way or another.

      In theory, you could just drop $5 on each test if budget is really tight, but your results won’t be as solid. I prefer to get the foundation 100% right so I can rely on it in the future without looking over my shoulder or having to re-test anything.

      However, if budgets are that tight, BookBub Ads wouldn’t be the first promo tool I’d reach for. This is what I’d recommend spending on first:

      1. Optimize your metadata – especially keyword and categories. Read Amazon Decoded (it’s free) and apply the advice in that. It won’t cost a penny. This will also increase the return from any future promos when you do have the budget for advertising. More here:

      2. Promo sites. These don’t scale very well for the most part, but they are some of the cheapest clicks around. Some sites are even free, but many are in the $20-$50 range. If my budget was very tight, I’d spend on this before AMS/FB/BB ads. More here:

      3. Newsletters/content marketing. Any advertising will cost you, both in terms of the spend for the campaigns themselves, and for the testing also. If budgets are going to be tight for the forseeable future, I’d look at alternative paths to readers which can be time intensive but low or no cost, like building up your mailing list or building up your Facebook Page via content marketing. More here:

    2. I appear to be quite confused today! Someone must have slipped me decaf…

      Were you referring to the amount I recommend you spend on a test ad (i.e. $10-15)? If so, my previous answer applies. If you meant the BID that you put in for a CPM Bidding campaign, then read this instead:

      The danger with bidding too low (and below $10 is often low, depending on your author targets), is that your ads won’t serve at all. You can try it, and you might get away with it if you aren’t seeking to scale that ad too much. But if you are looking to do a big, blast campaign where you are scaling that ad up to 10,000 or 30,000 impressions, you will almost certainly need to bid higher.

      At the other end of the scale, if you are pushing a permafree book with a low daily budget (like $5 or $10) then you could easily get away with a bid of significantly less than $10 CPM. So… it depends.

      All that said, if your aim is to reduce costs overall, the best level for doing that is increasing your CTR, rather than skimping on CPM. If you look at the detailed optimization examples above – with various screenshots I couldn’t include in the book – then you will see that what will radically reduce your CPC, and your overall costs, is getting that CTR higher and higher.

      It can truly make a radical difference and deliver far, far cheaper clicks than lowballing CPM (or using CPC).

  58. Oooh! A comments section! (I’ve kept checking back hoping for this.) Having read the book AND this page (several times), I’m trying to figure out where to go with my testing. A $10 budget never seems to get me 1000 impressions, even bidding above BookBub’s suggested max (which seems to be over $12 no matter how big or small the author, which seems odd). Bidding $15 eats a lot of cash really quickly while I simultaneously try to cross-check both author targets (one at a time, as suggested) and images (ditto). Even my best results tend to have an average CPM of over $10. Can that many people really be bidding that high for some of these smaller (under 5K followers) authors? I feel like I’m missing something fundamental and wasting money in the meantime.

    Oh, I’m testing with a permafree first-in-series, if that makes a difference?

    Thanks for ANY guidance here!

    1. Hi Brenda, sorry about the delay in rolling this out. I had some unexpected issues opening up this comment section. I think two things might be getting confused here. Perhaps I’m just confused. Or maybe we’re both confused! Also possible.

      BUDGET – this is your campaign budget. The total amount you are spending on this ad. I recommend putting in $10 or $15 as that’s usually sufficient to get what you really want: 1000 impressions on your test ad.

      BID – this is what you are bidding for those impressions under CPM Bidding. The recommended amount will vary depending on your targets and what you are doing with your ad. For test ads, I recommend bidding high because you want quick, clean results. You can be more conservative when running proper campaigns (much more conservative for something like a permafree).

      In this particular case, I recommend putting in a BUDGET of $15 for your first couple of test ads. And putting in a BID of maybe $12.50 or $13. Which sounds like a lot but that budget will limit your exposure.

      What will drive your click costs down ultimately is a good CTR (like in the optimization examples above), so don’t sweat the BID. Especially not when testing. The BUDGET will act as a solid cap on your spend.

      The permafree shouldn’t make a difference really – just factor in that a free book of any kind should result in higher CTRs. (But when you are done with testing and running a permafree campaign you will want to reduce those bids considerably.)

      1. Thanks, David. No, I wasn’t confused. I started out with $15 budgets and have tried scaling back to $10 but that rarely gets me 1000 impressions. On the bids, what’s weird is that the recommended bid range does NOT seem to change from one author/genre target to another. I’ll be setting up 4 test ads in a row with authors ranging from 2K to 15K followers and targeting the same genre with each and for every one, BB suggests bids of (for example) $4 – $12.06. To the penny. So I put in a bid of $14 for each ad. I’m very gradually weeding out nonperforming author targets and find-tuning my image, but haven’t yet come close to what I’d call a “good” CTR (say, over 2%) or effective CPC (say, under $0.50). Just wanted to check in, in case I’m missing something obvious.
        Of course, it’s also totally possible I simply suck at both images and targeting. 🙂

    2. Sorry, I see what you mean now. I was the confused one! Probably would have been safe assuming that from the get-go…

      OK, so both your bids and budgets are fine for testing. I often bid at the upper end of whatever BookBub is recommending, or even a little more, when serving is more important than ROI – and testing is definitely one of those cases. When running an actual permafree campaign though, I’ll scale that back considerably until I find the sweet spot which will roughly serve my ad enough to spend whatever I want for the daily budget. The lower the budget, the more you can skimp on that bid.

      You’re at the trickiest stage right now, where your initial test results aren’t quite there, but it’s not exactly clear what you need to change to get over the hump. Unless your targeting is way off, I’d guess that the image might just need a tweak or two. Hard to say without more details, but perhaps at this point, take your best performing target thus far and try a few different image approaches – just with that one author so you don’t waste money – and see if you can inch that CTR upwards to where you want it to be.

      And then when you have a winning approach to the image, you can go back and test some of those other authors who aren’t quite there.

      Some authors get through the testing process reasonably quickly, and for others it takes a little longer and can be more frustrating. That doesn’t necessarily indicate their relative future success though. It gets a lot easier once you get over this hump. The difficulty curve is all at the start with BookBub – it’s not an endless rabbit hole of complexity like Facebook.

      1. Hi David,
        I have a question about this… Now that BB is returning results splitting into individual authors, is there any reason to test individually?

      2. You can’t just lump all your authors together in one test as the serving will be gobbled up by the bigger ones. It’s kinda done on a pro-rata basis but can be a little uneven. To minimize such issues, you can group together authors with similar numbers of readers and test in small batches. E.g. test four authors with 30k-40k readers each together, and so on. This is needed because you’ll only get a CTR-by-author breakout if they hit around 1000 impressions.

Leave a Reply

Your email address will not be published. Required fields are marked *