Your author friends are lying to you about your book cover.
But they’re not doing it on purpose.
When you post two cover options in your author Facebook group or email and ask people which one is better, your friends look too carefully. They study the details, think about your feelings, notice which option already has more votes, and then give you an answer based on what they like.
News flash: what they like does not matter. What you like does not matter.
The goal of a book cover is to get complete strangers, who do not know or care about you, to stop scrolling on Amazon and click to learn more. If your cover cannot get the click, it is a bad cover, even if everyone likes it.
Your cover will always appear in a sea of thumbnail book cover images, even on its own Amazon page. In a bookstore, there is a 95% chance your book will be shelved spine-out, with no cover showing.
No stranger will ever look at your cover as closely as your friends do when you ask for advice.
So, stop asking for your friends for opinions. Instead, use a scientific method called split testing.
I asked Genesis Award-winning author, Navy veteran, and thriller writer J. A. Webb about his experience using split testing.
How did you use split testing?
Thomas: You had an award-winning book that struggled to get readership. No one will know if they like the writing if they cannot get past the cover. Tell us your story.
J. A.: Writing a book is like preparing a gift, wrapping it, and putting a bow on it. You want to give it to your reader to open it and enjoy it. But if they can never find your book, that will not happen.
I published my first novel about 15 months ago. Those in my target readership who found it loved it, but finding those people was difficult. The ads I ran were a total flop. It was a frustrating experience. I could not figure out what was going wrong or where the disconnect was.
Thomas: This is a common experience. People try running ads and don’t get any sales. Most authors jump to the wrong conclusion that, “ads don’t work.” The real question is whether something is wrong with the offer or the cover.
Just because ads are not working for you does not mean ads are not working for anyone.
How did you discover the real problem?
J. A.: Ads were not working for me. I am in Thomas’s mastermind group, and we analyzed my ads together to figure out where the disconnect was.
You suggested right away that it was probably a bad cover. As it turns out, you were right.
I knew when I chose my cover that I did not know enough to choose one, so I hired a professional. The designer did a fantastic job. The covers were gorgeous and well-received.
But, as you suggested, we learned the problem was that the covers, especially as thumbnails, were attracting the wrong reader. When the wrong reader clicked through, they said, “Not for me.” The right readers never clicked because the cover did not communicate that the book was for them.
That started me on a long and painful journey. It was not just cover placement. It was also genre placement. We had to figure out where my target reader was.
What makes a for a “good” cover?
Thomas: Everyone told you the cover was great. You got universally positive feedback from authors and readers, posted the cover online, and everyone said it was amazing. Even you liked it. Plus, you spent a lot of money on a designer.
But all of those data points are garbage.
J. A.: I did not throw something together. I knew I did not know how to design a cover, so I hired a professional, assuming a professional would know what works. I have since discovered that is not necessarily true.
Thomas: It is not necessarily true, especially if you do not have the right genre selected.
One best practice in cover design is to stay in the middle of the road for your genre. Your cover is not where you want to innovate.
If you design a dystopian-looking cover for a thriller, you will attract dystopian readers who may not want a thriller with dystopian elements. Dystopian often works as a flavor and not as the focus.
How does genre confusion affect your cover?
J. A.: That was one of our struggles. We had to determine what genre my books actually belonged to.
Aesthetically, they are speculative and set in a dystopian future, but they are not dystopian novels. They are thrillers in structure and narrative. That was something we worked through using your tools.
Until we pinpoint the genre, we cannot do a proper job with the cover.
Which Patron Toolbox tools did you use?
J. A.: I used all of them that were available at the time and have continued using them as new ones have been added. Every time I run a tool, I learn more about my book, how it fits in the publishing world, and where it should be marketed.
Thomas: One tool I developed and refined based on your feedback was the Trope and Genre Finder. It reads your book, identifies tropes, and gives an alignment score with different genres.
Instead of saying, “This is mystery,” it might say, “You have 70% alignment with mystery and 85% with thriller.” That helps you see where your book truly fits.
Sometimes authors think their book belongs in one genre when it actually fits better in another. Readers in that better-fit genre are more likely to read and review it.
How do you actually find a good cover?
J. A.: The Trope and Genre Finder and Comp Finder helped narrow things down, but we still needed a good cover.
My original strategy of merely hiring a professional had failed. I needed to recover my books, but I did not want to repeat that same mistake. Hiring a professional is still an educated guess. They know the market, but they cannot guarantee performance.
You only know if a cover works when it is tested in the marketplace.
Based on your recommendation, I tested multiple covers in the actual marketplace through split testing.
It can feel like a lot of work and expense, but it is cheaper than running ads that do not work.
Why do authors avoid split testing?
Thomas: Many authors hesitate to split test because they are not ready to run ads. When preparing their first book, they believe it will be a bestseller. They think readers will naturally find it.
After a few months, when sales drop, they realize they need to acquire customers. Then they start running ads with a cover and copy that often does not work in the marketplace.
What they should do is run ads before the book launches, before choosing a final cover, but that requires humility.
If you have a strong cover, your marketing, PR, advertising, and even Amazon’s algorithm all perform better.
What should authors test besides the cover?
J. A.: Test everything. Test your cover, ad text, sales copy, and headlines. Break them into elements and test each one. If you want to know your assets will work, you have to test them.
In my case, I needed a new cover. Following your method, I hired three professional designers and had each provide multiple concepts. It was not cheap, but it was cheaper than running ads and losing money because of a bad cover.
I ended up with 10 variants. I also included designs generated by your Book Cover Designer tool in the split test. The covers generated by your Book Cover Design tool consistently ranked at the top in my tests. The data showed that even a professional’s work is still an educated guess until tested.
I ultimately chose one of the tool-generated covers, even though I didn’t like it at first. It performed best, and I learned to love it. I refined it and used variations of that design for the first three books in my series.
Why do authors struggle to create covers readers want?
Thomas: Developing covers readers want is sometimes like developing food children like. My wife and I are true Texans, and we grew up eating Tex-Mex with bold flavors. Our children, on the other hand, do not like anything spicy. In fact, one of the first words my daughter learned was “’picy.” She would say, “This too ’picy for me.”
The bold flavors we love make food less appealing to our children.
One time we served boiled chicken, boiled vegetables, and rice. It was plain, but our kids devoured it.
To find food your children like, you have to get into their minds. The same is true for readers. You must get into the mind of your reader and realize your reader is different from you.
Being an author can be a disadvantage because you have sophisticated tastes. Let me ask you, do you like Academy Award-winning films? I hate them. If a film is nominated by the Academy, I assume I will not enjoy it. It will be boring or offensive, with nothing redeeming for me as a viewer. But people who are deeply into cinema love those films simply because The Academy voted for them.
How did you run you split tests?
Thomas: How did you run your split test? If I have three versions of my book cover, how do I find out which version will get the most clicks in a crowded visual environment?
J. A.: You test it in the marketplace. In my case, I used Facebook ads with the cover image by itself. No ad text. Nothing else. I wanted to know if the image worked. I did not want clicks because of a headline. I wanted to know which image got the most clicks, so I ran ads for all 11 covers and compared the results.
Thomas: If I am browsing Facebook and see one of your covers and tap or click on it, where would the click take me if the book is not for sale yet?
J. A.: I had the ads land on my website. I already had two books for sale. Even when I was split testing cover images and simple headlines, those clicks generated some sales. But the purpose of those ads was to find out which element performs best in the marketplace, not to sell my previous books.
Thomas: The first time I used this technique was for a political campaign around 2010. We ran Facebook ads with a picture of the candidate and tested single words like conservative, patriot, and Republican.
Every ad was identical except for one word. We discovered that “patriot” resonated far more than the other terms. He began using that word more often in his speeches because we knew it connected with voters. Just because it worked in 2010 does not mean it would work in 2014 or 2028. This tool is powerful because it gives you current data.
For about $50, you can put a split test in front of tens of thousands of people. Even those who do not click are voting. Not clicking is a vote that the image failed to capture attention. You use all of that data to inform your marketing.
Did you use bracket testing or multivariate testing?
J. A.: I have heard you recommend bracketed A/B testing, where A competes with B and the winner advances. Is that what you did with the political campaign?
Thomas: We would test two words, pick the winner, then test that winner against a new challenger. We also tested different photos. For example, a photo of the candidate alone versus a photo with his family.
At one point he was endorsed by Chuck Norris, which was a big deal in Texas in 2010. We tested different Chuck Norris images to see which performed best.
It was like boxing. The champion had to defend the title against each challenger.
There is a more complicated method called multivariate testing, where you test many versions at the same time. I do not recommend it for most authors. Did you use that?
J. A.: I did. I ran 11 ad variants simultaneously.
Bracket testing would have increased the cost and extended the timeline. Running four to five days per round for 11 rounds would take a long time.
I found a method developed by Steve Piper for author click testing. It allows you to test multiple variants at once, which compresses the timeline and lowers the cost.
The other advantage is that you get current data. I had a winter scene that performed very well before the holidays. In January and February, when people were tired of winter, it stopped performing. No one wanted to see snow anymore.
If you spread tests over a long period of time, you introduce another variable: time and audience. Audience perceptions shift. Demographics shift. People are on Facebook one week because of a current event, but don’t check in when that event is over. Running everything at once means you are hitting the same audience at the same time.
How does cultural change affect covers?
Thomas: We are in a period of rapid cultural change. On our other podcast, Author Update, we now discuss the changing zeitgeist every week.
That is big-picture change. You are describing small-picture change. Winter is magical until it is not. Then people are asking when spring will arrive.
There is a seasonal rotation. The kinds of stories and images that resonate shift throughout the year. A book someone wants to read on the beach is different from a book they want to read with cocoa in front of a fire.
Where should my ads send people?
Thomas: You sent people to a landing page where they could buy the book, even though the cover they saw was different. That is ideal. Some money is better than no money. But you do not need to have a published book for sale to run tests.
If you are unpublished, create a landing page with the book cover and a few paragraphs that say, “Amazing book coming soon. Click here to be emailed when it is available.” Include a newsletter signup form.
Will it be the most cost-effective way to grow your list? No. Will it add high-quality readers? Yes.
You can even say, “See which cover wins the test,” if that fits your audience. You can do all of this before your book is on Amazon or Kickstarter.
J. A.: For some audiences, that would be engaging. My readers are busy adults and wouldn’t care about the recovering process, but you have to know your audience.
But the signups from those landing pages are high-value. These are people who clicked because they are interested. That is very different from people chasing free books in a giveaway.
Thomas: Right. The pitch is “Find out when you can buy it,” which attracts buyers. Free offers grow your list faster, but buyers are higher quality leads.
Facebook works well for testing because it is noisy and chaotic. That environment is similar to Amazon. Every scroll reveals more covers and constant visual competition. Your cover must stand out but also fit in. It must appeal to the right reader in that noisy context. Facebook is a great place to test that.
To learn more about finding the right readers for your book, listen to our episode on How to Find Your Timothy.
What happened after you changed the cover and genre?
J. A.: After changing the genre positioning and the covers, I now have two months of data showing instantly profitable Amazon ads. It is only coffee money, but before the changes, the ads were guaranteed losses. I lost money on ads for a year before I changed the cover.
With the right cover and genre fit, the ads became profitable. Now I can think about scaling. Before, scaling would have meant bigger losses.
The tools, the mastermind work, and the split testing were huge.
Will my books become bestsellers? I do not know. I do not write romance or romantic suspense. I write fiction for readers not well served by standard genre fiction. It is a small slice of the market, and finding readers will be slow. But I could not have run more ads to find them if the ads were not profitable. Now they are, and that changes everything.
Thomas: You have no idea how high it can scale. I interviewed Connor Boyack, an author who got his ads profitable and started scaling. When I interviewed him, he had spent $3 million on ads in one year to make $10 million in revenue.
He scaled little by little. He took his winnings and kept putting them back into ads.
The difference between profitability and unprofitability seems minor, but financially it is a step change. It is like the difference between 33 degrees and 31 degrees. At 33 degrees, you get cold rain. At 31 degrees, you get snow. The shift between 34 and 33 degrees changes nothing, but drop two more degrees and everything changes.
Approaching that inflection point, the moment of profitability, can feel gradual. Then suddenly the rain turns to snow.
How did you choose your ad targets?
Thomas: I want to underline that these were marketing changes. You did not rewrite the book. You changed the categories and the cover so that they matched. By pivoting into a better-fitting category, you crossed that threshold.
J. A.: Better-fitting targets improved my Amazon targeting. Knowing that allowed me to choose better comps.
Thomas: If you target the wrong people, you get bad data. If I write for young men but my ads reach older women, their response does not tell me whether my intended audience will like it. That is an extreme example but targeting matters.
Thomas: Walk us through your targeting process. Once you choose it, you are effectively locked in for the test. If you change the target, you must restart your data. So how did you select it?
J. A.: Amazon ads and Facebook ads are very different. On Amazon, you can target narrowly. On Facebook, it is harder, especially because I write overtly Christian fiction. My first three books are adventure thrillers that address truth, faith, and specifically the Christian faith. That limits my targeting.
Facebook has made it nearly impossible to target Christians directly. If I wanted to target fans of the Left Behind series by Jerry Jenkins, I cannot. If I wanted to target people who like Christian thrillers, I cannot. Facebook targeting is difficult. For Christian authors, it can be especially challenging.
Thomas: This has been an issue for a long time. People outside the Christian writing world often do not believe it. But try promoting a Christian book on Facebook. It is tricky.
Every Christian author who succeeds with Facebook ads seems to find a different way through that wall. If you write secular books, you can target fans of similar authors or films. With Christian content, even that can be restricted. You can target HBO’s Rome, but not The Chosen, even though both are historical shows set in ancient Rome.
What strategies help Christian authors with Facebook targeting?
Thomas: It often comes down to targeting fans of brands like Hobby Lobby or Chick-fil-A. You look for the few Christian-adjacent brands that are still targetable.
Do you have other advice?
J. A.: I am not an expert. I am someone who had to learn this to make my books work.
I found Steve Piper’s training helpful. His Click Testing for Authors program offers extensive instruction with 12 hours of training. That is where I would start, especially for Christian authors.
Thomas: Another approach some Christian authors use is lookalike audiences. If you have 100 or 200 true fans in your micro-genre who are Christians, you can upload that list of email addresses and ask Facebook’s AI to find similar people.
That did not work for you because your early audience came from the wrong genre with the wrong cover. A lookalike audience based on that data would just replicate the mistake.
It is like starting a writers’ group on Tuesday nights and then asking attendees what the best night is. Tuesday will always win because everyone who could not come on Tuesday is absent.
Did your genre change or did you change genres?
J. A.: In my defense, I did not leave fantasy. Fantasy left me.
Thomas: Many authors feel that way. Genres evolve. Readers drift in and out of them. Think about your own reading. What were you reading 10 years ago? Twenty years ago? Has it always been the same genre?
Right now, for fiction, I am reading litRPG. For nonfiction, I am reading Aristotle. That has not always been the case. There were seasons when I read highbrow fiction and lowbrow nonfiction.
Readers oscillate. As you target your ideal reader, remember that reader is a moving target. Life changes and so do genres.
For more on cultural shifts and changing readers, listen to my other podcast, Author Update.
What advice do you have for authors who suspect their cover needs to change?
Thomas: What would you say to an author who loved their cover, received glowing praise from friends and subscribers, but sees that it is not working in ads or on Amazon?
J. A.: Something is probably broken. You need to identify it and fix it. It is painful. It was painful for me to decide to go down this road.
Education, time, and split testing require investment, but it is not as expensive as running Amazon ads for three years that do not work. It is not as expensive as the opportunity cost of never finding your readers.
We are offering a gift to readers, and you’re doing them a disservice if you’re making it hard for them to discover. You owe it to your readers to make your work discoverable.
Thomas: You also owe it to yourself. Your past self invested hundreds or thousands of hours writing this book. Your current self is the product of your past decisions.
Wrap that book in a cover that compels readers to click. All decisions about whether a book is worth reading happen before the reader makes a judgement on your writing or your story. Readers make instant, subconscious decisions to buy and read based on the cover. They scroll past hundreds or thousands of covers. The quality of the writing does not factor in at that stage.
You can hide excellent writing behind a beautiful but ineffective cover.
Often, no one in the author’s life will tell them the cover is not working. The only authoritative way to know is to test it.
I once pitched the idea of split testing to a vice president at a major publishing house. He loved it, but they never implemented it. Traditional publishers often avoid quantitative measurement, because people get fired over quantitative measurement. Numbers can lead to uncomfortable accountability.
Some also hesitate to test covers publicly. An agent once asked, “What about people who click the B cover if you do not use it?” He saw that as a tragedy.
But it is better to have a handful of people click a test cover than to let everyone see the wrong one forever. There are effectively infinite people online to test with. You are sampling a tiny fraction to discover what resonates. You owe your book that test.
Most of the expense is education, such as training from someone like Steve Piper. The ad spend itself is modest.
How much did you spend per round to test 13 covers?
J. A.: Per round, $50 to $70 was enough to get statistically relevant data. To test covers, headlines, and sales copy elements, you might spend $600 to $700.
But you will spend that much on Amazon ads in six months anyway. The question is, do you want them to work?
What was the real cost of reinvention?
Thomas: The real cost of this reinvention was not the ads. It was the training from Steve Piper and buying all those covers.
The worker is worthy of his hire. Just because a cover designer’s concept did not win the ad-click competition does not mean they should not be paid. It is not the designer’s fault that their cover lost. You may not hire that designer next time, but you absolutely pay them for the work they delivered. I do not believe in saying, “I will pay you only if your cover wins.” That is a terrible way to run a business.
No self-respecting cover designer will offer their services for free unless they win a contest.
J. A.: I agree. You need to purchase the cover concepts from your designers to run the tests.
Thomas: Be transparent. Tell them, “I am going to pay you regardless, but I want draft versions so I can test them.”
Sometimes a designer gives you three versions and asks which you prefer. You need to clarify in advance that you plan to test all three before choosing a final version and making final tweaks.
J. A.: When you ask for three versions, make sure you specify three completely separate concepts.
I had a situation where a designer gave me three variations of the same cover. That is not what you want. You want three entirely different approaches.
Thomas: Right. You don’t want simple font color changes on the same image.
What kind of covers often win in split tests?
Thomas: The Book Cover Designer tool in the Patron Toolbox often wins these tests because it creates simpler covers. There is very little nuance in the background. The tool is designed to create covers that work well as thumbnails.
Most authors want their cover to capture the full essence of the book, with layered visual elements and nuance. That often creates noise.
In the Book Cover Designer tool, you can only describe one primary visual element. That is it. The result is a simple design.
A simple cover with one strong visual element will usually outperform a complicated work of art that only works when someone studies it in a gallery, which is not how readers encounter covers while scrolling on Amazon.
Connect with J. A. Webb
Featured Patron
B.D. Lawrence, author of Final Jeopardy: A Jake Sledge Mystery
Jake Sledge, a hard-hitting PI must untangle a deadly family legacy and confront his own violent past in order to prove a troubled man innocent of murder. But will he have enough time when a relentless killer silences everyone who gets too close to the truth?
Related Episodes
- Ten Things Every Book Cover Needs to Look Legit (Ep 106)
- Book Cover Mistakes That Sabotage Sales (Ep 107)
- Effective Book Cover Design with Kirk DouPonce
- How to Create a Design Brief for Your Book Cover
- How to Create Powerful Ads for Authors Who Hate Math with Chris Fox
- Book Marketing 101: Product (the “making the right promise” framework)
Watch on YouTube
