One of the trickiest parts of advertising a book via display ads is finding the right combination of elements to make someone not only stop to look, but actually click. And the best way to understand what works and what doesn’t is to test your ad images with a series of split A/B tests.
Book Brush challenged five authors (myself included — I’m an author who also works at Book Brush!) to create A/B split test ads, each testing a different image element, with the goal of identifying a winning design formula for BookBub Ads. Our testing journeys led to some unexpected conclusions about which elements drive readers to click!
The testing process
The trick to an A/B split test is to create two ads — A and B versions — and change only one element in the B ad. When you’re testing an ad image, all the other campaign parameters you’ve set for budget, bid, links, targets, and schedule should remain exactly the same for both ads during the test. That way you’ll know the difference in results is due to the change in your ad image.
Book Brush assigned each author a specific image element to test: image background, adding a blurb, 2D vs. 3D book cover, placement of the book cover, and call-to-action (CTA). Each author ran the test ads until they had a minimum of 2,000 impressions each, then compared the click-through rates (CTR) of each ad.
All the ad images were created in Book Brush. This powerful design software was created specifically for authors, and has all the features you need to quickly design eye-catching marketing images, including images for BookBub Ads. You can easily save each rendition as you make any necessary changes to your BookBub Ad variations.
Here are the results of the test campaigns.
Test 1: Changing the background element
Elizabeth Briggs’s task was to change only the background of her ad image and see how that affected her results. The background image often takes up the most space on any ad, so choosing the wrong one can kill an ad campaign. We asked Elizabeth to create one ad with a plain background color of her choice and the other using the enlarged book cover as the background.
“I thought the image with the cover as the background would do better, but it was the opposite! Maybe simpler is better sometimes. However, the CTR (click-through rate) was lower than I’d like for both ads so I’d probably test some other images and targets.”
Takeaway: Try a bold, plain-colored background to draw attention to your book cover and blurb.
Test 2: Including an author blurb or not
Steena Holmes was tasked with using either a tagline, review, or author blurb on one ad and not on the other. Authors often feel torn between keeping their ad clean-looking with fewer words, or using a blurb that might pull in more readers. Which worked best for Steena?
“It’s a pretty close tie between having a blurb on the image and not having a blurb (my preference is without the blurb because it looks cleaner to me).
For these tests, I selected 10 authors that normally have a high CTR (the average author I selected has given me between 2.1%–4.6% CTR in the past). I added a category because a few of the authors I targeted write in a few different genres, so I wanted to make sure I hit the right readers. I went with a high budget for the simple reason that this was a test and I wanted the results quickly. I wasn’t too worried about the bid amount.
This book is not on sale — its full price is $4.99 — but my main goal was to attract KU readers. If I were to continue testing, I would tweak the graphic a bit more, go down to five of the best author targets, and place the book on sale to see how different the results would be. All in all, this was an interesting test for a full-priced book. I can’t wait to test this with one on sale!”
Takeaway: In this case, it appears to be a toss-up between including a blurb or not. We suggest running your own test for your book ads to see if adding additional copy to your image increases clicks or conversions. Which blurb you choose can also make a difference — for example, results could depend on how recognizable the author is in your particular genre.
Test 3: Using a 2D cover vs. a 3D cover
I was so interested in testing this element that I took this one on myself! What is more important in any ad than the book cover? In one ad I used a flat 2D cover image, and in the other a 3D cover, which is a simple change to make in Book Brush. I chose an iPad for my 3D image because I sell far more ebooks than paperbacks.
“I thought the ebook with the iPad would be a clear winner because it makes the cover pop. At first, the ads were running neck and neck, but the iPad version did eventually win. I think my $15 CPM bid was too high, but I wanted a clear test and wasn’t focused on driving the sales that a lower-cost drip campaign would need, so I was OK with the higher spend. As the ads were all going to Amazon readers, I didn’t mention Amazon in the ad, but I might test adding that next.
I will continue this ad with a few changes, including a lower bid. I may also need to find a different mix of author targets in order to win impressions with a lower bid. The overall budget will be based on what I feel I can spend at that moment in time, and naturally on the returns. So I will monitor the campaign results and adjust my budget up or down accordingly.”
Takeaway: For me, it was better to show off my book in a 3D ebook format. Try testing 3D images of your own ebook covers on mobile phones vs. tablets, or showing a paperback image if you sell a lot of print books.
Test 4: Changing the placement of the cover
Candee Fick had three ads to test! Her design element was the placement of the cover in the ad: left, center, and right. Naturally, the position of the CTA with the price had to move, too, but the CTA itself stayed the same. Cover art placement is often overlooked, but it can be an important element of an ad.
“I targeted a combination of two categories (Christian Fiction and Contemporary Romance) and 17 comparable authors. According to BookBub, if a reader followed at least one of the authors and was subscribed to one of the categories, then they could see my ad. That gave me a fairly good-sized target audience that I knew would enjoy this story.
The winner? I’d say the left-side image was the clear loser while the other two options were close.
Since this test concluded, I have started running a single CPC (cost-per-click) ad with the winning image and currently have a higher CTR than during any of my tests. I’m continuing to tweak the budget and bids to serve more impressions to this same audience of contemporary Christian romance readers.”
Takeaway: Consider your book cover placement carefully, since it might impact your results. Also keep in mind how it may interact with your background image, if you’re using an image rather than a solid color. As always, test this yourself to see what works best for your own ads.
Test 5: Changing the call-to-action (CTA) format
Pauline Baird Jones was tasked to change the CTA format. We asked that the CTA use the same copy and be the same bold color of her choice in both images, with one designed as a button and the other using just the text against the background.
“My ad was targeted to Chirp audiobook listeners, which presented some interesting challenges, such as finding audiobook authors who also wrote time travel. I think it probably limited the reach and possible impressions.
I was glad I ran the ad for as long as I did (18 days). My ad spend didn’t come close to my budget, but by looking at the nifty ‘Performance Over Time’ graph, I noticed the clicks picked up around day 10.”
Takeaway: Both a bold and subdued CTA can work well, as shown by the tie in this A/B test. You can also try testing different colors for your CTA to see if one draws more clicks than another.
Small changes to your image design elements can impact ad results, but we were also surprised to see that some changes didn’t have an effect on clicks at all! I hope you’ve found it helpful to see these actual ad tests, their results, and the authors’ observations.
Remember to test these elements in your own A/B split tests, as you may get different results depending on your particular book, the genre you’re reaching, or the precise audience of readers you target. Constantly running new tests is the best way to make sure you’re getting the best results possible for your ads!
Want to share this post? Here are ready-made tweets:
Click to tweet: Interesting… five authors ran A/B split test ads testing different image elements within a @BookBub Ads campaign. Check out their results! https://bit.ly/39yolYv
Click to tweet: Authors! Have you run A/B split tests for your @BookBub Ads designs? https://bit.ly/39yolYv