You may recall that I was skeptical about BookBub Ads when they were first announced. While it’s exciting to get a crack at that population of known book buyers, I didn’t really think CPM advertising (paying for impressions, rather than clicks) made sense, since they count opening the email as an impression. Yet you have to scroll down to see the ad. That means that a lot of those “impressions” aren’t really impressions at all, so we would expect the CTR (click-through rate) to be low. Well I’ve run a lot of tests, and I can tell you that my skepticism was justified. The CTR on BookBub ads is horrible. Pretty consistently about 0.4% for what should have been a very clickable ad. (See above.)
To make sure I was being fair to BookBub, I just ran a Twitter ad campaign. I’m using this tweet, which I think is pretty much the same creative as that ad:
Okay, so it’s got more going on than that ad, but BookBub gives you a pretty tiny footprint, so I did my best to capture the essence.
I tried targeting Erotic Romance, Contemporary Romance, and Chick-Lit readers. And I tried targeting Amazon and iBooks readers. And I got the same 0.4% CTR for all of them. I also found that all of those markets had the same CPM: $2.30. I couldn’t get any impressions at a $2 CPM bid, and if I bid $2.50, I’d get a bunch at that $2.30 level.
After spending $70 on BookBub, I had 128 clicks that converted to 5 sales. That’s 4%, which is neither good nor bad. It’s just a normal conversion rate. (My first test got a pretty high 7% conversion rate, but that didn’t hold up, so it was just sampling error, apparently.) 128 clicks for $70 is a CPC of $0.55. Let me put these numbers in context by comparing them to that Twitter campaign.
(CPM: Cost for 1000 impressions; CPC: Cost per click; CTR: Click-thru-rate; Conversion: % of clickers who buy; CPS: Cost per sale.)
In BookBub, you say “I’ll spend $2.30 for 1000 impressions” while in Twitter you say “I’ll spend $0.12 when someone clicks the link.” So the click-thru-rate in Twitter doesn’t really matter, except that if it’s low, you won’t get as many impressions for your bid. The first three lines in that table are really data about the marketplace. The next line is data about the audience (or how well I’m targeting the audience). The bottom line is the bottom line: What is each of these sales costing me. If BookBub had a killer conversion rate, it could make up for the high CPM or the low CTR. But it doesn’t.
Note that the CPS is not doing that great for Twitter, either. It’s costing me four bucks to sell a three buck novel that I make two bucks on. Also, those Twitter impressions are almost entirely coming from the UK. It seems the US advertising market has not settled down from the inflated US Election ad rates yet. (This especially sucks because I make even less on UK sales because of that insidious VAT that has to appear in the list price; in the US, sales taxes are on top of the list price.)
So here is what we can conclude: BookBub ads are a terrible way to sell books. I think other advertisers have figured this out, because the ads I’m seeing in my BookBub emails are all for home security systems. And I think, in a way, that’s the reason the book ads don’t work. The CPM for BookBub book ads should be about $0.50. But the CPM for ads in general is almost five times that high. And since BookBub is mixing book ads from their publishing partners with general ads from anybody, those general ads drive up the book ad CPMs to make them unjustifiably pricey.
A couple other notes if you want to try BookBub ads: their support people are very responsive and very nice. But they don’t understand the ad market at all. (The nice woman suggested I narrow my targeting to just followers of individual authors, to lower my CPM. I explained that narrowing would certainly increase my CPM, because that’s how markets work. She checked with someone and apologized for the bad advice.) Also, their dashboards are riddled with math errors. So be very careful when you look at the data. I reported all these errors, so I suspect they’ll fix them soon. Daily numbers seem to be reported reliably. It’s the roll-ups that are messed up.
If anyone from BookBub happens to read this post, here’s my advice if you want to make BookBub ads work for authors and publishers of books:
- CPC bidding, not CPM, since your impressions aren’t impressions
- Better demographic data, so we can at least tell who is clicking the ads
- Throw out the random advertisers and have a pure book market
Every one of those things is going to cost BookBub money, so don’t bet on any of them ever happening. For now, I’m going to stick to Twitter ads and lower my CPC bid so I’m not losing money on every sale. Because you should never pay people to read your book.