The past couple of weeks, I tried Facebook ads to boost my book sales for several of my novels. I learned several facts other authors don't bother to tell you that can cost a lot of money fast. I thought I would write up my experience for my cohort at the 2024 Taos Toolbox Writers' Workshop.
1. Facebook doesn't always sell your backlist effectively, but it does tell you how marketable something is, be it pitch or cover. If you're trying to decide between a set of covers, A-B testing will tell you fast which one people flock to. For example, book two of my Oni series gets clicks seven times more often than book one. Of my 34 novels, I can tell within a day and $5 to $20 of investment per experiment which ones will sell best, even if FB might not be the platform for long-term marketing. I knew this fact before, but these experiments underscored the lesson and allowed me to let go of a few of my orphan projects while encouraging me about others I had given up on. Make marketing decisions based on numbers not which darlings you love the most.
2. To verify which sales are due to FB ads or a specific ad rather than driver-by Amazon bargain hunters, you need to give each ad a different Amazon tracking URL, available for free for each campaign in your Amazon Advertising platform. Otherwise, you can't tell cause and effect reliably.
3. The possible high price per click means this is not a good vehicle for promoting free books. For an initial boost to an event, it can be great. For $50, I got 250 downloads in a couple days to put my new fantasy thriller Future-Proof at the top in each of its categories. However, this is not sustainable. Facebook was charging about 13 cents a click, but I was paying about 26 cents per download. Even if I used my cover photo and pitch verbatim, only half the clickers bothered to take my new book for free. FB optimizes for people who click LEARN MORE, not people who actually buy.
4. People surfing Facebook are cheap. Unless you have tens of thousands of rave reviews, nobody on the platform will buy for full price. Even then, it will be rare. Ninety percent of the ads are for 99 cent books. For a series, this means lowering the price for the first book or bundling the entire series for a just under a dollar a volume. For example, a five-book series would typically be $4.99. How do I know this? Because when you advertise on FB, you become a primary target for ads. Every third story on my feed is now a fantasy or sci-fi ad. If you can stick it out for a week, the FB algorithm can optimize the price per click downward. However, this means you only earn 30 cents a copy on Amazon for book one of a series.
5. Schedule ads ahead of time to be certain. I used KDP countdowns to lower prices for most experiments. You need to do this a couple days in advance. Once it's hot, FB has to be delivering the same hour, or you've wasted the perk for 90 more days. Especially over the weekend or overnight, approvals for new FB ads can have long delays. Have everything queued up in the missile silo ahead of time, ready to launch.
6. Facebook ads are only profitable if you're selling a series. My Jezebel series has five books and a high read-thru rate. That means for every Jez sale I make at 30 cents, I sell an average of 1.2 others at 3.99 each. So for that first week, I can spend a ruinous $5.10 a download and still break even while the algorithm finds my audience. On my two-book Oni series, I only sell an additional .25 of the lone sequel, limiting my budget to .3 + .25*3.99 = $1.30. This can work. For two days, I sold a copy of It Takes an Oni, every hour for a cost that dropped from $1.5 each to $1.25.
7. Once the targeting AI does down an ineffective rabbit hole, you can never change its direction. It equates clicks with success, no matter what return you see on the Amazon board. It could accelerate very fast in the wrong direction. On the Oni ads, I turned my back while I ate dinner, and FB spent the rest of my daily budget $8 for no return. I had to kill the ad immediately. On the flip side, I've had fully-funded ads stall out with no exposures for days because the AI gets stuck in navel-staring mode, even after I toggled it off and changed something trivial to restart the same ad.
8. When marketing book one of a series, use only the cover for that book. A carousel ad for my Gigaparec series using every cover in the series garnered 30 clicks in the first hour, but only one sale, wasting $8 fast. I saw this effect again and again. Be painfully honest with people about what they are getting. You want to attract the right audience for your product, not any rando off the street. If you get a loyal customer, they read other books in your portfolio, too.
9. It doesn't hurt to add an overlay on the ad to tell people you are available on KU. This helps the FB AI target people with Kindles who participate. However, people who add the phrase Free on Kindle Unlimited hack me off because I clicked based on the word "free" and feel cheated. Be honest.
10. AI generation of your pitch gets clicks, but seldom resembles your actual product. People who compare it to your actual Amazon page or sample content will be disappointed. Every time I experimented, I'd literally get ten clicks with zero conversion. It would claim my story was primarily a romance or be misleading in other way. You don't have to market your sushi as cold dead fish, but don't claim it's tenderloin with a side of fries. However, you can look at how it rewrites your pitch five ways and learn from it.
