In an earlier post, I discussed stalking you on Facebook. The you I stalked proved less than productive. I resolved to try again. This post documents my latest attempts to find a more productive you.

A proactive marketing department looks out into the world and identifies unmet or underserved customer needs. An author identifies emerging or surging areas of reader interest and crafts a story and a book to satisfy an audience’s hunger. Or their clues them in. Harry Potter’s success heralded years where Young Adult Fantasy was hot. Savvy authors with empathy for, and understanding of, that audience had an audience waiting for them if they could deliver a quality story.

Tensile Press’s marketing department (me) is not proactive. Raether’s Enzyme was written with only the haziest notions of who its audience might be and no data to back those notions up. Since the book’s publication, I have been struggling to divine the traits of readers who might enjoy the story. Facebook ads are my divining rods du jour.

WARNING: This post is NOT a how-to guide for Facebook advertising.

It’s a what-I-did story. Laugh or sigh at my mistakes. Root for me making slow progress toward enlightenment. Don’t try this at home and expect great results.

The Experiments

The previous experiments cast a wide net and caught no hungry eyeballs. Sad trombone.

The key to narrowing the net is hidden in Facebook’s ad-building user interface. With what is immediately visible, you can easily constrain demographics and assemble a list of interests. It takes a bit more poking around to unlock the amazing power of AND. With AND in your toolbox, you can take your base list of interests and say, “People interested in any of THESE, AND they must be interested in THIS as well.”

This narrows the audience for the ad considerably. Starting from a core audience and varying the AND portion, I set a series of divining rods aquiver. Facebook provides all manner of statistics for an ad’s performance. Key among them for my purposes was the Click-through Rate. Of all the people who saw the ad, how many at least clicked on it? Raether was enrolled in KDP Select. That made the e-book exclusive to Amazon, so I had all the ads link to its product page. I could see if there were any sales or Kindle Unlimited page reads associated with clicks on the current ad.

Experiment #1: Neal Stephenson

Location: United States
Ages: 24-65+
Interests: Speculative Fiction, Suspense, Thriller Novels AND Neal Stephenson.

A friend and early reviewer likened Raether to some of Stephenson’s work.

Experiment #2: Margaret Attwood (A/B Test with Neal Stephenson)

Location: United States
Ages: 24-65+
Interests: Speculative Fiction, Suspense, Thriller Novels AND Margaret Attwood.

There’s overlap between Megan’s and Offred’s predicament. Their freedom and agency are threatened, in part, because of something special in their biology.

Facebook allows for A/B testing. So, I ran Experiment #2 as an A/B test with the second half of Experiment #1. I might not be able to tell which clicks to Amazon came from which ad, but Facebook had something interesting to say about which of the two ads was more effective at getting those clicks. (Spoiler: Margaret, by a substantial margin.)

Experiment #3: Speculative Fiction

Location: United States
Ages: 22-65+
Interests: Suspense, Thriller Novels AND Speculative Fiction.

I shuffled the interests around. And lowered the age range by a smidgeon.

Experiment #4: Speculative Fiction – New Kids Mix

Location: United States
Ages: 18-49
Interests: Suspense, Thriller Novels AND Speculative Fiction.

Facebook was reporting that it was tending to show the ads to older readers. I was curious if we were missing out on younger readers due to the algorithm’s caprice.

Experiment #5: Gillian Flynn (no relation)

Location: United States
Ages: 18-65+
Interests: Suspense, Thriller Novels AND Gillian Flynn.

I was talking with another friend and early reviewer about the previous experiments. She suggested reaching out to Gillian Flynn’s audience. This was thinking outside my box and just the sort of help I needed.

Experiment #5: The Girl on the Train

Location: United States
Ages: 18-65+
Interests: Suspense, Thriller Novels AND The Girl on the Train.

Along with Gillian Flynn, my friend suggested Paula Hawkins in general and The Girl on the Train in particular. Facebook had the book, but not the author, as an AND-able interest.

Experiment #6: Hard Science Fiction

Location: United States
Ages: 18-65+
Interests: Suspense, Thriller Novels AND Hard science fiction.

Raether can be viewed as science fiction. It asks, “what if?” and explores the implications for its characters and their world. “What if cancer could be cured by a substance found in the bodies of a tiny minority?” This is somewhat fantastical given what we know about the varying causes of cancer. But it might still be possible.

Experiment #7: The Hunger Games

Location: United States
Ages: 18-65+
Interests: Suspense, Thriller Novels AND The Hunger Games.

Raether’s protagonist is a bit old for the Young Adult market. Her dystopia is ambiguous. I was curious if there might be any cross-over interest from fans of The Hunger Games.

Experiment #8: Young Adult books

Location: United States
Ages: 18-65+
Interests: Suspense, Thriller Novels AND Young Adult Books.

While I was in the neighborhood, I reached out to the wider YA audience. I left the age range open. There are a substantial number of adults who enjoy YA fiction.

The Results

In General

The audiences Facebook found for the ads tend to be older and female. There’s no way to tell if that’s an accurate picture of Raether’s audience or an artifact of who’s on Facebook these days and what Facebook happens to know about them. The age trend was bucked in Experiment #4 (artificially) and in Experiment #8 (organically, by Facebook’s reckoning).

Caveat Re Specifics

I was sloppy on controlling variables in these experiments. If an experiment was not competitive after a couple of days, I cut it short. Hey, I’m paying by the day for this advertising. This means that the raw numbers of impressions (how many times the ad was seen) and an ad’s reach (how many distinct people saw the ad) can’t be compared directly. The Cost Per Result and Click-Through Rates and the are comparable.

Experiment #1: Neal Stephenson

Cost per Result: $0.31
Unique Outbound CTR (Click-Through Rate): 11.68%

Experiment #2: Margaret Attwood (A/B Test with Neal Stephenson)

Cost per Result: $0.16
Unique Outbound CTR (Click-Through Rate): 11.30%

Experiment #3: Speculative Fiction

Cost per Result: $0.18
Unique Outbound CTR (Click-Through Rate): 9.96%

Experiment #4: Speculative Fiction – New Kids Mix

Cost per Result: $0.21
Unique Outbound CTR (Click-Through Rate): 6.9%

Experiment #5: Gillian Flynn (no relation)

Cost per Result: $0.16
Unique Outbound CTR (Click-Through Rate): 9.06%

Experiment #5: The Girl on the Train

Cost per Result: 31
Unique Outbound CTR (Click-Through Rate): 4.63%

Experiment #6: Hard Science Fiction

Cost per Result: $0.27
Unique Outbound CTR (Click-Through Rate): 7.17%

Experiment #7: The Hunger Games

Cost per Result: $0.20
Unique Outbound CTR (Click-Through Rate): 7.17%

Experiment #8: Young Adult books

Cost per Result: $0.62
Unique Outbound CTR (Click-Through Rate): 3.23%

Right, but what about actual sales?

I’m getting to that. While I was refreshing my Facebook Ads Manager dashboard, I was watching my KDP sales dashboard. I attributed book sales to the ad that was running at the time. Kindle Unlimited page reads ramped up and tapered off. Some of the page reads in the early part of one ad’s runtime may be readers who picked up the book during the previous ad.

Here are the KDP report graphs. I added colored overlays to illustrate which experiment was running. For the purposes of Kindle Unlimited, Raether’s Enzyme is approximately 500 pages.

Experiment #5 performed poorly and isn’t highlighted on this table. Which isn’t to say any of the ads did amazingly well, but I am happy with Gillian Flynn, Speculative Fiction, and Hard SciFi. There are intervals where the sales and reads come close to covering the cost of the ads for a given day. But…

It’s not NaN, nor is it a profit.

Previous attempts resulted in no sales. When you divided the cost of the ads by the royalties earned, you were dividing by zero. That’s Not a Number. This time around, the ads went to closer approximations of Raether’s audience. Some of the Unique Outbound Clicks translated into sales and Kindle Unlimited reads, both of which generate royalties. Some royalties are not zero royalties.

July’s Kindle Unlimited Royalties are still pending. In the month of June, I spent $140.14 on Facebook ads and earned $81.83 in total royalties. That’s a long way from breaking even, but I’m learning things and connecting with a few new readers, who, in turn, may recommend the book to their friends. For indie authors, that’s where the hope lies.