The Facebook advertising experiment concludes

I previously blogged about an experiment in Facebook advertising that I undertook. The gist of the experiment was to see if I could generate an appreciable bump in sales of my e-book The Taint by doing a Facebook ad campaign, and if so, whether that bump was big enough to cause the campaign to pay for itself. And now that we’ve reached the middle of June, the campaign is long over, the sales numbers are in, and I can offer some data and some conclusions.

Preliminaries

Facebook offers you a pretty wide variety of options for how you set up your ad campaign. I went with the defaults for pretty much everything. Most importantly, I chose the automatic bidding strategy, which means that I let FB choose how it was going to spend the money I allocated for the campaign. FB, like Google, allocates ads based on a “bidding” metaphor in which different ads indicate via a virtual “bid” how badly they want to appear in a particular slot. As I understand it, this means that every time FB wants to show you an ad, it goes through all of the active ad campaigns that match something in your profile and asks them how much they’re willing to spend to be shown at that moment. Based on some algorithm that looks at the advertising budget and the target user’s profile, FB decides how much that particular ad slot is “worth” to your ad, and the ad that wants to be shown the most wins.

However, with FB you only pay for clicks, not impressions. An impression is when someone merely sees your ad, while a click is when they actually click on it (duh) and see whatever you linked to.

A standard FB ad (which you’ve seen hundreds of if you’ve ever been on FB) consists of a thumbnail-sized image with a few lines of text next to it. I originally was just going to use my book cover as the thumbnail, but it turned into a mushy gray blob when it was scaled down to the proper size. Instead, I cropped out just the title and used that. My finished ad looked like this:

The ad as it appeared in Facebook

Advertising performance

Here’s the raw numbers:

Campaign Reach:
43,535

Frequency:
2.3

Clicks:
21

Click-Through Rate:
0.021%

Spent:
$29.99

First thoughts were, “Wow, that’s a lot of reach,” followed by, “Wow, that’s not very many clicks.” In FB lingo, “Reach” is the number of distinct people who saw the ad, “Frequency” is the average number of times the ad was shown to each person that saw it, and “Clicks” is the actual number of clicks generated because of this. I was pleasantly surprised at how many people saw the ad, even with my very modest bid. If my campaign were the sort of thing where merely getting the word out and increasing awareness helped, I’d consider that a good sign.

But as you can see, the “Click-Through Rate” (CTR) was awful. Or was it? I really have no idea what a realistic CTR is, and whether my ad did better or worse than the average. What is clear is that you need a lot of views to get even a modest number of clicks–which makes since, since I very rarely click on internet advertising, and I don’t expect that a lot of other people do differently.

So I spent $30 to get 21 people to click on my ad, meaning that I spent $1.42 per click. This is not a very good rate. When I’m spending that much per click, my “conversion rate” (the number of clicks that turn into actual sales) would have to be close to 100% in order to break even. Which leads me to my next point.

Effect on sales

I had to wait until the middle of June to get my royalty statement for the month of May, which showed exactly how many books I sold. Since clicking the ad brought you to the main Lyrical Press site, I took the number of sales directly from Lyrical as an indication of the number of sales resulting from the ad. Lyrical’s ebooks are available in a large number of venues, but most likely anyone who bought the book as a result of the ad also bought it directly from Lyrical, since that’s where the ad pointed to. So how many books did I sell directly from the Lyrical site in May?

Zero.

(I did sell books through other storefronts, though.)

This is not terribly surprising, given the numbers above. Based on what I’ve heard on the nets, ad conversion rates vary from 1%-5%, so getting zero conversions from only 21 clicks is possible, even likely. So while this report is kind of glum, we have to accept it as the consequence of our overall ad performance.

Conclusions

The primary question that I wanted to answer by doing this experiment was whether a Facebook ad campaign could create an appreciable spike in sales of an e-book at a low price point. And the answer to that is a resounding no. My small initial outlay did not pay for itself in terms of increased royalties. It would not have paid for itself even if I were self-published and getting 100% of receipts. At the price level I set, there was no perceptible bump in sales at all.

Nonetheless, I consider the experiment a success because it did furnish a clear and unambiguous answer to the question I posed at the beginning. The benefit derived from a low-cost ad campaign doesn’t pay for the ad, and so makes no financial sense for a small-press author such as myself.

There are a few ways I could tweak the experiment which could give different (more profitable) results. Sometime in the future I may try to repeat this experiment varying the parameters below.

  • Spend more money. It’s possible that my ad budget was so low that I was outbid for all of the high-performing ad slots (in terms of page placement, appropriateness to the user, etc.), and got only the dregs. If this is the case, then spending, say, $100 or $1000 might get me exponentially more results. This would mean that there’s a minimum amount that you have to spend to get an appreciable result from advertising, and that spending any amount below that threshold is simply wasted. In support of this theory, I note that FB told me that there were about 6 million potential viewers of my ad, but I actually hit less than 1% of those.
  • Get a better ad. I’m pretty happy with what I came up with for my first try, but I’m sure there are ways I could make it more effective. With a little reading on ad design and a little practice, I could probably come up with something more enticing.
  • Don’t send people to Lyrical. I might get better results if I had the ad point to the book’s Amazon page, since most people will already have an account on Amazon, and anything that removes the barriers to a sale would be a benefit. Furthermore, Amazon owns the currently dominant e-reader, and streamlining sales to the Kindle would be a huge win.

Hopefully this was interesting and helpful to anybody else who is published with small presses or self-published, and who may be contemplating such a move in the future.

Advertisements

3 Comments

  1. I think I owe you another beer. While your results don’t surprise me at all, this is still valuable information. Thanks for posting about it.

    I am deeply cynical about advertising. Having run a small business for several years, I know that advertising really doesn’t work for anyone other than the ad company. They get their money no matter what. The business? Nada. I suspect, as you also concluded, that you have to spend a certain amount of money (and over a prolonged period of time), before advertising brings any results. This is just not possible for a small business to do. If you spent $1000 on ads, how many books would you have to sell to cover that? Is there any hope at all that you’d meet your costs, let alone make a profit?

    Over the five years I was in business, I spent about four thousand on advertising. That’s not a lot, I know. But you know where ALL my business came from, besides word of mouth? Google search, which was free, and the national association’s search engine.

    In other words, not one single person came to me because they saw one of my ads. I had ads in local magazines, newspapers, and also online with websites that fit my client demographic.

    I spent $270 a year for a webpage, until I migrated it over to Blogger, which cost me nothing. I had even more clients after that, leading me to conclude there’s no reason to pay for a website, either.

    I don’t know how we’re supposed to get the word out about our books. But I’m pretty sure that paying for advertising is not the answer.

  2. I think you’re right about advertising. The kind of advertising that works costs a lot of money, and is probably out of reach for the typical self-pubbed author.

    As an interesting counterpoint to my experience here, twice in the past 18 months Lyrical Press has bought full-page ads in Woman’s Day magazine, and then allowed authors to buy a slot for their book cover for $250. I’m not sure how many cover slots there were in the ad (12? 16?), but that adds up to at least $3000, plus I believe that Lyrical itself kicked in a portion of the ad cost. The results of this advert (which were published on Lyrical’s internal mailing list) were impressive: something like a 50% increase in sales across the board for the month after the ad came out, with significant sales increases for the months afterwards, suggesting that they earned a big chunk of repeat customers alongside the people who just bought once. I didn’t participate in this because my book isn’t a romance, and marketing a horror novella in that magazine would probably be a waste of money. However, I believe that the authors who did participate made back their investment, as did Lyrical Press itself.

    Also, spending $270/yr for a webpage is ridiculous. I have a webpage that I own which I pay $60/yr for (not this one, which is free), and I don’t see any reason to pay much more than that.

  3. Is your horror book anything like the video game “Thief 3:Deadly Shadows” ?

    I remember the haunted, abandoned foster home / insane asylum, where you saved the ghost of trapped children from the evil “Home”, and gathered clues as to who the Hag was (main evil character).

    If yes, I’d love to read that book. Is it more psychological (lol) or supernatural?

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s