Adventures in Twitter Advertising

The Money Maker

The Money Maker

When last we met our hero, he had just stumbled onto Twitter Ads as a way to sell books. For context, I sold almost no books in the first three months of the year. I ignored this sad fact by focusing on the sequel I was launching. My rough plan was: 1) assume I’d get a lot of sequel sales with very little effort, because most of the people who read Entropy told me they really wanted a sequel, and 2) start trying to find ways to get new people to read Entropy after the launch (for free, if necessary), in order to build a base of readers who would buy the sequel Duality at full price.

I’m not a big fan of the “free first book” model, because I’ve never seen any hard evidence that it actually works. And I know from my personal behavior that books I get for free mostly sit unread. I read books I’ve paid for, and I read books where the author has personally asked for feedback. I never get around to reading free books. And my gut says I’m not alone. So even if I don’t make any money getting a book into the reader’s hands, I’d rather that the reader paid for that book. I believe they are much more likely to read it, and hence will buy my sequel. I have no evidence to back this up. I’m probably completely wrong. But that’s what my gut says.

After trying a lot of promotional tools, I stumbled into Twitter Ads. I LOVE TWITTER ADS. I’ve been using them for about two weeks, and one thing that’s clear is that you really need to use them a while to separate the signal from the noise. Some of my early conclusions have been debunked by later data. And there’s a chance some of the things I’m pretty sure about now will also prove false. But the following are my best guesses, based on the data I have so far.

Test a bunch of different tweets and go with the one that performs best

You can give Twitter a whole pile of tweets to promote, and then sit back and watch the click-through performance of each. The one I showed up top out-performed every other tweet I tried by a dramatic margin. So I’m only using that tweet in all the campaigns I’m running. Since you only pay-per-click, using a high-performing tweet doesn’t save you any money. But it does save time. Your bids will garner a fixed number of impressions, so the higher the percentage of people who click through, the faster you’ll get your buyers.

When you set up your campaign, use only Country and Followers for targeting

Twitter ads let you do all sorts of demographic targeting. Which seems cool, until you realize Twitter has no idea what the demographics of a user are. They don’t even know your gender, how could they possibly know your income or what TV shows  you like? However, by targeting people who follow authors and publishers of books in your genre, you are guaranteeing you’ll hit the right audience. I’m targeting people who follow these accounts:

Who's Who in Romance

Who’s Who in Romance

The list of authors and publishers you want to target are probably different. But it’s relatively easy to build up a list. Start with the ones you know. Twitter will suggest similar ones, which tend to be right on. I also googled for “Romance authors to follow on Twitter” and got a bunch more. The more the better, because targeting a larger group will lower the bid at which you will get impressions.

Run one campaign for US/Canada, and another for UK

The US/Canada Twitter ad market is totally different than the UK market. The bid at which you get impressions differs greatly. But in addition to that, the buying behavior of people in the two markets is different. So if you separate them into two campaigns, it will be a lot easier to tune your bidding. Note that Twitter Ads have been completely ineffective for me in any other market where people can buy Kindle books. Even Australia, where I have a lot of readers, didn’t produce a single sale despite 51 clicks. So in my experience, US/Canada and UK are the only markets to bother advertising in.

$0.99 sales do improve conversion rates

In a Twitter Ad campaign, you pay roughly your bid price every time someone clicks the link to look at your book. Some percentage of these people will go on to buy your book, which you can track using the KDP dashboard. Since I had no sales before I started advertising, I’m generously assuming that every sale is due to these ad clicks. That’s probably not exactly right, since I’m continuing to do social marketing on Twitter and GoodReads. But it’s almost right. Since sales only last a few days, and I’ve found a lot of day-to-day variation, these conversion numbers might not hold up if I ran a lot more tests. But here’s what I found:

US/Canada Market: $0.99 conversion rate: 18.2%
US/Canada Market: $2.99 conversion rate: 4.7%

UK Market: £0.99 conversion rate: 9.8%
UK Market: £1.73 conversion rate: 6.8%

The higher price conversion rates I’m seeing are holding up pretty well over time. Those sale rates might be inflated because I was doing a lot of social media marketing about my sequel launch at the time the sales prices were in effect. It’s clear that running a sale boosts conversion rates, though. However, since Amazon slashes your royalty if you drop below the $2.99 rate, the numbers clearly indicate that for me lowering my everyday price to $0.99 would be a loser. I would not increase volume nearly enough to compensate for the 9x drop I’d see in my royalties. (When you factor in the $0.07 that Amazon charges to deliver my book, my net 70% royalty at $2.99 is $2.04, but my net 30% royalty at $0.99 would be only $0.22.)

Bid to break even on sales

My original thought was that since I expect a high percentage of those who read the first book to go on to buy the sequel, I should bid based on the expected royalty for both of those sales. However, it turns out I don’t need to do that. If I take my royalty and multiply by my conversion rate, I get a break-even bid per click of $0.14 in the UK, and $0.10 in the US/Canada. That means that if I run a $10 campaign with that bid, I’ll make about $10 in royalties. This strategy is yielding no profit, of course. But I’m building a base of people who I think are actually going to read my book. Since, you know, they paid for it! So if they go on to buy the sequel, that’s where the profit comes in. And if they don’t, I haven’t lost any money.

Actual Sales Due to Twitter Ads

Actual Sales Due to Twitter Ads

The first three days on the left were during my $0.99 sale. Then I stopped getting any clicks for a bit, while I tested whether really low bids would get impressions (they don’t). And then I started tuning my bidding based on that break-even calculation I just explained.

Next Steps

The next thing I want to do is test what increasing my price does to my conversion rate. But I can’t do that for a few more days because Amazon doesn’t let you change your price for two weeks after you run a $0.99 sale. Once that blackout period is over, I’m going to boost my price by a full dollar, and run at least a week of ads at my current bid price. Depending on the results of that, I might try increasing or decreasing and doing another test. Increasing the price should lower sales volume, but increase profit per sale. So the question is exactly what price yields the highest profit. And the only way to find it is by testing. But since I now have a steady-state way of getting buyers (there are millions of people in my target audience, so there is no chance of market saturation), I should be able to find that given enough time.

I’ll keep you posted.


4 thoughts on “Adventures in Twitter Advertising

  1. Thanks for more on this topic! I made my first visit to the Twitter ad pages and I must say, they aren’t welcoming at first glance. I appreciate your advice here about whom to target, since making that decision was a first hurdle. This gives me direction.
    Do you know of a clear, step-by-step guide for learning the basics of this process? Or is there a specific section of the Twitter instructions that makes it transparent? I admit I didn’t spend as much time there as I wanted to, but any guidance you feel like giving will be a joy to me.
    Thanks again!

Leave a Reply

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

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

Google photo

You are commenting using your Google 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 )

Connecting to %s