Joshua Edward Smith | From the blog of Nicholas C. Rossis, author of science fiction, the Pearseus epic fantasy series and children's booksI’ve mentioned Joshua Smith’s (aka Alfageek’s) excellent blog in the past (in Alfageek Shares His Bookbub Ads Experience and The Benefits Of KDP Select: One Author’s Experience).

Well, the man has done it again and produced the best guide in setting up Twitter ads I’ve ever seen. He has detailed everything in six (!) consecutive posts aptly titled Step By Step Instructions for Promotion of your Book with Twitter Ads. I will be repeating here the gist of it in a single, easy-to-bookmark post with my experience added to Joshua’s, but be sure to check out the complete posts on his blog if you decide to go down the Twitter ads route.

How To Set Up Twitter Ads For Your Books

twitter tips | From the blog of Nicholas C. Rossis, author of science fiction, the Pearseus epic fantasy series and children's books

Photo by

Using Twitter Ads, Joshua has been able to sell an average of almost 3 copies of his first novel a day, day in and day out, every day. However, these offer him zero net profit on those sales. He does this because he hopes that people will then move on to the rest of his series, thus making him money on the sequel. If you feel this is a successful plan, then read on for a basic structure of his strategy:

  1. Create book or character accounts on Twitter
  2. Get your book ready
    1. Fix the first chapter
    2. Fix the back matter
    3. Fix the blurb
    4. Fix the price
    5. Have reviews
    6. Set up a link
  3. Run a test campaign
    1. Write five tweets
    2. Figure out your targeting
    3. Spend $10
    4. Sell nothing
    5. Pick your winner
  4. Run a real campaign
    1. Segment your markets
    2. Track results
    3. Compute your conversion rate
  5. Run a continuous campaign
    1. Tune your bid
    2. Test price elasticity

Sounds complicated, right? Thankfully, he has kindly broken down everything in a series of easy-to-follow instructions, starting with the creation of a book or character account on Twitter.

Step 1: Create book or character accounts on Twitter

I assume that most of you already have a Twitter account and/or have read Rayne Hall’s Twitter For Writers. The most important thing to remember here is to be the kind of person whose tweets are entertaining or personal, with the occasional bit of news or promotion of your books. Don’t be the nothing-but-promotion account with lots of links to Facebook posts and retweets of other author’s promotions and quotes from the book and hashtags and on and on and on. For further information on how to achieve this, refer to My Book Marketing Secret post, if you haven’t read it already.

Step 2: Get your book ready

By ready, Joshua refers to everything that will help conversions. While a campaign will get you leads, i.e. clicks and visits to your book’s page, it is up to your book page on Amazon to convert those leads to customers. That process is called “conversion” and the percentage of leads you manage to turn into customers is called your “conversion rate.” To improve that, you will need to be sure that everything, and I mean everything, encourages a visitor to click the coveted “buy now” button.

The first thing to do, which Joshua doesn’t actually mention, but which I’ve found out to be imperative, is to have a killer first chapter. When I first started advertising Pearseus on Facebook (see below), scores of people would visit the Amazon page but no one would actually buy the book. A friend suggested that my first chapter simply wasn’t enticing enough. So, I’ve rewritten it over a dozen of times since. Each time, the conversion rises.

As for the back matter, you may put there the first chapter of the next book, or simply a link to the next book plus a plea to leave a review.

The blurb needs to be as close to the ad as possible. You need a great one, but you also need to make sure that people who have clicked on the ad don’t feel like it’s a different book they’ve landed on. Check out my post, A Perfect Blurb, for some helpful tips on writing your blurb.

The price is another contentious issue. I have a bunch of posts on the subject, which you can find on my Marketing page. Joshua suggests you set it to $2.99 to begin with, and I personally agree with him.

Wait until you have a few reviews before you advertise. This will increase the chances of people buying your book. If you don’t know how to get reviews, check out my post, How to Score Great Amazon Reviews: Resources and More.

Finally, set up a link. This will automatically direct people to the right Amazon shop when they click on your link. This is the link you will use in your ad.

As you can tell, the whole point of Step 2 is to make the transition from seeing the ad to buying the book as smooth as possible. This is the golden rule; we don’t want anything to trip up the customer on their path to the buy now button.

Step 3: Run a test campaign

You’re finally ready to get to work with your campaign, starting with the tweets. These will be seen by people outside your follower circle. This is an important thing to remember, as these people will not have heard of you or your book. You’re about to make a first impression, and, as everyone knows, these count.

Start by writing five tweets. Ask yourself the following two questions:

  1. Can my plot (or even one of the subplots) be summed up in a loaded question?
  2. Would I click on this ad?

What I mean by “loaded question” is a question that will create an emotional response. One of the best examples I’ve come across was for a thriller that asked the question, “Would you kill your wife to save your daughter?”

For the purposes of this step, do not include a picture or any hashtags. Just some text and that link you made back in Step 2. The formula that’s worked best for Joshua is five stars (type star emoji if you know how to do that, or just copy & paste them from this tweet), followed by a sentence from a 5-star review. Joshua points out that sentences that make your book sound interesting/different work better than ones that say it is good. The five stars kind of already said it’s good. So you can use a sentence that captures attention.

Then, figure out your targeting.

Who will receive the tweet on their timeline? Make a list of accounts to target. Your tweets will be shown to the people who follow these accounts. Find authors who write books similar to yours. Find the publishers who sell those authors’ books. Dig deep. Try to come up with 30-40 accounts. Obviously, bigger is better. And they don’t even need to be authors you like. Just authors that are famous. If you can’t come up with that many, Twitter is going to suggest “similar” accounts when you start to enter these, so that might help.

It’s really important that you not target people who are unlikely to want to read your book. A big part of achieving conversion is talking to the right audience. But it’s also really important that you target a big group. Because if you don’t target very many people, you will not be able to get your ad seen at the price you can afford. Coming up with this list of accounts is the single most important thing you will do.

Once you have that ready, spend $10:

  • Go to and give them your credit card number.
  • Create a campaign: “Website clicks or conversions”
  • Name it “Test” and set it to “Run continuously”
  • Do not check the box for Twitter Audience Platform. See this post to understand why.
  • Set the locations to: United States, Canada, UK.
  • Gender: Any gender (Twitter has no idea what gender people are)
  • If you are sticking with US, CA, UK, then there is no need to set language.
  • Add followers, and put in the accounts you listed. After you enter each one, Twitter will suggest some. Write those down. They might be good additions to your list.
  • Do not do any demographic targetingNone. Here’s the problem: Twitter has no idea what demographics anyone is in. It doesn’t know gender or interests or any of that. If you touch any of the demographic controls, you will ruin all that careful targeting of accounts you just did. Don’t do it!
  • Only put tweets in people’s timelines, because those other places are annoying.
  • Set a daily maximum and a total budget of $10.
  • Set your bid to $0.09. Yes, Twitter is going to yell at you and say it should be 16 times that big. Twitter lies. $0.09 will get you plenty of impressions.
  • Select the 5 tweets that you made earlier.
  • Launch your campaign.

After all this, you will sell nothing.

You won’t get any impressions or clicks for a little while. Don’t worry about that. Eventually you’ll start to get both. You might not sell anything during this test. But that’s okay. Our objective here is only to find the best tweet to use going forward.

You will be lured by the analytics interface to look at demographics and which accounts worked best and other random things. Don’t fall for it. Twitter has no idea about any of that. Those numbers are all junk. All you care about right now is which tweet to use, and all you’ll care about at the end is how many books you sold.

At $0.09 per click, your $10 campaign will generate 111 clicks. If you want, you can lower your bid and you will get more clicks, but it will take longer. If you don’t get any clicks after a day, you probably aren’t targeting enough users. If you get a lot of clicks in the first few hours, you should lower your bid to slow things down. You can adjust the accounts you target and your bid and anything else you want while this test is running.

Now, you’re ready to pick your winner.

By the end of the test, you’ll see that each of your tweets has a “Click Rate” associated with it. One of them will have the best click rate, and that’s the winner. Having a tweet that draws clicks well is a good way to ensure you get the maximum exposure, fastest, for the lowest cost. A tweet with a better click rate is better for you economically.

Step 4: Run a real campaign

Change the locations to just US and Canada. Scroll down to the bottom and set your budget to $10/day and $10/total again, and pick your winning tweet to promote. You can add an image if you wish, and test which tweet works best for you – with or without an image. Note that Twitter suggests you promote three, but it’s fine to just do one. And it eliminates a variable, which is helpful. Here is Joshua’s winning tweet:

Launch that campaign. Now copy that campaign, and change the locations to just the UK. You’ll need to set the tweet again. Launch that.

If you are doing any other countries, make separate campaigns for them.

This is called “segmenting your markets” and there are two reasons to do it. First, the ad markets vary a lot by country. So in some countries you will be able to get impressions with a lower bid than will yield impressions in other countries. You only get to set one bid per campaign, so this alone is a good reason to keep them separate. Second, buying behavior varies by country, so the break-even bid is different for that market.

Track results

Twitter will report clicks in pretty much real-time. Don’t look at any of the other numbers because they are irrelevant and will confuse you. Plus they are mostly wrong. All you care about in the Twitter Ads interface is clicks.

Over in your KDP dashboard, you can see sales. Since you segmented by market you can look at the sales in each market separately. You can use the excellent – and free – Book Report to track your sales, but keep in mind that royalty report has some time lag in it, which can be confusing.

It is okay to tinker with your bid now and then. Raise it a little if you are not getting impressions in that market. Lower it if you are getting a lot of them. Ideally, a $10 campaign should last a few days, so you get a good cross-section of data.

Compute your conversion rate

  1. Number of copies sold: _____
  2. Number of clicks: _____
  3. 100 times line 1 divided by line 2: ____%

Your conversion rate will vary a lot at the beginning, but you’ll find it stabilizes to some number, usually around 5%.

We are going to use this conversion rate to compute our maximum bid going forward.

Step 5: Run a continuous campaign

You can just add more money to those existing campaigns and they will start up again. Joshua’s approach is to feed each campaign segment in $10 increments, and continuously monitor the clicks and conversions.

There is some evidence that Twitter will stop giving impressions to tweets that are old. So, take your money maker tweet and copy it exactly into a new tweet every two weeks. Then swap that into the campaign.

Tune your bid

You can tune your bid constantly. At low bids like $0.10 or less, you will see that impressions tend to come in bursts. One day Joshua had 500 impressions in the US market all day, and then he suddenly got 10,000 in the hour starting at 10pm. It really doesn’t matter. Whenever you get impressions, you’ll get your clicks, and then you’ll convert some of those leads to customers. If you bid higher, you’ll spend more and speed up the process. If you bid lower, you’ll save a little money and slow things down.

You need to set a cap, though. At $2.99, your royalty is about $2. Your bid should not ever be higher than $2 times your conversion rate. If your conversion rate is 5%, you should not bid higher than 0.05 x 2 = $0.10. You can bid lower if you want. Your campaign will go slower, but you’ll do better than break-even. You should not bid higher. That’s called paying people to read your book, and it’s a rookie mistake.

Also, feel free to add more accounts to the list if you find one that you think has good prospects following it. Don’t forget to add them to every campaign geography segment.

Test price elasticity

You don’t actually have to do this; you can simply trust your gut or the results below, and continue with the book priced at $2.99. But, if you want to test other prices, run your ad campaign for at least two weeks before going to this step. You want to make sure that you’ve tuned your bid and have a consistent flow of clicks going to your page. Pause your existing campaigns, and copy them.

Increase your price. Joshua’s first experiment was to go from $2.99 to $3.99. You can leave everything identical (you’ll have to select the tweet again). Launch the copy. The reason for doing the copy is that it’ll be easier to count clicks. We expect that our conversion rate will go down at a higher price, which is bad. But we make a lot more money per book, so it might be good. There’s only one way to know: test it.

Book price elasticity | From the blog of Nicholas C. Rossis, author of science fiction, the Pearseus epic fantasy series and children's books

Graph by Alfageek

Joshua has done just that and has produced the graph above. To be profitable, one should set the price to the place where the cross is the highest. Two things emerge from it:

  1. Higher prices yield fewer sales. Yes, it seems obvious, but here it is in black and white. And blue and green dots and crosses.
  2. The highest profit is to be made at the lowest price.

In other words, the increase in royalties does not compensate for the loss of buyers at a higher price. Note that Amazon has been trying to tell authors this forever. They say very clearly on the KDP setup page that the biggest earnings will come if you set your price to $2.99. Hugh Howey and The Data Guy also agree in their report.

Still, if you wish to test other prices, who am I to stop you? Just be sure to share your experience in the comments!