So, I’ve had a bit of a break since I got back from Las Vegas. With an eight hour time difference, it turned out the jet lag was much worse than I anticipated.
But I’m back now, I’ve got a new novel out, and I’m using it to try out a blurb testing technique I learned from Bryan Cohen at 20 Books Vegas.
(Note: I do not claim to have come up with this system myself. It’s all Bryan. What I hope to do is to be able to show you how it works in practice so you can learn from what works for me, and, importantly, what doesn’t.)
Bryan is the author of How to Write a Sizzling Synopsis, and runs a company that specializes in writing blurbs. So he knows this stuff inside out.
I’ve written before about the process of writing blurbs and, to a lesser extent, testing them. I learned a great technique from Orna Ross of the Alliance of Independent Authors which involves identifying hot words that are used in your genre. Suddenly the process of writing blurbs, which I’d always hated, became fun. Yes, fun!
But I still have plenty of questions about blurbs, and you might do to. Questions like:
- How do you know if a blurb is performing or not?
- How can you tell whether it’s the blurb affecting sales, or the cover?
- How many clicks to your book page do you need before the test becomes statistically significant?
- If the blurb isn’t performing, what do you change about it? And how do you test that?
Bryan’s talk answered a lot of those questions.
Let’s take a look at them one by one.
How do you know if a blurb is performing or not?
If you’re getting around 1 sale to every 10 clicks, that’s healthy. If you can get more, fantastic. But if you convert at a rate of less than 1:10, your blurb is underperforming.
How can you tell whether it’s the blurb affecting sales, or the cover?
This is a tricky one. Before you start testing blurbs you should ensure your cover is good quality and (most importantly) on-genre.
I found myself in the middle of a twitter spat about this topic this week. Someone read my post on what your book cover should do and disagreed with the idea that book covers should be on-genre. He sent me a list of books which had off-genre covers but nonetheless sold well.
All of these books were classics, from traditional publishers. If there’s one piece of advice I’ve received that I constantly pay attention to it’s this:
Ignore what the traditional publishers are doing when you market your independently published books.
This isn’t because traditional publishers don’t know how to market books. But it’s because they operate in a very different market: one where print still outsells digital, and a book can sell massively because of publicity, not because of the cover, or the blurb, or even the contents.
It’s much more useful to follow what other Indies do.
Anyway, this post isn’t about covers, or about tradpub. But how do you know if your cover is good enough?
The answer is in Amazon ads. For many Amazon ads, you can’t add any text: all people will see is the cover and the number of reviews. If your ads get a healthy number of clicks, then your cover is doing its job.
And if your also-boughts are in the same genre as your book, that gives you a clue that your book looks like other books in your genre. Which is good (regardless of what people on Twitter tell me).
How many clicks to your book page do you need to test before the test becomes statistically significant?
Bryan’s answer to this was simple: 100. When you have 100 clicks, you should have 10 sales. If you have less than that, you need to change something about your blurb.
Which leads me to…
If the blurb isn’t performing, what do you change about it? And how do you test that?
I’ve always struggled with this one. What if it’s just part of the blurb that’s a problem? What if it’s the whole thing? How can I possibly know?
Bryan’s advice was to change just one thing: the hook. That’s the opening line, which should stand alone and hook people into buying the book – or at least reading more of the blurb.
Change that, and test again.
So let’s look at Bryan’s technique for testing your blurbs.
The 4 Step Blurb Testing Process
OK, here goes.
Step 1: Get 100 ad clicks to your book page
Bryan recommended using Amazon ads to do this because it’s the easiest.
He suggested using yasiv to generate keywords for your ads so that you can run enough ads to get 100 clicks.
Gather 10-50 sets of keywords and bid 20-40 cents on each of those keyword sets. That’s a higher bid than you might want in the long run but for testing purposes, you want to get those clicks.
If you struggle to get your Amazon ads to deliver (and who doesn’t?), you might want to use Facebook ads instead, which is what I’ve been doing. But only if you’re comfortable creating FB ads: it is significantly more work than Amazon ads.
If you really want to make life simple, run some automated product display ads on Amazon, that don’t even require you to add keywords: Amazon generates them for you. In fact, I often find that Amazon is better at generating keywords for my books than I am.
Step 2: Analyse your conversion rate
Ignore most of the data Amazon gives you in the advertising dashboard: things like ACoS (average cost of sale), CPC, sales etc.
As any author who’s worked with Amazon ads will tell you, that data is unreliable anyway.
You just need one figure from your ads data: the number of clicks. So grab that from your ads dashboard. It should be more than 100.
Then go to your KDP dashboard and find the Sales screen. Look at the period during which you were running those ads to that book. Note how many sales you made.
You now have two figures:
- Number of clicks
- Number of sales
Here’s an important caveat: if you’re running a promotion on your book (a discount, Bookbub featured deal, anything like that), don’t test during that time. It will skew results.
Now divide the number of clicks by the number of sales.
- If the figure is 10 or less, your blurb is performing. Run more ads to that book, constantly keeping an eye on the conversion rate. Try to get your cost per click down, so your profit margin improves. And move on to testing the next blurb.
- If the figure is more than 10 (i.e. it takes more than 10 clicks to get a sale), you need to review your blurb. Advance to step 3.
Step 3: Tweak the blurb
So your ads aren’t converting well enough.
Assuming your cover is good (if it isn’t, you shouldn’t be testing blurbs), you need to tweak your blurb.
Tweak just one line: the hook.
If your blurb doesn’t have a hook, write one.
Step 4: Rinse and Repeat
Once you have your new hook, repeat this process.
Keep testing until your book converts at a healthy rate.
But what if it doesn’t?
If after five iterations of the hook your book page still isn’t converting at 1:10 or better, you probably have a problem with the entire blurb.
Maybe your blurb is really a synopsis.
Maybe you’re adding too much detail.
Maybe you aren’t introducing your characters powerfully enough.
If this is the case, see my post on writing blurbs for some tips on starting from scratch.
Groups like this will critique your blurb for you. They’ll often rewrite it completely. If you’re stuck, it will be cheaper than hiring someone.
If you really get stuck, you can hire someone to write blurbs for you. But I always prefer to write and test my own. This is for three reasons:
- The process of testing will make me better at blurb writing, a skill which is an investment in my future career.
- I can spend the budget I would have spent on a blurb writer on testing with ads.
- I’ll make some sales in the process, even if only a few.
So hopefully that’s encouraged you to test your blurbs, and given you a process you can follow. I’m already testing it with my new Mystery book. And I’m finding three very interesting things:
- The ads are getting massive engagement, like nothing I’ve seen before (the books are funny and feature a quantum cat, which helps). Likes, laughs and shares: it’s wonderful.
- In the UK, they’re converting at 1:7, which is fantastic.
- In the US, they’re converting at 1:20, which is dreadful.
Sales are the important thing, not ad engagement. You don’t forge an author career on Facebook likes.
And the US is the benchmark for testing ads. It’s the hardest market to sell to, because of the high levels of competition. So it’s the market in which you should test (and in which you should get 1:10 conversions). And that means two things:
- My blurb isn’t good enough.
- My blurb doesn’t appeal to Americans.
So this week I’ll be tweaking the hook and testing again. Hopefully the next time you hear from me, this will have translated into sales.
Watch this space!