For any new author, trying to establish an audience is one of the most difficult things we face. Before the age of self-publishing through Amazon, B&N et al., if you weren’t picked up by a publisher then getting your work in front of people was a very difficult and time-consuming process. Nowadays, anyone can write a novel and have it visible to the public in moments.
The downside of that, of course, is a massive increase in competition. Instead of a shop with hundreds of books vying for readers’ attention, there are hundreds of thousands on Amazon. It’s a bookstore of near-infinite shelves, and if you’re not at the front table, it’s a journey you have to persuade your customers to make.

tf2-warehouse-raiders-of-the-lost-ark
So how to make sure that your book stands out and gets people interested?
One way is to advertise. Popular ways of doing this online are Amazon KDP select advertising, Facebook ads, discount / free offer mailing lists and Twitter streams.
Ask other indi authors on Facebook or internet forums which is best and you’ll get lots of different responses. Unfortunately,  advice from one person will usually be in direct conflict with advice from another. This is to be expected; what works for one person or genre may not work for someone else. It’s trial and error at the end of the day, to find what works for you.
This all costs money and often the expense unfortunately far outstrips any return. These adverts should be considered long term investments, but even so it can feel like flushing money away into a black hole most of the time.

“She watch channel Twitter”

hoxfgjbjvd
There are many companies offering to advertise your book on their Twitter streams, and costs vary from about $14.99 a month to a lot more. They often boast a huge reader base and the opportunity of selling hundreds of books, usually with testimonials to match. Nothing wrong with that, as long as they are genuine; they are a business after all trying to get your custom.

Not all Twitter-based companies and their channels are created equal, though. Here are some questions to ask and things to consider before choosing one.

  1. Do they publish their overall hit rate metrics? How good are they at driving traffic to pages?
  2. If they do give a figure on their site, it’s often an average rather than the median. It only takes a few really big successes (possibly from authors much further on in their career than you) to skew the numbers beyond manageable expectations.
  3. It’s in their interest to get as many customers as possible (again, nothing wrong with that). The more books they show, however, the quicker your tweet will fall off the bottom of the page. Humans are notoriously lazy and the odds of them stumbling across your book goes down the further they have to scroll. Therefore check how many times a day your tweet gets sent out and at what times. If they have several channels and cross-post, that’s great. But if the tweets all go out at once at 2am then it’s not ideal.
  4. You have no idea how many of their user base actually look at their streams and how often, even if they are ‘real’ Twitter users. I’m a follower of all of the lists I’ve used in the past and rarely look at any of them. I’m part of the problem!
  5. Some companies will write your tweet for you from the book description. This can result in some hilariously out-of-context tweets if you don’t check.
  6. Check if you will get a summary of hits at the end of the advertisement. If you don’t and you don’t sell, you’re still in the dark where the problem may lie.
  7. Is the readership on the stream the correct one for your book? Your sci-fi vampire tale based on an alternate Earth where the Elizabethan period never ended (hey, that sounds good!) may not be the best fit for people looking for a contemporary character romance or Bourne-type action thriller.
It’s worth noting that the best any promotion company can do is get traffic to your Amazon page. After that, it’s up to your blurb, cover and reviews to make the sale.
If people come to your page based on a recommendation then you’re halfway there already. If they’re there on a whim and a click, then it has to be top-notch.

My initial experience with Twitter

I tried a few Twitter companies when I first released In Vitro Lottery with pretty much no success at all. At the time, however, I had little to no reviews and no buzz, so in hindsight that wasn’t totally unexpected.

Now, though, my novel has over twenty reviews on both Amazon US and UK sites, averaging 4 and 4.5 respectively which isn’t bad. It also has 38 ratings on Goodreads. So I decided to give it another try.

One of the companies I currently use is BooksGoSocial (aka YourBookPromoter.com), who offer a variety of packages and tweet duration. Although the tweets run for a week or more, the actual membership lasts a year and includes lots of inclusive extras that many companies charge a premium for. BGS is run by Laurence O’Bryan, who is himself an author in addition to running the business, and he has some great insight on how the traditional and indi publishing world works.

For this experiment I signed up with BooksGoSocial’s ‘Reader Insight Service’ add-on, which is a bit different from their standard offerings. This package would include a month of tweets across all their networks of over 700k users, an initial Skype support call of what things to try, and email support. They also designed a new cover for my book (cover 2, below) and some graphics for the tweets.

The experiment

291483

As a starting baseline, in a month where I’m not promoting it, my book will sell a few copies and get maybe a couple of hundred page reads through Kindle Unlimited. Not exactly retirement-fund worthy, but it is what it is.

The aim of the experiment, therefore, was two answer two questions:
  1. Can general reading Twitter streams get people to my Amazon page?
  2. Do the cover and blurb make a difference to any resulting sales numbers?
For this promotion, I had 4 x 1 week blocks of tweets, so we tried the following combinations, focusing on the story and the emotions behind the story, and two covers. Due to a slight communication error (my fault) the plot tweets were actually a mix of two different ones.
Emotion-centric tweet, emotional-blurb and cover 1
Emotion-centric tweet, emotional-blurb and cover 2

Plot-centric tweet, plot-blurb and cover 2
Plot-centric tweet, plot-blurb and cover 1

Cover 1
Cover 2

The tweets were linked to a Bit.ly page which collected data, and sent it to a Google Analytics page for analysis.

To make sure money was not a factor, for the period of the promotion I set the price to $0.99 / 99p

The results.

I’ve split the results up into click-throughs to my Amazon page and sales. I’ve also included Kindle Unlimited page reads, although these are difficult to interpret as the book could have been downloaded at any time. Due to the lag between Amazon’s blurb and cover changing systems the boundaries aren’t perfect for the different covers, but it gives a good enough idea.

Hits according to bit.ly Plot Emotion
Cover 1 49 (25) 49 (25)
Cover 2 52 (26) 60 (30)
Total 101 (51) 109 (55)
Sales Plot Emotion
Cover 1 3 1
Cover 2 4 2
Total 7 3
Page reads Plot Emotion
Cover 1 529 196
Cover 2 238 20

So in all, my results weren’t much of an increase over my baseline, which obviously for me was very disappointing.  Still, as my boss often says; “the numbers are the numbers”, so instead of going off on some self-entitled Facebook rant about the unfairness of life, let’s look at what data there is more closely.

Click rate:

Considering that about half of those clicks are probably from robots (based on reading around the internet and that the goo.gl link wrappers showed about 50% non-bit.ly traffic) the adjusted numbers are shown in parentheses.

Over the month I got about 25-30 clicks per week (adjusted), which considering the number of followers on the various streams, is very low. The nature of the tweets or the two graphics used made no real difference to the click rate.

Click to purchase ratio:

This is where it gets a bit more interesting.  For the plot blurb, that’s a click to buy ratio of 14% overall, compared to 5.5% for the emotional.  The two covers didn’t make much of a difference to the sales.
That said, the numbers overall are low (or as we say at work, ‘a very small n’), so it would only take a few sales or hits in either direction to make a big difference.

Still, based on the result, my blurb will remain focused on the story from now on.

So, what went wrong?

face-palm-life-homer-simpson-facepalm-fail-demotivational-poster-1277576990
“Well, maybe your book just sucks and no one wants to read it.”
Maybe that’s true! It may be that my particular blend of Orwellian dystopia, genetic engineering and scientific ethics in the face of annihilation isn’t for everyone. And that’s fine. To counter this, I’ll offer my experience with a different promotional company, Book Barbarian.

Book Barbarian is an email mailing list company focusing purely on science fiction and fantasy (Disclosure – I have a testimonial on their website). They only advertise discounted and free books, and I used them last time I had a free promotion for In Vitro Lottery back in July. Although the email only goes out once (plus a mention that day on their Twitter stream and Facebook groups), I had my book for free for two days to pick up any stragglers and then dropped the price to $0.99 for a few days after that to entice anyone who read the email late, but was still interested.

Tip – with any promotion company, if there’s an option of a premium package to put your book on the front page and top of the list, take it. At least then you’ll know customers can see it.

For the day until the month after my promotion with BookBarbarian I got:
Free book downloads: 1237
Sales: 24 (most of these happened a few days after the promotion)
Page reads: 11,773 (an increase of about 1000% over baseline)

Which, for one email letter and a couple of tweets, is pretty good!

This isn’t an article trying to pitch one company against another, though. They both offer very different services, after all. But it is an example that different approaches can give completely different results, which is why it is vital to cast your advertising net over as many different things as possible when testing.

Conclusions

So what can I conclude from this experiment?
General readership Twitter streams currently isn’t the best choice for me and probably won’t be unless my book really takes off. If I write a more general-fiction style book in the future, I’ll probably re-visit them.

Do I regret using BooksGoSocial? Not at all. I’ve been a customer of theirs for about six months now, and even though I made very few sales on both attempts through their Twitter streams, I’ve found the free extras they provide as part of the service (mailing lists, review clubs, Facebook support groups etc) more than make up for it. It’s just a really nice community to be a part of and I’ve learnt a lot from the experience. Laurence and co have always been very helpful and quick to respond to emails throughout the experiment, even at weekends and evenings.
I just wish I had a better result for them really!

I do think, however it would be a fantastic idea for BooksGoSocial (and other Twitter companies) to offer genre-specific streams with a member base who are interested in that area. I’d much rather pitch to a few thousand people who really like sci-fi, than hundreds of thousands who have no interest in my genre. As my experience with BookBarbarian shows, it’s a fertile ground for a great response.

Ed Ryder is a research scientist by day and author of In Vitro Lottery, which is available now from Amazon.