Amazon Giveaway Follow-Up
Last week I ran an Amazon giveaway and reported on my initial results. Here’s a bit more of an in-depth analysis a week later. First, a brief look at what happened overnight:
Within an hour and a half I’d gained 100 Twitter followers. To put that in perspective, I gained 147 followers throughout the entire month of May. As of this morning when I write this, I’ve increased from 436 Twitter followers to 631, a total increase of 195 followers or a 44% jump. The contest is not yet over as of this writing – there’s still one more eBook to give away.
Book sales are up – and by more than the units that I paid for. This experiment happened as part of the book launch, so I can’t tell for sure that it’s directly related to this giveaway. But book sales are actually up as compared to the previous day.
Good results, right? Yes – and I’m definitely glad I did it. However, the rest of the week didn’t live up to that first rush.
I scheduled the giveaway for 3 books, with a 1-in-100 chance of winning. That should have netted between 200 and 300 followers – and it did. My follower count went from the 436 followers noted above to a peak of 702 followers over the weekend. However…
The first two books went quickly, with 195 followers literally overnight. That was great. But it took another five days to finish the giveaway and award the third book, and also to hit that peak.
What does this tell us? Does that mean this is a one time deal? I don’t think so. I think it tells us two things.
First, a large portion of that gain in followers likely came from people who already followed me retweeting it. In other words, it was friends of friends. This is good and bad. Many of those followers are likely to stay, which is good. On the other hand, that resource is probably tapped out for the near future – until I gain significantly more new followers, or manage to convert a large portion of this batch of new followers into actual “fans.”
Second, as small as it was this singular giveaway probably saturated the market. That means that running one every day or even every week is going to hit diminishing returns very quickly. My guess is that the sweet spot will be running one once a month or even once a quarter. I think that once a month might work OK if you’re giving away multiple products and can cycle through them. Once a quarter would work a lot better if you only have one product. Even with multiple products, once a month might be too often. I plan to experiment further to nail this down. Either way, I don’t expect to pull of overnight 44% growth again. I believe that getting that kind of growth was largely a function of having such a low follower count to begin with.
So far I’ve maintained an overwhelming majority of those followers – but they have started to trickle off. I’m down into the low 690s now, so I’ve lost about 10 followers of the 266 I gained during this event (about 4% of them). I expect to drop more over the rest of the month, although it might be hard to tell as they trickle away and I continue to gain new followers on a day-to-day basis. I expected to lose many of these followers, and 4% after one week isn’t bad at all. It’s actually far better retention than I’d expected.
The units that I paid for did count as actual sales. They showed up on my KDP sales dashboard, and they did effect sales rank on Amazon. However, this effect was diminished due to the length of time it took for the contest to end. All three sales showed up at once at the very end of the giveaway. Had they shown up at the beginning – when sales were already good from our launch – they might have propelled the book into the top 20 for its category. As it was, they didn’t even push the book back into the top 100.
Given this, and the way Amazon’s sales rank algorithm works, I think the best use of this kind of giveaway is during a product launch. Experiment beforehand and get a good idea of the right way to setup your giveaway to ensure that it ends early and the sales show up in your launch ranking, helping to propel you to a better slot as part of your launch. But realize that if you try to game the system by buying a particularly good sales rank, you’ll probably have to buy far more copies than you’ll sell as a result. In other words, the value of this is likely to be limited.
At the same time, you’ll want to make sure that you have a good number of reviews up early. If we’d had more reviews up when the giveaway happened, I think we’d have had far more success with it.
The final verdict is that this is an immensely useful tool when used properly. But don’t let my click-bait headline from the last post fool you – it’s not going to cure all of your sales woes, at least not on its own.