I’m sure you remember my step-by-step tutorial on how to advertise with Amazon. Three weeks later, I have some preliminary results, which, as promised, I’m sharing with you, so that you can learn from my mistakes.
In a nutshell: I’ve failed. The experiment has so far been a disappointment, albeit one offering fascinating insights.
You may remember that I had done my math and knew that, to make any profit, I had to bid under the proposed $0.50.
Now, for a peek behind the scenes: whenever Amazon is about to show a product page, a small bidding war will rage for a few nanoseconds. Amazon will check to see if that product is linked to any ads. If more than one ads want to be displayed, they will bid for that space, until they reach their cut-off threshold. So, if I have specified 5c as my maximum, then an ad bidding for 6c will be displayed instead, and its owner will be charged that amount.
I had assumed a conversion rate (how many of the people who see my ad will click on it, and how many of them would actually buy the book) around 1%. This was based on my experience with Google ads, which I run for our Istomedia clients. The percentage of people clicking on one of our Google ads is usually between 3 and 7%.
If that were true in this case, each sale would require some 100 clicks, and each click would require another 100 impressions (views). So, you’d basically need around 10,000 impressions before you make a sale.
Conclusion #1: Target by product
I started my CPC bid, as this kind of bidding is known, low. Very low: at 3c, well under the suggested 5c. You may remember that I had run two campaigns. I had targeted one by product, and the other by interest. I immediately noticed that targeting by product was far more efficient. So much so, that a week later I terminated the second campaign and started a new one, targeted by product this time.
Conclusion #2: Bid high
This was my first failure. The second was that, even when targeting by product, too few impressions were made. I started increasing the bids, until I was over the 10c mark – twice as much as Amazon’s initial suggestion. I knew I’d make no money this way, as the ads would eat up any profits from my sales, but figured it was worth a shot.
This finally got me some traction, but I still had to raise the bid even higher to get any real kind of action.
Conclusion #3: Price matters
I was simultaneously running a campaign for a friend, who’s a romance author. Her campaign was about three times more successful than mine! She needed a 5c CPC bid to have the same results that I got with my 15c ones.
There were two differences between the campaigns: one was our genre. However, romance is one of the hottest genres, outselling fantasy, therefore it should have required higher bids, not the the other way around.
The second difference was that her book was on a 99c sale, whereas mine were full-price. Therefore I concluded that your book’s price matters a great deal. Everyone loves a good deal, and it’s practically pointless to advertise about a full-priced product (unless you want to raise awareness instead of making sales – more on that later).
Conclusion #4: Conversion rates are low
The final conclusion was that my initial estimate of a conversion rate of 1% was way too high. Indeed, it turns out the real figure was a mere 0.14%. In other words, it takes some 700 views before someone clicks on your ad, instead of a hundred!
I have no real data on how many people need to click before you actually make a sale, but I suspect it will be far lower than my initial estimate of 1%. If it were 1%, you’d need some 70,000 people to look at your ad before you made a sale. If, however, that, too, is 0,14%, then it may take up to half a million eyeballs before you even sell a book! Even with a low CPC bid of 5c, you will have paid $35 to make that sale. If said book sells for $0.99, then you will make 33c for every $35 spent. Include taxes, and you may end up with some 20c. Not exactly a stellar return on investment…
So, is it a bust?
In a nutshell, yes – as far as sales are concerned. A few months ago, I stopped advertising my books on Google because I spent too much for each sale. That’s been the case here as well – at least for me. Ideally, you want to get caught up on a virtuous circle, where your ad sales feed into your organic ones. By the term organic, I mean those sales you get from the “people who bought this also bought…” mentions etc. Amazon is really good at this, and it’s one of the pleasures of working with them.
As I ended up paying Amazon almost nothing, due to the poor conversion rates, perhaps the extra impressions would help raise awareness of my book, so that its organic sales would be helped. Or at least that’s how I consoled myself.
However, one of the first things I noticed was that my organic sales for the two books advertised plummeted as soon as I started advertising them. I can think of two reasons why this might happen:
- Amazon is no longer promoting books organically, as it focuses on paid promotions
- Amazon stops promoting organically books advertised, to make room for other books
Whichever the case, this was a shock to me. I’m used to selling a couple of books daily when not advertising. Selling none for days on end was a terrible outcome, even if I didn’t actually pay Amazon anything.
Use it wisely
As you may know, I’ve started selling my fantasy book Pearseus: Rise of the Prince for 99c. As I want to raise awareness to the fact, I will advertise it with Amazon. However, I will do it knowing I will end up paying through the nose for each sale. The real benefit will be the huge number of people who will have caught a glimpse of the first book in my series by the time my $100 budget runs out – perhaps as high as 1,5 million eyeballs. This will raise awareness, and hopefully help make readers buy the rest of the Pearseus series as well.
To make sales, however, I will advertise it on other places. For example, when I last advertised with Book Gorilla I made 29 sales for my $50. Even better, ENT gave me 28 sales for only $15 (I am preparing a joint post with the lovely Tara Sparling on the subject of paid ads and which ones offer best value for money, so stay tuned).
Obviously, this is a far better outcome, if you’re interested in sales. If you’re interested in raising awareness for your book, however, you may still wish to give Amazon a serious look.
Either way, I’m looking forward to hearing about your experience with paid promotion!
Author CJ Heath has shared the following information, in response to my post:
“I have run an advertising campaign on Goodreads with CPC. I ran a six day campaign with $15. I played with the price of CPC and averaged 63c a click. My book was seen by 27,000 people. 13 clicked through and judging by my sales, 10 bought.
My first campaign (when my book launched) cost $60 at 70c. It was viewed by 143,000 people and 87 clicked through with sales in the high 70s.”
The above numbers suggest that:
- My CPC bid was too low to make a difference. It needed to be some ten times higher.
- My estimates as to how many people buy the book once they click on the ad was way too conservative. Heath seems to have had a number closer to 80% – much higher than my assumed 1%.
I will now head over to AMS and adjust the bid, as I just started a campaign to push Rise of the Prince, at the new price of 99c.
This is what Chris McMullen, our resident authority on all things Amazon and the man who has literally written the book on it, has to say:
“AMS ads are intriguing. I’ve now run 13 different ads, but still need more data to draw significant conclusions. I can say that 0.1% clicks-to-impressions is typical; I have hundreds of thousands of impressions to show that. I don’t have enough sales data to for a good estimate of sales-to-clicks. I’ve seen a couple of my ads exceeds 10% here, but that’s been the exception. I don’t have enough data for that to be statistically significant. I think it’s a good measure, though; if sales-to-clicks is well above 1%, it’s a good sign; if not, something’s wrong (product page, content, topic, targeting). That could be valuable info.
Clicks-to-impressions really don’t matter. You only pay for clicks, so wasted impressions don’t cost you. The problem is that it’s not easy to make a ton of impressions. So if you’re trying to generate significant sales frequency with the ad, the limited impression rate and click rate will make that a challenge.
For most of my ads, my sales actually went up, and I mean sales not reported with the ads. Even paperback sales went up at the time. Amazon is still promoting books like they had been; I’ve been on Amazon enough to see that, and still see the usual marketing emails. I suspect that the ads can generate indirect sales. In my case, customers seeing the ad are more likely to buy the paperback, and that won’t show on the ad report.
The ad reports were screwy the past few days. Numbers were going down. How is that possible? It got so bad, Amazon put a message above the reports briefly. So I’d say the data for the past few days isn’t reliable. It looks like it may have been fixed today. But there are also delays, and numbers are still growing for ads that have been paused for days. Sales themselves may be delayed for up to 14 days in the report. Cross your fingers; things might look slightly better than they do now.
These ads don’t seem to be quick or easy, but they probably have value to series authors, authors running hot promos, supplementing external promotions and marketing, better-known indies waving a flag, etc. The trouble is that short-term ROI can be dismal, and it’s hard to gauge long-term potential; and there are unique situations where that long-term is more apt to pay off.
Remember, the tool is new, everyone is eager to throw money at it, the bids are high, the impression rates are low. In time, things may be different. I have reason to believe the tool will change in the coming months.”
Don’t forget that you don’t need to pay anything to read my children’s book, Runaway Smile on my blog!