The advice in KDP author communities is consistent: send external traffic to Amazon. It improves your visibility, signals to the algorithm that your book is worth showing to more people, helps you rank. The principle is correct. What the advice almost never explains is what kind of external traffic actually does that work, and why a lot of the traffic authors are actively sending doesn’t.
The short version: Amazon’s algorithm responds to conversion rate, not click volume. When someone lands on your book page from an external source and buys, the algorithm records that an outside visitor converted. Enough of those conversions, quickly enough, and your visibility goes up. That’s the mechanism the advice is pointing at. The problem is that click volume and conversion rate are entirely different numbers, and conflating them is what produces most of the external traffic effort that doesn’t work.
Here is what that looks like in practice. If you’re driving 2,000 people from a social media campaign to your book page and 30 of them buy, your external conversion rate is 1.5%. Amazon has now logged 1,970 data points of people who looked at your page and left without purchasing. That can suppress your page’s overall conversion metrics in ways that don’t appear as a headline number but affect how the algorithm classifies your book over time. High click volume with low conversion doesn’t register as signal. It registers as noise.
What warm external traffic actually looks like
The external traffic that earns algorithmic benefit has done the qualifying work before the reader lands on your page. Your newsletter list is the clearest example. A subscriber who signed up after finishing your last book, who has opened your last three emails, and who clicks through on your new release announcement: that reader is arriving at your Amazon page with intent. The conversion rate on that cohort is high because the selection happened before the click.
BookBub Featured Deals work the same way. The reason a Featured Deal produces ranking movement isn’t just volume. BookBub’s reader list is segmented by genre and purchase behaviour, meaning the people clicking through have already self-selected for your category. The traffic converts at scale, and that’s what the algorithm is registering. It’s not just that a lot of people are clicking. It’s that a high proportion of the people clicking are buying.
Promo stacks function on the same logic. Readers engaging with a genre-specific newsletter service are arriving because they read your genre. The conversion rate is higher than cold traffic because the audience has already been filtered before landing on your page.
What cold traffic produces
Social media followers are not the same as warm readers, and this is where a lot of external traffic strategy breaks down. Someone who follows you on TikTok because they enjoyed a video about your writing process is not the same as someone who follows you because they have read and enjoyed your genre. When a post goes viral and drives 8,000 clicks to your book page, the number that matters isn’t 8,000. It’s how many of those 8,000 people bought.
If the answer is 70, your conversion rate from that spike is under 1%. Amazon has registered a significant volume of non-converting visits. That doesn’t erase the sales, but it can flatten the algorithmic benefit you might otherwise have expected from a sudden traffic spike.
This doesn’t mean social media is useless for an indie author’s marketing. Social media builds name recognition and warm audiences over time, both of which eventually show up as converting traffic. But the mechanism is indirect and slow, and it’s different from the mechanism that produces ranking benefit. Expecting a viral TikTok post to function as an external traffic boost in the algorithmic sense is asking it to do a job it’s not designed for.
Broad paid advertising aimed at cold audiences produces the same pattern. Traffic from a Facebook ad aimed at general interest categories will convert at a low rate, and if the volume is high, the ratio works against you. The more productive use of paid traffic for ranking purposes is advertising against warm audiences: your own email list, lookalike audiences built from previous buyers, or genre-specific interest segments where the reader has already self-selected.
The number worth tracking
If you’re actively managing external traffic efforts, the metric to watch is your click-to-sale ratio by source, not your total click count.
A newsletter click that converts at 20-30% is doing real algorithmic work. A paid ad converting at 1% may be generating sales but not the ranking benefit the effort was aimed at. When both numbers are available, the ratio tells you more about whether your external traffic strategy is functioning than either number alone.
Most authors track sales and not much else. The sales number matters, but for evaluating whether your external traffic is doing what external traffic is supposed to do, conversion rate by source is the diagnostic figure. It’s also usually the figure that explains why "I sent traffic to Amazon" didn’t produce the ranking results the advice implied it would.
If your current launch plan includes external traffic as a ranking strategy, look at where that traffic is coming from and what its likely conversion rate is before launch week. Warm sources are doing the job. Cold sources may be filling the column without filling the function.
