Launching on Amazon KDP: What Nobody Tells You
Everyone who talks you into self-publishing tells you the same thing: write the book, upload the file, watch the sales come in. Nobody tells you what actually happens between “upload” and “watch the sales come in,” which is where most of the real work lives. Here's what we learned putting our first title live on Amazon KDP.
The listing is a storefront, not a page
A book's Amazon listing isn't a description with a buy button attached to it. It's a storefront window, and it gets judged the way a storefront gets judged: in about two seconds, mostly on the cover thumbnail, before anyone reads a word of the copy underneath. That means the cover has to work at the size of a postage stamp on a phone screen, not just at full size on a desktop. Titles that read beautifully in a design file can turn to mush at thumbnail scale. We redesigned ours twice after seeing it next to actual search results, not in isolation.
The cover has just as much work as the interior
It's tempting to treat the cover as the last five percent of the project, the part you knock out once the real work, the interior pages, is done. In practice the cover carries more of the sales weight than the interior does, because the interior only gets seen after someone has already decided to click. Back cover copy, spine width, trim size, they all have to be locked before the interior is finalized, not after, because KDP calculates spine width from page count and paper type, and a late page count change means redoing the whole cover file. We found this out the hard way when a 20 page count needed to move to meet KDP's 24 page minimum, well after the cover was supposedly finished.
Categories and keywords decide who ever finds it
Amazon doesn't have a “browse everything” button. People find a book through search terms and category browsing, and both of those are entirely in the author's hands, not Amazon's. Seven keyword slots, two categories at listing, more available on request, and each one is a bet on how a real person searching for something to buy would phrase it. Generic terms get buried under thousands of competitors. Specific terms get less traffic but a much higher share of the traffic they do get. Getting this right takes more research than the interior design does, and it happens before a single review exists to help the listing along.
Price, territories, and the fields nobody reads
The pricing and rights screens in KDP look like formalities, checkboxes between you and the publish button. They're not. Price sets your royalty rate directly, and the difference between two price points can be the difference between a book that earns and one that doesn't, once printing costs come out of the math. Territory selection decides which countries can actually buy the physical copy. These are easy to click through on autopilot and easy to get wrong in a way that only shows up weeks later, in a sales report that doesn't add up.
Launch week, and the hype that matters
Amazon's algorithm pays outsized attention to what happens in the first few days a listing is live, sales velocity, review activity, click-through, all of it. That means launch week isn't a celebration lap after the work is done, it's the last and most consequential phase of the work. Whatever attention a title is going to get, it earns more from that attention in week one than it will from the same attention spread out over month three. We treated our launch like a deadline with a fixed date, not a soft opening, and it changed how the first month of numbers looked.
None of this is a reason not to self-publish. It's a reason to go in knowing that the listing, the cover, the categories, and the launch date are the actual project, not the paperwork around it. The writing gets you a manuscript. Everything above gets you a book someone can actually find.
Have a launch of your own in the works? We'd like to hear about it: dylan@twobentspoons.com