With 86 Twitter followers, Kindle royalties reach ¥75,000 a month. Ten books, two years, and what an all-in bet on SEO looked like

When individual Kindle publishing gets discussed, “SNS influence” is almost always assumed to come with it — the premise being that you gather followers and sell books to them. This case removes that premise. The author’s Twitter follower count is 86. And yet monthly royalties have reached ¥75,000.

The source for this case is an interview published on the note account “Kan | AI × Kindle Publishing,” featuring author Nagataki. All figures are based on the subject’s own statements, as of December 2022.

¥75,000 a month on 86 followers

At the start of the interview, the interviewer, Kan, introduces the guest this way: “What’s amazing about Nagasaki-san is that their Twitter follower count is under 100. I just checked, it’s 86, and their monthly royalties are ¥75,000.”

A follower count of 86 is effectively negligible as an SNS acquisition channel. The core of this case is right there: the buyers behind this ¥75,000 are strangers who discovered the book inside Amazon rather than Twitter followers.

The publishing pace is “roughly one book every two months,” and by the time of the interview the cumulative total was 10 books. The genres aren’t specialized: “savings, job changes, side hustles, asset management, IT, I put them out for beginners and those without experience.”

Two years of royalty growth (in the author’s own words)

Nagataki verbally recounts the monthly trajectory. These figures are the author’s own rough recollection, not precise bookkeeping to the yen, and should be read with that in mind.

Nagataki’s own summary: “It went up gradually, over two years.” Looking back at when the first book earned ¥3,000 a month, the reaction at the time was: “I thought that was plenty, honestly kind of a lot.”

The turning point wasn’t a “lucky hit”

Looking at the shape of the graph alone, the November 2021 game-engine book looks like the turning point. The author even says, “it sort of just happened to hit, and I earned ¥30,000 from one book.” But over the following year, the number only oscillated between ¥20,000 and ¥40,000, growth had stalled. One hit built a peak, but it didn’t raise the floor.

The floor changed in February 2022. That month, Nagataki purchased a course covering Amazon-internal SEO and completely restructured how titles, keywords, and descriptions were built. “I felt the Amazon SEO content itself fit me, so I bought it, actually tried it, and it started ranking well, I got to where I could earn about ¥30,000.”

Then, in late December 2022, using a tool learned about through a different course, thorough SEO measures were applied to a new title on AWS certification. That’s where the level jumps. “From there, royalties (which had been bouncing around ¥30,000, ¥20,000, sometimes ¥40,000) went up to ¥70,000.”

Lining things up: 11 months stagnating in the ¥20,000-40,000 range, then one new SEO-implemented title brings it to the ¥70,000 level. What turned the trajectory was changing the same book to be findable through search, not “writing a good book.”

Why did it work?

Nagataki’s measures can be organized into three layers. The reason each worked differs.

Layer 1: Blanketing Amazon’s internal search surface. Against the recommendation that “titles should stay under 60 characters,” the approach was “getting it to 59 or 57 characters, packing it in right up to the limit.” For the CCNA book: “I was targeting the keyword ‘CCNA,’ but not just that, related terms that come up, like CCNA workbook, textbook, CCNA 200-301, and so on, as long as it’s not unnatural, I’d put those in the title too.” Words that didn’t fit in the title went into the seven keyword fields. The description was used right up to its 4,000-character limit, pasting in the full table of contents to expand the vocabulary further. A+ Content was used up to its five-image cap, embedding keywords in the images as well, and an author page was created to build “traffic from other books.” The reason this worked is straightforward: choosing a genre (certifications, tools, product names) where the search terms are already fixed in advance, and then filling in the surrounding vocabulary alone, one person, exhaustively. Unlike social media, the search surface is won not by influence but by comprehensiveness and match.

Layer 2: Buying the first three launch days with a free promotion. “Every one of the ten books I’ve published so far, I’ve run a free campaign at launch.” The window is three days, and the reasoning is simple: “It felt like there’d be a lot of missed opportunity, so I intuitively picked three days.” Promotion targets a LINE open chat with a few hundred members, plus friends, acquaintances, and coworkers. “If it were priced at ¥99, I don’t think anyone would buy it, so a free campaign is really the only option.” Downloads during the free period serve to seed the initial ranking and reviews. After switching back to paid, the author fires off roughly 15 promotional tweets tagged with hashtags like “connecting with junior engineers.”

Layer 3: Borrowing other people’s influence. For the AWS book, the author reached out on Twitter to influencers and companies featured inside the book, saying “I wrote about you like this,” and “they helped amplify it.” The result: about 315 downloads during the free campaign and a paid ranking of 1,206th. Since the author personally has no reach with only 86 followers, reach has to be borrowed, and the practical move is manufacturing a pretext to borrow it, by creating the fact “I featured you” beforehand.

What isn’t being done, and where the limits lie

Nagataki has essentially abandoned SNS operations. “Tweeting multiple times a day is a hassle, and if I’m going to put in that effort, I’d rather just publish my next book faster.” Monetization is royalties alone. Content sales or a mailing list weren’t part of the picture at the time. The two-year, ¥75,000-a-month level is also a consequence of that choice.

The author also sets limits on the applicability of this approach. “If you want to earn big, honestly, my method isn’t going to be much of a reference.” But: “if you’re someone who just wants to do Kindle without doing much SNS promotion, my method could be useful as a reference.”

The risk location is also clear. Revenue rests on two variables the author doesn’t touch, Amazon’s search algorithm and the Kindle Unlimited payout rate. If per-page rates move, royalties move. And the core tactic, packing related keywords into the title, is the kind of thing that could be invalidated overnight if platform guideline enforcement changes.

How far can this be replicated?

What’s easiest to replicate is genre selection and search-surface design. Choosing a theme (a certification, tool, or product name) where the terms readers will type in are already fixed, and consistently placing the related vocabulary across the title, keyword fields, description, and images. This work requires neither influence nor capital, has near-zero additional cost per book, and, once 10 books have accumulated, generates cross-traffic through the author page too.

Three things resist replication. The sheer continuity of publishing 10 books over two years is the hardest. As Kan puts it, “the fact that you kept at it steadily for nearly two years is probably why you’re seeing results like this”, whether one can keep publishing a book every two months without giving up at ¥3,000 a month is the biggest barrier. Having a writable area of specialized expertise, in this case IT certifications, is the quieter requirement. Timing of entry rounds it out: the number of competing titles in the same category differs between 2021 and now, and there’s no guarantee the same SEO tactics produce the same ranking.

And it shouldn’t be forgotten that the source for this article is a note account run by someone who sells Kindle-related courses. The courses Nagataki purchased are also that source’s own products. The figures can be treated as the subject’s own statements, but the causal claim, “sales rose because I bought the course”, is a claim voiced on the seller’s own platform, and should be discounted accordingly. What’s verifiable is only the timeline fact: monthly royalties shifted from ¥20,000-40,000 to the ¥70,000 range before and after SEO implementation.

Related reading

• 24 Kindle books earning ¥50,000-70,000 a month: real monthly data from 16 months and ¥750,000 cumulative from a “mass production” self-publishing model — the fully-disclosed, month-by-month version of the volume-driven approach
• Hitode’s blog revenue — the long-term trajectory of a solo media outlet running mainly on search traffic