The tactical advice that dominated KDP communities for most of the past decade was built for a different algorithm. Understanding precisely what changed between A9 and A10 — not in vague terms, but in specific ranking signals — tells you exactly which of your current practices to keep, which to change, and which to abandon entirely.
| 10-minute read | Intermediate |
The problem with most KDP advice in circulation is not that it was wrong when it was written — it’s that it was written for a system that no longer exists. Authors following 2019-era keyword stuffing advice, 2020-era launch spike strategies, or 2021-era PPC bidding frameworks are not just failing to benefit from best practices — they’re actively working against the signals that A10 rewards. Understanding the structural difference between the two systems is the prerequisite for understanding any specific optimisation advice.
The A9 Model: How the Old System Worked
The A9 algorithm was a relatively transparent relevance engine built around three core inputs: keyword presence, sales velocity, and advertising spend. Keyword presence meant the algorithm compared search queries directly against the words in your title, subtitle, description, and backend keywords — the more times a relevant keyword appeared, the stronger the relevance signal. Sales velocity meant that books selling more copies ranked higher, creating a compounding advantage for established titles and a high initial barrier for new ones. Advertising spend affected organic rank directly — a book generating substantial sales through Sponsored Products campaigns accumulated enough sales velocity to improve its organic position, effectively allowing authors to buy their way up the rankings.
The system was gameable, and the self-publishing community was good at gaming it. Keyword stuffed subtitles that read as strings of search terms rather than descriptive text. Coordinated launch teams engineering artificial velocity spikes. Aggressive PPC bidding by authors with deep pockets or a willingness to operate at a loss for ranking purposes. Mass-publishing of low-quality content at high volume to accumulate sales across many titles. All of these tactics exploited specific mechanics of A9 that A10 was explicitly designed to make obsolete.
The Structural Differences: A Comparison
The shift from A9 to A10 can be understood as a move from measuring inputs to measuring outcomes. A9 evaluated what an author did — keywords added, money spent, sales generated. A10 evaluates what the reader experienced — did the book match what the listing promised, did the reader engage with the content, did the purchase satisfy the intent behind the search.
On primary ranking signals: A9 prioritised internal sales velocity and PPC volume. A10 prioritises organic engagement, relevance quality, and seller authority. A book can rank well under A10 without advertising if its organic engagement signals are strong; a book with strong advertising but poor engagement signals will find that its ad spend generates diminishing returns on organic rank.
On search logic: A9 used lexical matching — comparing strings of characters. A10 uses semantic understanding — inferring intent and context. A search for “enemies to lovers fantasy romance” under A9 surfaced books containing those exact words. Under A10, it surfaces books that the algorithm’s contextual understanding classifies as matching that reader intent, regardless of whether those exact words appear in the listing. This is why natural, reader-oriented copy now outperforms keyword-dense text — the algorithm’s semantic layer responds to coherent, purposeful language rather than keyword proximity.
On external traffic: Under A9, external traffic had minimal impact on organic ranking. Under A10, sales from external sources carry approximately three times the ranking weight of sales from internal advertising. Amazon now treats off-platform traffic as an authenticity signal — evidence of genuine demand that the platform wants to reward and encourage. This single change has made email list building and social platform presence part of the SEO strategy for every serious KDP author.
On seller authority: A9 had no meaningful equivalent of the A10 seller authority concept. A10 builds a composite reputation score for each publishing account based on account age, return rates, feedback scores, and content policy compliance history. Established accounts with clean histories receive preferential treatment in search results. This is why consistent, quality publishing over time has compounding ranking benefits that no short-term tactic can replicate.
What A9 Tactics Now Actively Harm Under A10
Several specific practices that improved A9 performance now suppress A10 performance. Understanding which ones is essential for auditing your current approach.
Keyword stuffed subtitles — subtitles formatted as lists of search terms rather than descriptive text — now create negative signals. A10’s semantic understanding recognises unnatural language patterns. A subtitle that reads “Mystery Thriller Suspense Detective Crime Novel Fiction Book” is not interpreted as a relevant, high-quality listing; it’s interpreted as a low-quality listing attempting to manipulate keyword presence signals. The A10 effect is that stuffed subtitles can actually suppress the ranking they were intended to improve.
Artificial launch spikes — coordinated purchase events designed to create a single-day sales velocity extreme — have diminishing value under A10’s decay-weighted ranking model, which averages performance across three to four weeks rather than rewarding point-in-time peaks. A spike that isn’t sustained has negligible impact on sustained organic rank, and if the spike was generated through manipulative means that result in high return rates, it can damage seller authority while generating negligible lasting ranking benefit.
Category accumulation — listing a book in irrelevant categories to collect bestseller badges — is now structurally impossible since Amazon’s 2023 restriction to three category slots per format. But the underlying tactic was already counterproductive under A10 even before the restriction: category misalignment creates recommendation mismatches that generate browser abandonment, high return rates from mismatched reader expectations, and negative review sentiment — all of which the A10 engagement depth evaluation picks up and uses as negative quality signals.
What A10 Rewards That A9 Ignored
Listing quality as a conversion driver has no meaningful A9 equivalent but is now a direct ranking signal under A10. A listing that converts a high percentage of browsers — because it accurately and compellingly represents the book to the readers who are genuinely right for it — tells A10’s engagement signals that the book is a high-quality match for its search terms. This generates ranking reinforcement from conversion data alone, separate from sales volume. The implication is that copywriting quality is now a search engine optimisation function, not just a marketing function.
Review velocity and sentiment — how frequently your book receives new reviews and what those reviews say — now inform A10’s quality assessment and specifically feed Rufus’s recommendation decisions. A book with 50 reviews that received its last review six months ago is treated differently from a book with 30 reviews that received three in the last month. Maintaining fresh review input through ARC programmes, strategic review request timing, and consistent reader engagement is now part of the SEO maintenance of a published book. The Amazon Book Reviews guide covers the specific tactics for maintaining review velocity.
Engagement depth — the amount of time browsers spend on your product page, how deeply they scroll, whether they open the Look Inside preview — is evaluated by A10 as a quality signal. A listing that generates engagement before a purchase tells the algorithm that the book is genuinely interesting to the readers it’s shown to. This is why A+ Content, a strong Look Inside opening, and a description that earns the scroll rather than losing readers above the fold are now ranking considerations, not just conversion considerations.
The tools that perform best in this environment are those that apply both data and expert copy judgment simultaneously. KDP Rank Fuel combines real Amazon market data with the copywriting methodology Vappingo has developed across 15 years of KDP listing work — identifying the opportunities the research reveals and then producing the copy that wins them. The Alliance of Independent Authors provides a detailed overview of the evolving Amazon discovery landscape at allianceindependentauthors.org — useful context for understanding how A10’s changes sit within the broader self-publishing environment.
The Tactics That Still Work Under A10
While A10 has made many A9-era tactics counterproductive, it has not changed the fundamentals of what sells books. A great cover that accurately signals genre, a description that opens with a compelling hook, targeted keyword selection based on actual search data, and consistent review acquisition — all of these work as well or better under A10 as they did under A9. What A10 has changed is which execution approaches serve these goals. The keyword research is now about semantic intent mapping rather than keyword density planning. The description writing is now about natural, specific copy rather than keyword repetition. The review strategy is about velocity and authenticity rather than volume alone.
The tools that consistently produce the best results under A10 are those built on a combination of real Amazon data and genuine copywriting expertise — not keyword lists alone, not generic text generation, but the specific understanding of what language converts readers in each genre that comes from years of direct KDP listing work. That combination is what separates the research-only tools from platforms like KDP Rank Fuel, where the data from the Competition Analyzer and Niche Navigator feeds directly into listing copy built on 15 years of KDP copywriting experience. Written Word Media publishes an annual reader survey at writtenwordmedia.com that provides useful genre-specific data on reader discovery and purchasing behaviour — valuable context for authors calibrating their A10 strategy to their specific genre audience.
The most revealing test of whether your current practices align with A10 or A9 is simple: look at your last three published books’ BSR trajectories over 90 days. Under A10, books with strong listing quality, accurate genre signalling, and consistent external traffic activity show a gradual sustained rank rather than a spike-and-crash pattern. Books formatted around A9-era tactics — keyword-stuffed subtitles, aggressive short-term PPC without supporting organic engagement — show a spike at launch and rapid decay to dormancy. If your books look like the second pattern, the issues described in this guide are the most likely causes. The KDP Sales Rank Decline guide covers the specific diagnostic and remediation process for each cause.
A10 Rewards the Book That Delivers
Return rates, review sentiment, and engagement signals all feed back into your A10 ranking. A professionally proofread manuscript that earns positive reviews and low return rates is now a ranking asset. Vappingo’s manuscript proofreading service is the production step that protects your A10 signals.
