Amazon’s “A10 Algorithm” for Authors: What We Know, What We Don’t, and What to Do
A fact-checked guide to Amazon book discovery—separating official guidance, reasonable inference, and unsupported ranking claims.
Amazon changes constantly. Agencies and authors must form hypotheses when public documentation is incomplete. But a hypothesis should be labeled as a hypothesis—especially when it could influence an author’s budget, launch, metadata, or decision to republish a book.
This update separates what Amazon explicitly documents, what is a reasonable marketing inference, and what remains unverified. That distinction does not weaken the strategy. It makes the strategy more useful.
Is Amazon A10 real?
“A10” is widely used by marketers to describe an assumed evolution of Amazon’s search and recommendation systems. In the current Amazon and KDP documentation reviewed for this article, Amazon does not name a book-search algorithm called A10, announce that A9 was replaced for books in July, or publish a weighted list of ranking factors.
That does not mean Amazon’s systems stand still. They undoubtedly evolve. It means outsiders should not present a private, changing system as if Amazon had released a public technical specification.
For authors, the A9-versus-A10 debate can become a distraction. The practical question is simpler:
What can we improve that helps the right reader discover, evaluate, buy, borrow, and continue through the catalog—and how will we know whether it worked?
Use three evidence labels
Verified
Amazon or KDP explicitly documents the capability, input, or customer-facing purpose. “Verified” does not mean we know its exact weight in search.
Reasonable inference
The idea fits customer behavior, business logic, or observed account data, but Amazon has not confirmed it as a direct ranking factor.
Unverified
The claim assigns a specific rule, priority, formula, or causal effect that current public Amazon documentation does not support.
A fact-check of common Amazon A10 claims
| Popular claim | Status | Responsible interpretation |
|---|---|---|
| Amazon replaced A9 with A10 for books in July. | Unverified | Amazon’s current author documentation does not publish that change or date. |
| External traffic is now a major direct ranking factor. | Unverified | Amazon permits outside promotion and offers Attribution to measure its on-Amazon impact. That does not disclose a ranking weight. |
| Keywords, categories, and book details affect findability. | Verified | KDP explicitly recommends relevant keywords and categories and says search can use keywords, book details, and book text. |
| Amazon now values organic sales more than ad-attributed sales. | Unverified | Amazon does not publish an organic-versus-paid weighting or ratio for book search. |
| A strong detail page can improve conversion. | Verified | KDP’s advertising guidance says a strong detail page can help convert clicks into sales. |
| A+ Content directly raises organic rank. | Unverified | KDP describes A+ as a way to engage and inform readers; the reviewed guidance does not promise a rank increase. |
| Time on page and repeat visits are known book-ranking factors. | Unverified | No public weight or book-search rule for these signals appears in the reviewed documentation. |
| Ten thoughtful reviews rank better than fifty short reviews. | Unverified | Reviews can inform purchase decisions, but Amazon does not publish that ranking rule. Review manipulation is prohibited. |
| Amazon has a special 30-day algorithm window for every launch. | Unverified | Sustained launch support is sensible, but no universal 30-day ranking rule is documented. |
| Consistent promotion can support ongoing discovery. | Inference | Amazon Ads recommends sustainable, always-on backlist advertising, while stopping short of publishing an organic-rank formula. |
The useful parts of many “A10” articles survive this review. Accurate metadata, better conversion, qualified traffic, steady discovery, and reader retention still matter to an author’s business. The unsupported certainty does not.
What Amazon officially tells authors about discovery
Search is dynamic
KDP says results can fluctuate and may use keywords, book details, and the text of the book. A short-term movement is a signal to investigate—not proof of an update.
Metadata supports findability
KDP recommends accurate keywords and relevant categories. Its keyword guidance says placement can also reflect sales history and Amazon Best Sellers Rank, without publishing weights.
The retail page must convert
KDP describes the description as a short pitch, and its advertising guidance says a strong detail page can help convert clicks into a sale.
Catalog paths matter
Accurate series metadata can group connected titles on a Series Collection page. Author Central gives readers a place to discover an author’s catalog.
Amazon Ads creates discovery
KDP says ads can place books in shopping results and detail-page positions. Amazon’s current backlist guide supports sustainable, always-on visibility for suitable titles.
Outside promotion is measurable
KDP allows email, websites, and other outreach. Amazon Attribution can measure eligible non-Amazon search, social, video, email, and influencer activity.
Why external traffic still belongs in the plan
If Amazon does not confirm an external-traffic ranking boost, why invest in social content, email, paid social, podcasts, or a website?
Because they can create business outcomes without needing a mythical algorithm reward:
- More qualified readers can discover the author.
- Followers and subscribers become reachable audiences.
- A launch no longer depends on one platform or one ad auction.
- Readers may search the author or title later instead of clicking immediately.
- Messaging and creative can be tested before a larger campaign.
- Attributable sales can be measured where Amazon Attribution is available.
A real example from Gary McAvoy’s organic-video program
In one seven-video sample managed by Adverley, the clips generated:
Gary reported that his audience rose from about 6,800 to 7,100. He also observed a small sales spike on days when a new video was published and said readers were responding positively to the clips.
The gap between 109,253 views and 49 visible link clicks is instructive. Content can build recognition and prompt later branded search, profile visits, shares, or purchases that a visible link-click column does not capture. That is why we examine several signals instead of declaring success or failure from last-click data alone.
Amazon Ads should be judged as part of the reader journey
The original version of this post said A10 had reduced reliance on paid ads. A more defensible statement is:
Amazon Ads can buy relevant discovery. Whether that discovery creates economic value depends on conversion, KENP, read-through, later-book royalties, organic demand, and the author’s acceptable payback period.
Amazon Ads ACoS divides ad spend by ad-attributed retail sales. At Adverley, our Adjusted ACoS adds estimated ad-attributed KENP royalties to that denominator when the available reporting supports the estimate. It gives Kindle Unlimited authors a better front-end view, but it is still not a complete profit-and-loss statement.
That distinction becomes especially important when advertising Book 1 of a series.
For a standalone, a low-retention series, or a book with weak conversion, the conclusion can be very different. The correct target is not “low ACoS at any cost” or “high ACoS is fine.” It is profitable reader acquisition under the conditions of that catalog.
An evidence-led visibility system for authors
Instead of trying to satisfy a secret formula, build four connected layers.
1 · Retail foundation
Make the promise clear before sending more traffic.
- Genre-aligned cover and positioning
- Accurate keywords and categories
- Reader-focused description and sample
- Series, Author Central, and useful A+ Content
2 · On-Amazon discovery
Reach readers while they browse and search.
- Relevant titles, searches, and product targets
- Discovery separated from proven targets
- Sustainable budgets and bids
- Optimization over meaningful windows
3 · Off-Amazon demand
Give readers more ways to encounter and remember the author.
- Email and newsletter growth
- Organic short-form video and social
- Measured paid social
- Website, interviews, and communities
4 · Measurement & economics
Connect channel metrics to the catalog.
- Ads, KDP, and Attribution where eligible
- Sales, estimated KENP, and trends
- Read-through and reader value
- Acquisition cost, lag, and payback
What “A10-aligned” means at Adverley: our shorthand for an integrated visibility and reader-acquisition approach. It is not an Amazon certification, endorsement, or claim that we have access to a private algorithm.
A 12-step evidence-based action plan
Retail foundation
- Audit the cover at thumbnail size. Confirm that genre, tone, and professionalism are clear before spending more for clicks.
- Strengthen the description and sample. Make the promise, stakes, and reader fit obvious; ensure the opening rewards the click.
- Review metadata for accuracy. Use KDP’s seven keyword fields thoughtfully, select relevant categories, and remove promotional or misleading phrases.
- Connect the catalog. Verify series order, formats, Author Central, and A+ Content in each important marketplace.
Discovery
- Separate visibility from profitability. Diagnose impressions, clicks, conversion, orders, KENP, and catalog revenue independently.
- Use Amazon Ads for relevant discovery. Structure campaigns so you can learn which searches, products, and creative angles attract the right readers.
- Choose a sustainable test budget. Predetermine the learning budget and the conditions for continuing, revising, or stopping.
- Avoid daily overcorrection. Evaluate enough data to distinguish a pattern from normal volatility while respecting cash-flow limits.
Demand and measurement
- Build a reachable audience. Use email, organic content, or community-building so every launch does not begin at zero.
- Test external traffic with a purpose. Define whether the campaign should generate sales, subscribers, followers, video engagement, or creative insight.
- Use Amazon Attribution when eligible. Measure non-Amazon activity without assuming every untracked purchase had no influence.
- Judge the catalog, not only the advertised order. Include estimated KENP, later-book behavior, formats, acquisition cost, lag, and payback—without treating projections as settled royalties.
If sales or KENP decline, diagnose before blaming an algorithm
| What changed? | Inspect first | Do not conclude yet |
|---|---|---|
| Search or ad impressions fell | Eligibility, budget, bids, targeting, competition, categories, keywords, format availability | “Amazon penalized my account.” |
| Click-through rate fell | Cover, title, price display, target relevance, creative fatigue, competing releases | “Paid traffic no longer works.” |
| Clicks held but orders fell | Description, sample, reviews, price, format, audience match, page changes | “The algorithm stopped converting.” |
| Ads held but total sales fell | Organic baseline, seasonality, launches, promo calendar, catalog demand | “The ads caused the decline.” |
| KENP fell | KU enrollment, page-read lag, reader progression, release cadence, seasonality | “A10 suppresses KU books.” |
| Video views rose but visible clicks stayed low | Hook, call to action, profile path, branded search, follower growth, Attribution | “The videos had no effect.” |
Compare like-for-like periods when possible. Annotate launches, price changes, promotions, outages, metadata edits, and major budget changes. One metric rarely identifies the cause by itself.
What not to do in the name of A10
- Do not keyword-stuff titles, subtitles, or descriptions.
- Do not buy, trade, or improperly incentivize reviews.
- Do not republish solely because someone promised a ranking reset.
- Do not send broad traffic to a page that does not convert.
- Do not stop a series-acquisition program because one front-end metric crossed an arbitrary threshold.
- Do not scale an unprofitable catalog because someone said external traffic is always rewarded.
- Do not describe correlation as proof of algorithmic causation.
The strategic takeaway
Authors do not need certainty about a private algorithm to make better decisions.
The strongest response to changing visibility is to improve the retail page, maintain relevant on-Amazon discovery, build off-Amazon demand, and measure what happens across the reader journey. Some of those actions may influence Amazon’s systems. All of them can improve the business when executed for the right audience and evaluated honestly.
Replace the algorithm question
Do not ask only, “How do I make A10 like my book?” Ask:
How do I make it easier for the right reader to discover this book, choose it, enjoy it, and continue—and can the economics support doing more of what works?
That is a strategy an author can control.
Frequently asked questions
Is Amazon A10 an official book-search algorithm?
Not in the current public Amazon and KDP documentation reviewed for this article. “A10” is an industry label. Amazon may change its systems without publishing every detail, but marketers should not present an undisclosed formula as confirmed fact.
Did Amazon replace A9 with A10 for books in July?
We could not verify that claim in current official author documentation. A change in sales or KENP after July does not, by itself, establish a platform-wide algorithm replacement.
Does external traffic help Amazon book sales?
It can. Qualified external traffic can create awareness, followers, subscribers, branded searches, sales, and KENP. Eligible KDP authors can use Amazon Attribution to measure some on-Amazon outcomes. A special direct ranking bonus for external traffic is not publicly confirmed.
Are Amazon Ads less valuable now?
Amazon’s current KDP and Amazon Ads guidance still presents advertising as a tool for book visibility and reader discovery. Ads should be judged against the catalog’s conversion and reader economics, not assumed to guarantee organic ranking.
Does A+ Content improve organic rank?
KDP says A+ Content can make a detail page stand out and give readers more information. The documentation reviewed does not promise a direct organic-ranking increase. Use A+ to improve the reader’s evaluation experience.
What ACoS should an author target?
There is no universal answer. Standard ACoS measures ad spend against ad-attributed retail sales. Adjusted ACoS can add estimated attributed KENP, but neither automatically captures later-book value or full costs. The sustainable threshold depends on royalties, formats, read-through, acquisition cost, lag, cash flow, and the author’s goals.
What should I do first if orders or KENP fall?
Compare meaningful periods and identify which part of the funnel changed: visibility, clicks, conversion, page reads, or later-book progression. Then inspect campaign, retail-page, catalog, seasonal, and promotional factors before assigning the decline to an algorithm.
Continue reading
Want a clearer view of your author marketing?
Adverley helps authors connect Amazon Ads, off-Amazon discovery, retail-page readiness, and catalog economics. We do not promise a secret algorithm shortcut. We help identify what can be tested, measured, and improved.
Request an audit Explore AdverleySources and further reading
Editorial note: Platform documentation and product availability can change. This article reflects the public sources reviewed in August 2026. “A10-aligned” is Adverley terminology, not an Amazon program or certification. Client examples are observational and are not guarantees of future results.
