How many pages does a product site need?
Eight to twelve pages cover a product completely: what it does, what it costs, who you are, how to start. Every page after that has to earn its place by answering a search or closing a doubt. This page is about which pages must exist at all; how fast you can add more, and how often, are separate questions.
By Théophile Louvart, founder of Porteur · Updated 15 September 2026 · Markdown
Page count is an outcome, not a target
There is no number of pages that ranks. Google has said there is no word count that ranks either, and the same logic applies upward: a site of thirty pages that each answer a search outranks a site of three hundred that do not.
- A page can only rank for one cluster of searches, so more searches means more pages. That is the only honest reason to add one.
- Every page is crawled, judged and stored. Pages with nothing to say spend that budget without returning anything.
- Every page is also yours to keep accurate. Three hundred pages is three hundred things that can go stale.
So the question is never "how many". It is "which searches do I want, and does each one need its own page".
The pages every product needs
These exist because people look for them by name, not because of search volume. A product missing one of them looks unfinished to a buyer and to an assistant answering questions about your category.
| Page | Why it exists | The search it answers |
|---|---|---|
| Home | The promise and the proof. | Your brand name, and the job you do. |
| Pricing | The first thing a buyer checks. | "yourproduct pricing", "yourproduct cost" |
| What it does | One page per real capability. | The job: "import csv to postgres" |
| About | Who is behind it, where, how to reach them. | "who makes yourproduct", and trust checks |
| Docs or getting started | Proves the product is real and answers setup. | "yourproduct api", "how to set up yourproduct" |
| Terms and privacy | Buyers check before paying. | Rarely searched, always looked for. |
| Changelog | Shows the product is alive. | "yourproduct updates" |
| Contact or support | A way to reach a person. | "yourproduct support" |
Eight pages. Add an integrations page if you have integrations, a security page if you handle data, and a status page if you sell uptime. Ten to twelve, and the product is completely described.
What every page after those must earn
Beyond the product pages, each addition should pass one of four tests. A page that passes none is a page you will be deciding what to do with in a year.
- It answers a search that has demand and that no page of yours answers. Check the second half: an existing page probably covers it.
- It closes a doubt that stops buyers: the comparison, the alternatives page, the security page, the case study.
- It is something people link to: a tool, a study, a table nobody else publishes.
- It is a task your users have to perform, which is documentation.
The point where more pages start to hurt
Adding pages is not neutral. Three specific things go wrong, and they arrive in this order.
| What goes wrong | How it shows | The fix |
|---|---|---|
| Two pages answer one search | In the Performance report, two of your URLs trade places on one query and neither settles. | Merge them, keep the stronger URL, redirect the other. |
| Pages nobody links | Orphan pages, reachable only from the sitemap, with no impressions after months. | Link them from the relevant hub or remove them. |
| The site gets judged | A batch of thin pages sits in "Crawled, currently not indexed", and older pages slip at the same time. | Cut the batch back to the pages that hold something of their own. |
The third is the one to take seriously. Google's spam policies describe scaled content abuse as many pages made primarily to rank rather than to help, whoever or whatever made them, and the judgement is about the site as much as the pages.
What a library is for, and when to build one
Some products end up with hundreds of pages legitimately. The difference between a library and a pile is that a library answers many distinct questions and is maintained.
- Build one when you can name twenty distinct questions your buyers ask, in their words, that no page of yours answers.
- Build it in families where the pages differ genuinely: one per error, one per term, one per tool comparison, one per task.
- Structure it as hubs and pages, so nothing is three clicks deep and nothing is an orphan.
- Maintain it: a quarterly pass over the pages with the most impressions, checking every claim.
A hundred maintained pages is an asset. A hundred abandoned ones is a liability you will spend a quarter unwinding.
A worked count for a small product
Take yourproduct.com, a fictional invoicing tool for freelancers, eighteen months old, one founder.
| Group | Pages | What decided the number |
|---|---|---|
| Product | 10 | Home, pricing, four capabilities, about, docs index, terms, privacy. |
| Documentation | 24 | One per task a user performs. |
| Comparisons and alternatives | 6 | The five tools buyers weigh against it, plus one alternatives page. |
| Guides | 15 | The jobs and the problems, one per question that support actually receives. |
| Glossary | 12 | The terms the guides use without defining. |
| Total | 67 | Every one of them answers something somebody asks. |
Sixty-seven pages, built over eighteen months, maintainable by one person in a few hours a month. No page in that list exists because a number needed reaching.
Questions
There is no number. A site of ten pages can rank for ten clusters of searches if each page answers one well. More pages help only when each answers a search nothing else on your site answers.
Only if the pages are good. Google has said there is no penalty for having many pages, and also that low-quality pages can affect how a site is assessed. Size is a consequence of covering more subjects properly, not a cause of ranking.
Eight to twelve: home, pricing, what it does, about, docs or getting started, terms, privacy, contact, plus a changelog and an integrations page if they apply. That describes a product completely.
Not automatically. Check whether the page has links, serves a purpose for users, or is new. Then decide: improve it, merge it into a stronger page with a redirect, or remove it with a 410. Do it on data over months, not on a hunch.
A useful rule: whatever you can review in an afternoon each quarter. For most people that is somewhere under a hundred pages, unless the pages are short and stable like a glossary.
Check my site, free
Paste your URL and the free check reads your site and the searches around it in about thirty seconds, then names the pages worth building and the ones already competing with each other.
- Free check, no card
- Read-only, your own accounts
- Readable by your agent
Read next
- GuideContent velocity: how many pages a small site can add without harm
- GuideShould your startup have a blog?
- GlossaryThin content
- GuideA content audit for a small site: one table, one decision a row
- GlossaryIndex bloat
- GuideOrphan pages: finding the pages nothing links to
- GuideDelete or redirect an old page: 301, 404 or 410
- GuideDoes a startup need SEO? A straight answer by stage