usdigitalmarketingindex.com SEO
How can a large FAQ library be organized without generating hundreds of thin pages?
Updated November 15, 2026
Group the questions by the task the reader is trying to complete, answer the short ones together on one page per task, and give a question its own URL only when its answer needs space and stands as a complete resource. A library built this way has a small number of substantial task pages, a handful of standalone answers, and an index that leads to both. The failure it avoids is the one-question-per-URL library, where hundreds of pages each hold two sentences and the same header and footer. No word count separates a thin page from a sufficient one. The test is whether the page fully serves the person who asked.
Grouping questions by user task
Sort by what the person is doing, not by department or product name.
- Collect every question, with its current answer, in one sheet.
- For each, write the task behind it: setting up, paying, fixing a fault, returning an item, cancelling.
- Merge questions that are the same question in different words.
- Group the rest under their task.
- Order each group in the sequence a person would meet the questions.
A group of fifteen billing questions is one page about billing with fifteen answers. A reader with one billing question may have a second, and finds it without another search.
Choosing inline answers versus standalone pages
Decide per question, with the same four checks.
| Check | Inline on the task page | Standalone page |
|---|---|---|
| Length of a complete answer | A few sentences | Several steps, a table, or images |
| Does the answer branch by situation? | No | Yes, with different paths |
| Do people search for this question by itself? | No evidence | Search Console or site search shows the phrase |
| Does it need its own updates? | Changes with the group | Changes on its own schedule |
A question earns a standalone page when at least two checks point that way. The task page then carries a one-line answer and a link to the full page, so the reader is never sent away empty-handed.
Here is a small architecture for an imaginary appliance maker. The questions and placements are examples.
/help/ index of tasks
├── /help/setup/ task page, 12 answers inline
│ "How do I level the machine?" inline
│ "Which hose goes where?" inline
│ "How do I connect to Wi-Fi?" one line + link to standalone
├── /help/setup/connect-wifi/ standalone: steps differ by router type
├── /help/billing/ task page, 9 answers inline
├── /help/faults/ task page, 14 answers inline
│ "What does error E4 mean?" one line + link to standalone
└── /help/faults/error-e4/ standalone: diagnosis tree with 4 branches
Thirty-seven questions become three task pages and two standalone pages.
Navigating the library
A library is only useful if a question can be found in it.
- The index lists the tasks in plain words, with the questions under each visible or one click away.
- Each task page opens with a list of its questions, linked to their answers on the page.
- Each answer has a stable heading, so support staff and other pages can link to the exact question.
- Standalone pages link back to their task page.
- Site search covers the library, and its log is reviewed for questions people type that the library does not answer.
One expectation should be corrected before planning. Marking up questions and answers no longer produces an expanded result in Google. Google’s documentation changelog records that the FAQ rich result “will no longer appear in Google Search starting May 7, 2026” (Google Search documentation updates). Organize the library for readers. There is no display feature to be gained by splitting it.
Quality guardrails for thin entries
Apply these checks before any question is published as its own URL, and once a year to the existing ones.
- Could the answer be read aloud in one breath? Then it belongs inline.
- Does the page contain anything besides the answer and the site template? If not, it is thin whatever its length.
- Is the page a near copy of another with one noun changed? Google’s spam policies name, among examples of doorway abuse, “Creating substantially similar pages that are closer to search results than a clearly defined, browseable hierarchy” (Google’s spam policies for web search). A hundred question pages differing by a product name fit that description more closely than a helpful library does.
- Does the standalone page get visits or searches? One with neither for a year is merged back into its task page, with a redirect.
The guardrails apply in the other direction too. A task page that has grown to sixty answers is failing its readers in a different way, and should be split by sub-task.
This guide covers a library of questions. Whether one particular service page should carry a short set of questions and answers of its own is decided by different criteria, set out in the guide to when a commercial service page should include an on-page FAQ section.
The rule that keeps a library healthy: a question gets a URL when its answer is a resource, and a heading when its answer is a sentence.
Add your comment to this answer