As of 13 August 2026, AI can only partly improve your help centre SEO.

What to actually do

1. Hand it to a person

   The route this page recommends

   A person who owns the outcome does this end to end, worth it when the failure is dear.

2. Use a tool built for this

   Second choice

3. Do it yourself

   The distant third

   A chat interface, power-user skill, and roughly 2 hours until you can act on the result.

   How to actually do it

   1. Open your help-centre CMS and export the current article titles, URLs and full text into a document, keeping each URL attached to its article.
   2. Export the relevant Google Search Console queries and pages, plus internal site-search terms and unresolved support themes if you have them.
   3. Gather your current product policies, feature descriptions and approved wording so factual claims in the articles can be checked against the source of truth.
   4. Paste the business context, help-centre content, customer questions and search data into the prompt, then ask the chatbot to produce the audit and prioritised change table.
   5. Compare every proposed factual change with your approved product and policy documents, and remove recommendations based on invented or unsupported search evidence.
   6. Check proposed titles, headings, URLs and internal links against your CMS, existing URLs and customer intent before asking a colleague with SEO experience to review the priorities.
   7. Publish a small group of approved changes, record the old and new versions, and monitor search queries, clicks, article visits and support contacts before applying the next group.

   Prompt

   Act as an SEO editor for a UK help centre. I will provide help-centre articles, their URLs, a list of important customer questions, and any available Google Search Console or site-search data.
   
   Your task is to identify practical improvements that make the help centre easier to find and easier to use. Do not invent product features, policies, customer questions, search data, rankings, traffic figures or Google rules. Treat supplied search data as evidence, and label any inference as an inference.
   
   For each recommended change, provide:
   1. The article or URL affected.
   2. The customer intent it serves.
   3. The evidence from the supplied content or data.
   4. A proposed page title, H1, meta description and URL only where a change is justified.
   5. Missing questions or sections to add.
   6. Internal-linking opportunities using only supplied URLs.
   7. Duplicate, overlapping or potentially competing articles.
   8. Any factual statement that must be confirmed by the business before publication.
   9. A priority of high, medium or low with a short reason.
   
   Preserve the meaning of the support content and use plain UK English. Do not make medical, legal or financial claims. Do not recommend hiding content, keyword stuffing, misleading titles or changes based only on generic SEO advice. Separate content recommendations from technical recommendations, and say when a technical recommendation needs a web developer or SEO specialist.
   
   At the end, give me:
   - a table of the highest-priority changes,
   - a publishing checklist,
   - tests to run before and after publishing,
   - and the limits of this analysis based on the information I supplied.
   
   Here is the business context:
   [BUSINESS CONTEXT]
   
   Here are the help-centre articles and URLs:
   [HELP CENTRE CONTENT]
   
   Here are the customer questions or support themes:
   [CUSTOMER QUESTIONS]
   
   Here is the available Search Console or site-search data:
   [SEARCH DATA]

   Open it prefilled in ChatGPT or Claude, or copy it into Gemini, which takes no prefill link.