10 Claude Prompts for a Repeatable Weekly SEO Workflow

A good prompt is not magic. It is a compact brief that gives the model evidence, scope, evaluation criteria and a useful output format. The ten prompts below work best as a sequence. Start with one priority page, export the relevant data and review every recommendation before implementation.

Prompt: Audit this page for retrieval and citation readiness. Review crawlability, entity clarity, first-hand evidence, authorship, source quality, structured data and answer clarity. Separate confirmed issues from assumptions, then prioritise fixes by impact and effort.

Use the result with the E-E-A-T Checker and the AI citations guide. Avoid asking for unsupported guarantees such as a predicted citation rate.

Prompt: Compare this page with its Search Console queries and the current result set. Group missing needs into informational, evaluative, transactional and post-purchase intent. Recommend a new section or page only where the user task is materially different.

Prompt: Build a topic map for [subject] around the decisions our audience makes. Identify a pillar, supporting pages, unique evidence needed, logical internal links and a publishing order. Remove topics that do not support our expertise or commercial goals.

Prompt: Review the page for missing concepts, entities and relationships that a knowledgeable reader would expect. Suggest precise additions with a reason and source requirement. Do not add synonyms merely to increase keyword frequency.

Prompt: Find sections that answer a clear question. Rewrite the opening two sentences so they answer directly, preserve necessary conditions and remain accurate when quoted alone. Then list the evidence required to support each answer.

Google says useful, original, people-first content and sound technical SEO remain the basis for its generative features. Read the official guidance before adopting any tactic presented as an AI-only shortcut.

Prompt: Identify where this page makes an important claim without showing experience, expertise, authority or trust. Recommend concrete evidence such as methodology, author credentials, original data, examples, dates, limitations and reputable sources. Do not invent any credential, client or result.

Prompt: Using this URL inventory, propose contextual internal links for the page. Give the source URL, destination URL, natural anchor and placement rationale. Prioritise user navigation and topic relationships. Flag orphan pages and competing destinations.

Validate suggestions with the Internal Link Checker and review the keyword cannibalisation guide when several pages target the same need.

Prompt: Review this crawl, rendering test and Search Console evidence. Classify issues under discovery, indexing, canonicalisation, rendering, structured data and page experience. Rank each by affected URLs, likely impact, confidence and implementation effort.

Prompt: Compare the current page with its historic queries, recent competitors and the latest reliable sources. Identify facts that need updating, sections losing demand, new user questions and internal links that should change. Preserve sections still earning qualified traffic.

Prompt: Assess whether this page can serve traditional search, AI Overviews and cited assistant answers without splitting into duplicate versions. Recommend improvements to structure, evidence, entity consistency and answer quality. Label platform-specific assumptions and link them to a source.

1. Monday: select one commercially important page from the SEO Opportunity Finder.
2. Tuesday: run intent, topic and semantic reviews.
3. Wednesday: review technical and internal-link evidence.
4. Thursday: implement the smallest high-confidence changes.
5. Friday: record the change, request indexing where appropriate and define the measurement window.