We scan new podcasts and send you the top 5 insights daily.
After Claude's audit found missing SEO descriptions, crawlability issues, and long titles, James McAulay said a growth person could hand the report to an agent inside the repo. The agent can test and refine until the issues are solved, replacing the weeks he once spent waiting on agencies and developers.
Traditional website optimization focused on human experience and SEO for search bots. A third pillar is now essential: optimizing for AI advisory tools and recommendation engines through structured data like product feeds and APIs.
The web is entering a new optimization era. It started with SEO (can a search engine understand a page?), evolved to AEO (can an AI cite a page?), and is now moving to WebMCP (can an agent complete a job on a page?). This shift prioritizes machine actionability over simple readability.
The first step to influencing AI is ensuring your website is technically sound for LLMs to crawl and index. This revives the importance of technical audits, log file analysis, and tools like Screaming Frog to identify and remove barriers preventing AI crawlers from accessing your content.
Technical audits, like reviewing all Salesforce flows, are laborious. By feeding Salesforce metadata into an AI like Claude, teams can automatically generate documentation and analysis of each automation. This drastically cuts manual review time, reducing a multi-day task to just a few hours and accelerating project roadmapping.
You can automate search engine optimization by creating an AI loop that runs on a schedule (e.g., monthly). The agent connects to Google Search Console, analyzes ranking data, experiments with on-site changes, and learns from the results to create a continuous improvement cycle.
James McAulay remixed Corey Haynes's SEO audit and wired in Firecrawl plus DataForSEO. He says Claude can research any term for cents and assemble a keyword universe, bypassing clunky SEMrush workflows. The agent can analyze his site and competitors, then point him at the keywords worth attacking.
Use Playwright to give Claude Code control over a browser for testing. The AI can run tests, visually identify bugs, and then immediately access the codebase to fix the issue and re-validate. This creates a powerful, automated QA and debugging loop.
Optimizely runs AI agents for SEO (EAT) and brand voice directly within their content platform's workflow. The agent completes its task like a human team member, leaving comments on the content record. This avoids forcing marketers into a separate tool and seamlessly integrates AI into daily work.
The lowest-hanging fruit for AEO is often technical site optimization. If LLMs cannot easily crawl and understand your site's structure via things like schema.org markup, meta descriptions, and alt text, even the best content may not get included in answers.
The traditionally lengthy programmatic SEO process can now be done in a single session. AI agents can perform live keyword research via APIs, generate thousands of optimized pages, and prepare them for deployment, a task that once took months.