brightonSEO · April 202620 min · Intermediate
Gus Pelogia, Indeed
Gus explains MCP servers, the standard that lets an AI assistant call other tools, then shows five ways he uses them with SEO data providers such as DataForSEO, Ahrefs, Semrush and SE Ranking: sorting keywords into groups, checking fresh results in other cities, pulling search features by location, chaining tools together, and measuring how many ranking keywords show AI Overviews. He recommends Claude Code over the desktop app for bigger jobs and warns about usage limits and API costs.
Why watch A practical starting point if you want an AI assistant to do the legwork across your SEO tools. You'll need paid plans with API or MCP access.
Tech SEO Connect · December 202526 min · Advanced
Tyler Gargula
Tyler sets out a step-by-step framework for turning SEO data into decisions: define a clear goal, pull and clean the data, look at distributions rather than single averages, score priorities, and share findings people can act on. Examples include finding which page attributes went with 'crawled, currently not indexed', and measuring content decay with trend slopes instead of comparing two dates.
Why watch Strong on avoiding bias and cherry-picking, which is where a lot of SEO reporting goes wrong.
Tech SEO Connect · December 202520 min · Advanced
Sam Torres, The Gray Dot Company
Sam argues that language models are the wrong tool for numbers and makes the case for classic machine learning on Search Console and GA4 data instead. She shows a workflow using Google Colab notebooks, with an AI assistant writing the code, and compares three ways to find unusual days: isolation forest, local outlier factor, and a forecast-based approach using Prophet that copes with seasonality and holidays.
Why watch A practical way into anomaly detection if you can follow a notebook but don't want to build one from scratch. She gives her notebooks away.
brightonSEO · April 202523 min · Beginner
Oliver Brett, Screaming Frog
A funny, practical talk on the reporting traps Oliver has seen in ten years at Screaming Frog: mixing up traffic metrics, the many meanings of average position, authority scores that move when a tool's index grows, and link and rank tools that only see part of the picture. One example shows brand and non-brand clicks in Search Console adding up to much less than the total, because some searches are hidden for privacy.
Why watch Good for newer SEOs and anyone who writes a monthly report.