Content People Actually Search For — SEO Intelligence on Tap

The Situation

The content calendar was full of ideas. Whether anyone was searching for those ideas was a separate question that took a few hours in Search Console and Ahrefs to answer, so it mostly went unanswered. Content got written because it seemed like a good idea, then quietly underperformed.

Meanwhile the site was already ranking for questions it had never properly answered. Nobody saw that either, because it takes three tools and a pivot table to notice.

What I Built

An SEO and content intelligence layer with Claude connected directly to the sources:

• Google Search Console for what people actually type and which pages show up for it.
• Ahrefs for keyword volume, competitors, and what is ranking around you.
• Google Analytics for what visitors do once they arrive.

Then a set of questions the team can ask any time, in plain language:

• What content is performing, and what has started slipping?
• Which queries are we getting impressions for without a page that really answers them?
• What should we write next, ranked by search demand and how close we already are to ranking?
• Which existing pages are one update away from page one?

The Results

• Content decisions are backed by search data by default, not by whoever had the strongest opinion in the meeting.
• Content gaps get pinpointed, down to the specific question people are asking that the site does not answer yet.
• Under-performing pages surface on their own, before someone notices a traffic chart three months later.
• The team still writes the content. The system says what to write and why. Humans decide whether it fits and make it good.

Why It Worked

This is the same pattern as every other automation on this site. The expensive part was never the thinking. It was pulling the data together so the thinking could happen. Automate the pulling, and the thinking starts happening every week.