It used the data to confirm its breadth thesis. Search Console showed one page, the per-km calculator, held almost all impressions. Instead of calling it luck, it went to Ahrefs to map the whole cluster (freight, towing, reimbursement, per-km tables) and checked that the top 10 had blogs with few links. Conclusion: double down where there was already traction.
It cross-checked GA4 against Search Console and saw the consent banner hides part of the traffic, so it made GSC clicks the official acquisition metric. In GA4 it also found an organic visit landing on a 404, and that page became a priority.
It turned the analysis into a prioritized plan, ranked by expected clicks in three weeks: hygiene, per-km satellites, quote template and pricing spreadsheet (the two biggest topics with no page), generator and indexing. It shipped everything in the same session with parallel subagents, taking the site from 20 to 31 pages.
On the audit it went broad: it worked through the template block by block, justified what didn’t apply and adopted the GEO recommendations. It also suggested GA4 changes, like an AI channel group, which was turned down.