Start with revenue, not traffic
On a site with thousands of pages, total traffic is the easiest number to watch and the least useful. A page type pulling 3,000 visits that converts into sales can be worth more than one pulling 40,000 that does not. That reframing changes priorities: instead of chasing rankings on high-volume keywords, the question becomes which page templates actually carry the money.
Scale rules out hand-editing every URL, but optimising everything at once just spreads the budget thin. The workable unit is the page type: product cards, categories, blog posts, service pages. Each has its own job in the funnel. Segment first, then decide which types earn and which only make noise.
Metrics at template level
Sitewide averages for CTR or position say little. What matters is measured per template: order or lead conversion, revenue per page, and the share of pages with no traffic at all. A page type that sells on modest traffic is a candidate for expansion across more URLs. A high-traffic type that does not convert is a candidate for rework or for being blocked from indexing.
Automation belongs on the routine work: templated meta tags, broken-link checks, sitemap updates. On content and internal linking it can do damage, since blanket insertion of links to popular products across every article dilutes link weight and pulls relevance off course. Automate the data collection and the prioritisation; leave content decisions to people.
Tasks always outnumber capacity. Give priority to what moves revenue fastest. A page type that already sells but carries weak meta tags comes first. Pages that have sat in the index for years with no traffic are candidates for removal or rework, and clearing them frees crawl budget for the pages that do work. Manage the money, not the pages.
