Programmatic SEO without publishing thin pages
Templated pages at scale work when each one answers a question a person actually asked. Here is the threshold we hold them to before publishing.
Key takeaways
- A templated page needs unique data, not unique adjectives.
- Hold every generated page to a minimum inventory and word threshold before it is indexable.
- Publish in cohorts and let indexation data decide whether the next cohort ships.
What separates useful programmatic pages from thin ones?
Page-specific data. A generated page earns indexation when it carries information the template cannot supply on its own — real inventory, prices, availability or comparisons — and when it clears a published minimum for content depth. Everything else is a duplicate with a different heading.
The failure mode is always the same
A team builds one good template, generates nine hundred pages from a spreadsheet of place names, and watches impressions arrive for six weeks before the whole set quietly stops being served.
The pages were not penalised for being templated. They were filtered because 880 of them contained no information that the other 19 did not already provide.
The gate every generated page passes before it is indexable
| Check | Threshold | If it fails |
|---|---|---|
| Page-specific words | 300+ | Noindex until content exists |
| Inventory or data rows | 5+ | Redirect to the parent hub |
| Unique title and description | Required | Blocked at build |
| Internal links in and out | 3+ each | Added to the hub backlog |
Publish in cohorts
We never publish the full set. The first cohort is fifty pages across the range of data quality. Four weeks later, indexation and impression data tell us whether the threshold is set correctly.
If half the cohort is not indexed, the threshold was wrong and we fix it before generating the other eight hundred. That single discipline has saved more programmatic projects than any technical fix.
Give the hub a real job
Generated pages need a parent that is worth reading on its own — explaining the dataset, comparing the options and linking to the pages worth visiting first.
Without it, the set has no internal authority source and every page competes alone against established sites. With it, the hub carries the topic and the leaves carry the long tail.
FAQFAQ
Frequently asked questions
About the author
Aisha runs acquisition across search, paid and lifecycle. She writes about the work that connects a channel report to pipeline — crawl and indexation, topical authority, incrementality testing and attribution that survives a privacy-first web.
- Technical SEO
- Topical authority
- Performance media
- Marketing attribution
