Template Entropy

Template entropy is a working SEO metric for how much unique information a page contains compared with repeated template elements. On large sites, that matters fast: if 10,000 location, product, or category URLs share 80% of their HTML and copy, Google often treats them as low-priority crawl targets.

This is not an official Google metric. Still useful. It gives SEO teams a concrete way to audit thin templated pages before they become an indexation problem in Google Search Console.

What template entropy actually measures

At a basic level, you are comparing page-specific content against boilerplate. That can include body copy, product specs, reviews, FAQs, internal links, images, structured data fields, and local data modules. A simple version is:

Unique page elements / total page elements

Some teams calculate this with text tokens only. Better teams include rendered HTML blocks, repeated components, and structured data properties. Screaming Frog exports, custom Python scripts, and BigQuery are common setups. Sitebulb works too, but Screaming Frog is usually faster for rough segmentation.

Why SEOs use it

The main use case is prioritization. If a faceted category set averages 18% unique copy and a city-page set averages 42%, you know where to fix first.

• Indexation: Low-entropy URL sets often correlate with high counts of Discovered – currently not indexed or Crawled – currently not indexed in GSC.
• Ranking ceiling: Pages with mostly repeated copy rarely compete for anything beyond ultra-specific queries.
• Internal equity waste: Linking heavily to weak template variants spreads authority across pages that add little distinct value.

Ahrefs and Semrush can help quantify whether these pages attract any non-brand keyword footprint. If 5,000 pages rank for fewer than 200 total keywords outside branded terms, the template is probably doing too little.

How to audit it in practice

1. Crawl a representative URL set in Screaming Frog with rendered HTML enabled.
2. Group pages by template type: product, location, category, comparison, programmatic pages.
3. Extract repeated blocks such as nav, footer, filters, trust badges, and generic intros.
4. Measure how much remains page-specific.
5. Cross-check against GSC indexation and log files.

A practical benchmark: pages under 25% to 30% unique content are usually risky unless demand is high and the page has strong external signals. Pages above 40% tend to perform better, especially when the unique content is useful rather than padded.

The caveat most teams miss

More entropy does not automatically mean better SEO. Adding 600 words of AI filler, a spun FAQ, and a stock image gallery can raise the score while making the page worse. Google's John Mueller has repeatedly said that uniqueness alone is not enough; the page still needs a reason to exist. That is the part weak audits skip.

Also, some low-entropy pages deserve indexation. Product variants, legal pages, and tightly structured inventory URLs can rank with limited unique copy if demand, links, and site architecture are strong. Use entropy as a diagnostic model, not a ranking factor.

The practical fix is simple: add modules that change the decision value of the page. Real reviews. Store-level inventory. Pricing tables. Original comparison data. Local proof. Not another generic paragraph Surfer SEO says should be there.