A published URL is not the same thing as an indexed SEO page. That distinction matters because teams often measure output in pages shipped, then assume the cost of those pages reflects search coverage. It doesn’t. The real unit is the page that becomes eligible for search traffic and actually enters the index. If a large share of pages never get indexed, your per-page economics are distorted by wasted production, weak information architecture, crawl constraints, or quality issues that never show up if you only look at publishing volume.
What you are actually measuring
Cost per indexed SEO page is the total cost of creating, publishing, and maintaining a page family divided by the number of pages from that family that are indexed over a defined period.
That sounds simple, but the boundary definition matters.
You are not measuring:
- cost per draft
- cost per published URL
- cost per page requested by the backlog
- cost per page with impressions
You are measuring the cost required to produce one indexed page that the search engine has accepted into its index, within a cohort and time window you can defend.
That means the numerator should include more than content production. If your team spends time on template work, data normalization, QA, internal linking, sitemap work, redirects, and cleanup for bad batches, those costs belong in the model when they are part of the path to indexation. Otherwise you end up pretending infrastructure is free.
This is the same reason a cost per published SEO page style metric is useful but incomplete on its own: publishing is a delivery step, not the end state. If you want the economics of indexation, you need to include the downstream work that makes pages eligible and maintainable.
Define the cohort before you do any math
Do not calculate this across the entire site unless the site is unusually uniform. The metric only becomes useful when you measure by page family, template, market, or publishing wave.
For example:
- location pages launched in Q2
- product-led landing pages generated from one template
- directory pages built from a specific data source
- refreshed pages that were republished after a content system change
Why cohorting matters:
- Different page types have different indexation behavior.
- Different templates create different maintenance burdens.
- Different data sources produce different rates of duplication or thinness.
- Different publishing windows have different crawl lag.
Averages across all of that hide the actual system problem. A site can appear efficient while one page family is quietly consuming a large share of labor and producing few indexed URLs.
If your operation is already tracking broader publishing metrics, it is worth aligning this measurement with your content operations efficiency framework and your SEO publishing velocity model. The first tells you how much waste exists in the workflow; the second tells you how quickly work moves. Cost per indexed page combines both with the indexation outcome.
Build the numerator from real operating costs
Start with a cost map. For each cohort, include the categories that actually contribute to producing indexed pages.
Typical inputs:
- strategy and planning time
- data preparation and normalization
- content writing or generation
- editorial review
- template design and engineering
- CMS or pipeline configuration
- QA and validation
- publishing operations
- internal linking work
- crawl/indexation debugging
- maintenance and refresh work attributable to that cohort
- rollback or cleanup work when pages fail
Do not force every team into the same cost model. If engineering only touched the template once but it supports thousands of pages, amortize that cost across the cohort you are measuring. If a content manager spent five hours fixing internal links for the page family, that belongs in the numerator.
A useful approach is to separate:
- fixed setup cost
- variable per-page cost
- recovery and maintenance cost
That separation makes the metric more actionable. If fixed cost is high but variable cost is low, the system may scale well after the initial build. If variable cost stays high, the workflow itself is expensive. If recovery cost keeps rising, the system is creating rework that is not visible in page counts.
Use indexed pages as the denominator, not published pages
This is where many teams make the measurement too flattering. Published pages are a leading indicator; indexed pages are the outcome.
A cohort with 1,000 published URLs and 250 indexed URLs should not be treated as a 1,000-page success. The meaningful denominator is 250, assuming those pages were intended to be indexable and had enough time to settle.
That said, you need a lag window. Indexation is not instant, and crawl schedules vary. Measure after a reasonable cohort age, such as 30, 60, or 90 days depending on the site and page type. The point is not to force a universal lag number; it is to avoid counting pages before they have had a fair chance to be crawled, processed, and indexed.
If your team is still diagnosing why pages are crawled but not indexed, or why a page family is under-indexed, this metric should be paired with a separate indexation analysis. That is where a focused indexation rate measurement is useful. Cost per indexed page tells you what the system costs. Indexation rate helps explain where the loss is happening.
A simple formula
At a cohort level:
Cost per indexed SEO page = total cohort cost / number of indexed pages in cohort
Example:
- Template work: $4,000
- Data prep: $2,000
- Content production: $6,000
- QA and publishing: $1,500
- Internal linking and maintenance: $1,500
- Total cohort cost: $15,000
- Indexed pages after the cohort matures: 300
- Cost per indexed SEO page: $50
If the same cohort produced 600 published URLs but only 300 were indexed, the published-page cost would look like $25 per page. That number is misleading if your actual goal is indexed coverage.
What makes the number rise
A high cost per indexed page usually reflects one or more of these failures:
- the team is producing more pages than the system can support
- the page family has weak differentiation, so indexation is suppressed or slow
- the data model creates duplicates or near-duplicates
- the template requires too much manual intervention
- internal linking is not helping discovery
- crawlable surface area is bloated with low-value URLs
- maintenance work keeps accumulating after launch
None of these is solved by publishing more pages. In fact, volume often makes the ratio worse because the numerator grows faster than the denominator.
This is why teams working on large-scale page families need to understand whether they have a page creation problem or an indexation problem. If the workflow itself is the bottleneck, use SEO publishing velocity. If the system is producing more waste than value, use content operations efficiency. If the pages themselves are structurally weak, the cost metric will tell you that the system is expensive even before traffic analysis gets involved.
Use the metric for decisions, not vanity reporting
The number is useful when it changes how the team operates. A few examples:
- Compare two template approaches before expanding a page family.
- Decide whether a data source is too noisy to justify automation.
- Determine whether engineering work on templates lowered the real cost of indexation.
- Evaluate whether a refresh program is cheaper than creating new pages.
- Identify cohorts where maintenance costs make the model uneconomic.
It is also useful for setting thresholds. If a new page family has a high projected cost per indexed page and a weak business value per indexed page, it may not deserve to exist. That is a better decision than publishing first and explaining the waste later.
The main trap is treating the metric as a one-time audit number. It should become part of a publishing system that gets remeasured after template changes, crawler behavior changes, data-source changes, and cleanup work. Otherwise you are just looking at a historical artifact.
Measured well, cost per indexed SEO page gives technical SEO teams a way to talk about page economics without collapsing everything into traffic, which arrives too late, or page count, which is too optimistic.