Published, crawled, indexed: the three numbers to compare first
When a site with good content does not rank, the cause is usually mechanical. Compare these three numbers before commissioning another word.
A team publishes consistently for a year and the traffic line stays flat. The instinct is to publish more. Usually the right move is to find out whether anything published so far can be found at all.
Three numbers, in this order.
1. Published
How many pages you believe you have. Take it from your own content system, not from a crawler.
2. Crawled
How many pages a search engine has actually fetched. If this is well below the first number, something is blocking access: a robots rule, a redirect chain that gives up, pages linked from nowhere, an address structure that generates far more combinations than you have content, or responses slow enough that crawling gets throttled.
3. Indexed
How many pages are eligible to appear in results. A page can be crawled and still not indexed, and that gap is the most useful signal available to you. It usually means the page is a near duplicate of another page, it is thin enough that the engine decided it was not worth storing, or a canonical tag is pointing somewhere you did not intend.
Reading the gaps
- Published far above crawled. An access problem. Check robots directives, internal linking, the sitemap and your status codes.
- Crawled far above indexed. A quality or duplication problem. Look for parameter variants competing with each other, near-identical product descriptions, and canonical tags pointing at the wrong page.
- All three close together and still no traffic. Now it is a relevance and competition question, and now content work is the right answer.
That last case is the only one where publishing more helps. In the other two, every new page joins the same queue and meets the same blocker.
The part worth insisting on
Ask for these three numbers before you approve an SEO retainer, and ask for them again each month. Anyone who cannot produce them is not doing technical work. And when a report shows only what improved, ask what was tried and abandoned. That is the part that tells you somebody is actually testing rather than reporting the weather.
Founder & Lead Penetration Tester at Nextralix. Manual web and API penetration testing with a focus on authentication, authorization, account lifecycles and business logic.