A page cannot lose a ranking it never had. Most "my site doesn't rank" threads - and most paid audits - start at keywords and backlinks. That skips the sentence this whole post stands on: was the page ever in Google's index at all? Google's own documentation estimates that for large sites, not every crawled page gets indexed (Google Search Central, July 2026) - and small sites bleed the same way, one silent URL at a time.
Here is why I trust that sentence more than any SEO tool I have used: it happened to us this week.
This week, our own blog page quietly went from two posts to one. No error page. No warning email. The listing just rendered fewer cards, and if our editor had not been reading the page, the site would still be showing one post right now - sitemap included. The post was not deleted by an algorithm or a hacker; a single routine dashboard action removed it, and the site obeyed. What struck me was not the bug. It was the silence. The web does not notify you when your pages disappear. Search engines behave the same way: pages fall out of indexes every day - blocked, duplicated, outclassed - and nothing pings you. You find out by looking. Most people never look.
The funnel your page has to survive
Every URL lives or dies on a three-stage funnel:
- Discovered - Google learned the URL exists (a link, a sitemap, a submission).
- Crawled - Googlebot fetched it.
- Indexed - Google judged it worth storing and serving.
"Ranking" is stage four, and it is the stage everybody argues about. But each stage can silently drop your page, and the fixes are completely different. Optimizing keywords on a page that never passed stage three is repainting a house the mailman cannot find. Google's crawl documentation puts it plainly: after crawling, every page must be evaluated - and low-value, duplicate, or near-duplicate pages are where crawl budget goes to die (Google Search Central, 2026).
The five checks, in order
1. Ask Google directly (2 minutes)
site:yourdomain.com/your-page in Google. Result appears? You are indexed - your problem is stage four, and this post just saved you from the wrong rabbit hole. No result? Move to check two. Also try the cache checker to see what Google last stored: an old snapshot tells you the page was indexed once and then dropped, which is a different disease than never-indexed.
2. URL Inspection (5 minutes)
Search Console's URL Inspection tool is the only authority: it reports exactly which funnel stage failed - "URL is not on Google", "Crawled - currently not indexed", "Discovered - currently not indexed", "Excluded by noindex". Each phrase maps to a different fix. Write down the exact phrase; guessing between them is how weeks get wasted.
3. Blockers audit (15 minutes)
Three settings remove pages without any error: a noindex meta tag (the page renders fine to humans - that is why it survives so long), a robots.txt rule (which blocks crawling entirely - test your rules here), and a canonical tag pointing somewhere else (which merges your page into another URL - validate your canonicals). Then check the plumbing: a redirect chain can quietly hand your page's value to a URL you forgot existed. All three blockers are invisible on screen. All three are one tag deep. Generate a clean robots.txt if the audit finds your rules grew teeth by accident.
4. The value verdict (20 minutes)
"Crawled - currently not indexed" is Google's polite way of saying: fetched it, not impressed. The documented causes are low value, duplication, and near-duplication - thin pages, tag archives, near-identical service pages for ten neighboring towns. The test that works: is this page genuinely useful to a human who arrives from search, or does it exist to exist? Consolidate the duplicates, deepen the thin ones, and point internal links at the survivors (the link analyzer shows which of your pages are orphans - pages with no internal links are pages Google learns about last and trusts least).
5. The sitemap reality check (10 minutes)
Your sitemap is your list of URLs you consider canonical - submitted pages get discovered faster and their indexation rate becomes measurable in Search Console (submitted vs indexed, per sitemap). Validate it (sitemap validator), keep it to real, indexable, canonical URLs (generate a clean one if yours grew by accident), and watch the submitted-to-indexed ratio over time. That ratio is your site's indexation vital sign - one number, trending, honest.
Here I'll be blunt
My take: "publish more content" is the worst advice in SEO while indexation is broken, and it is the most repeated advice in the industry. Because: new pages on an unhealthy domain enter the same funnel that just dropped the last batch - you are scaling the leak, not fixing it, and Ahrefs' landmark crawl study found 90.63% of pages get zero Google traffic (Ahrefs, 2020 - old number, still the cleanest map of how much of the web is invisible). Volume multiplies whatever exists. If indexation is broken, volume multiplies a graveyard. What would change my mind: evidence that thin-page publishing at scale still achieves durable indexation for small domains - e.g. a documented case of a small site sustaining 90%+ indexation across hundreds of near-duplicate pages through a core update cycle. I have never seen one.
The Devil's Advocate
The steelman: "Big sites win by publishing everything and letting Google sort it out - indexation follows authority, so the real advice is to grow authority, not to audit tags." Correct about the big sites, and misleading for everyone else. News domains and marketplaces win volume games because their domain authority makes Google invest crawl budget in them by default. A small site running the same strategy is pouring water into a funnel with a hole in it. The order of operations is not negotiable: fix the funnel on the traffic you already have, then scale. Authority growth and indexation hygiene are not rivals; the first multiplies only what the second protects.
The 2am Version
Before you chase rankings, prove the page is in the race. site: search, URL Inspection, blockers audit (noindex, robots, canonical, redirects), value verdict, sitemap ratio. Google will not email you when a page drops - it dropped two of our own blog posts' worth of visibility this week with zero notifications, and the only reason we caught it is that a human looked. Look. Then publish more.
Analyses - the RankForge frame
Data lens. The Ahrefs 90.63% is a 2020 study - outside our preferred recency window, cited because no newer study has replaced it as the landmark measurement, and the structural fact it maps (most of the web gets no search traffic) has no reason to have inverted. Its limitation: it sampled pages, not intent - "zero traffic" includes pages nobody wanted indexed. The 2026 Google crawl doc is current but written for large sites; small sites extrapolate from it at their own risk, because crawl budget is barely a constraint under a few thousand URLs - value judgments are.
Risk lens. The expensive mistake this post prevents is misdiagnosis by deletion: discovering weak traffic, assuming the page is worthless, deleting or rewriting pages that were the only indexed URLs on the domain. The loss mechanism is precise - indexed URLs carry history, links, and rankings; replacing them resets all three to zero. Diagnose the funnel stage first. Rewrite second, if at all.
Ethics lens. Indexation anxiety is a retail opportunity for audit-sellers, and the fear line "Google is hiding your site" is usually false. We show the diagnosis path free, with no signup wall, and we say plainly when the answer is boring: most indexation failures are one tag, one redirect, or one thin page. Education, not scaremongering; this post promises no rankings - it promises you will know which stage failed, which is the part you can actually control.
RF stance. Audit the funnel before scaling content; keep a sitemap ratio you actually watch. Because the funnel is checkable in under an hour with free tools, silence is the default failure mode of the entire web, and scaling a broken funnel is the only SEO mistake that gets more expensive the harder you work.
Change-my-mind. A documented small-domain case sustaining high indexation across hundreds of near-duplicate pages through a core update cycle - publish that case study with its Search Console data, and the "fix the funnel before you scale" order of operations gets a genuine exception.
Sources
- Google Search Central (July 2026) - "Crawl Budget Management": not every crawled page is indexed; evaluation, consolidation and value judgments between crawl and index. developers.google.com
- Google Search Central Help Community (Feb 2026) - official community answer on "crawled but not indexed": low value, duplicate, or too-similar pages as the documented cause. support.google.com
- Ahrefs (2020) - "Search Traffic Study": 90.63% of pages in a ~1B-page crawl get zero Google search traffic. Landmark study, cited despite age as the standard measurement of web invisibility. ahrefs.com



