We inspected all 86 of our own URLs
This is a report on being crawled - currently not indexed, written from our own site's data rather than from theory. Getting a page stored by Google is the floor under everything else, including the practice the industry calls AEO or GEO — we call it Answer Optimization (AO). A page Google never kept cannot be quoted by anything.
Our site is built in matched pairs. Every English page has a Spanish twin: 43 and 43, 86 pages total. That was a marketing decision, not an experiment. But it accidentally produced a clean one, because both halves share a template, a domain, a sitemap and a publishing date range. This morning we ran all 86 URLs through the Search Console API and read the verdict on every one.
English: 36 of 43 pages indexed. Spanish: 23 of 43. Across the whole site, 59 of 86, which is a little over two thirds.
That much we half expected. We have known since our first full indexing sweep on August 15 that the Spanish side was lagging. What we did not expect was where the gap lives.
Keeping them. This is the whole story, and it surprised us.
Google Search Console sorts a missing page into one of three buckets. The page can be unknown, meaning Google has never heard of it. It can be discovered but never crawled, meaning Google knows the URL exists and has not bothered to go get it. Or it can be crawled and then declined, which shows up as crawled - currently not indexed. The first two are reach problems. The third is a judgment.
| Status on Aug 20, 2026 | English | Spanish |
|---|---|---|
| Indexed | 36 | 23 |
| Crawled, then declined | 0 | 12 |
| Never crawled (discovered or unknown) | 7 | 8 |
| Total published | 43 | 43 |
Look at the bottom two rows. Pages Google never went and fetched: seven English, eight Spanish. Basically identical. Google's crawler reached our Spanish pages exactly as readily as our English ones. Discovery is not the problem, and our sitemap is not broken for one language.
Now look at the middle row. Twelve Spanish pages were fetched and then turned away. Zero English pages were. Not one.
Put it as a keep rate and it gets starker. Of the English pages Google actually crawled, it kept 36 of 36. Of the Spanish pages it actually crawled, it kept 23 of 35. Same template, same domain, same sitemap, same week.
Google found our Spanish pages just as easily as the English ones. It just didn't keep them. The entire gap between 36 and 23 sits in a decision Google makes after it has already read the page.
We don't know, and this is the part where most articles would hand you a confident reason. We measured what happened. We cannot prove why.
Our three best guesses, in the order we'd bet on them: a translated page on a young site may read as a near-duplicate of something Google already stored; almost nothing on the open web links to our Spanish pages, so they arrive with no outside support; and indexing budget is finite, so the version Google thinks fewer people will need loses. Google does not publish a rule that says any of that. If someone tells you they know exactly why your page was declined, they are guessing too — they're just not telling you.
One more honest limit: this is one site, ours, with 86 pages and not much authority yet. It is a real measurement, not a law of the internet.
Two things, and both are worse than the language gap.
First, the article we published three days ago about how Siri picks businesses is still, as of this morning, unknown to Google. Not declined. Not queued. Unknown, in both English and Spanish. We wrote a piece about being findable and Google does not yet know it exists.
Second, and this one stings: an article we published on July 19 has never been crawled. Not once, in 32 days. It sits in our sitemap. It is linked from our homepage footer and from the bottom of other articles Google visits regularly. None of that was enough. A sitemap entry is a suggestion, and we had quietly been treating it as an instruction.
No, and this is the useful part. On August 17 we had 17 pages sitting in the declined bucket. This morning, five of them are indexed. Google re-crawled them on August 18 and changed its mind. Four of those five were Spanish.
The word doing the work in that status is currently. It is a queue position, not a verdict. The wrong move is to delete the page or assume it's cursed. The right move is to give Google a reason to come back: strengthen the page, link to it from somewhere it already crawls, and ask for indexing directly in Search Console.
Open Search Console and inspect your own URLs one at a time. Most owners have never done it and assume that a live page is a found page. Those are different things, and the gap between them is where a lot of invisible businesses live.
Then read which bucket you're in, because the fix is different for each. Never crawled means work on links and requesting indexing. Crawled and declined means the page needs to be more worth keeping than the thing Google already has. And if you publish in two languages, check both halves separately. Nobody would have caught our 12 to 0 split by looking at a single site-wide number.
Because 23 of 43 is our real number, and we sell honesty about numbers. We run a tool that grades other people's visibility. A tool like that has no business hiding its own scoreboard, and the failures teach more than the wins do. Our last few reports from these logs were a failing self-audit, a schema error inside our own schema checker, and a warning we deliberately left in place. This one is a third of our own site missing from Google.
We will re-run this exact sweep next week and publish whether the 12 moved.
Want to know if AI engines and Google can actually see your site? The scan is free, takes about a minute, and shows every check, including the ones you fail.
It means Google's crawler successfully fetched the page and then decided not to put it in the search index. It is not a technical error and not a penalty. The page is reachable, the server answered, nothing is blocking it. Google simply judged it not worth storing, at least for now. Because it is a judgment rather than a fault, there is no error message to fix. On our own site this status hit 12 Spanish pages and zero English pages on the same day.
No. On August 17, 2026 our site had 17 pages sitting in that status. Three days later, five of them were indexed after Google re-crawled them on August 18. Four of those five were Spanish. The word currently in the status is doing real work. Treat it as a queue position that can change, not a final verdict, and keep the page live and linked instead of deleting it.
We measured that this happens on our site but we cannot prove why. Our leading guesses are that translated pages on a young site read as near-duplicates of pages Google already stored, that almost nobody links to our Spanish pages from anywhere else on the web, and that Google spends limited indexing budget on the version it thinks fewer people will need. None of that is confirmed by Google. What we can say from our own data is that discovery was not the problem. Google reached our Spanish pages as readily as the English ones and then kept fewer of them.
No. One article on our site had been in our sitemap since July 19, 2026 and was cross-linked from the homepage and from other articles. Thirty-two days later Google had still never crawled it once. A sitemap is a suggestion, not an instruction. Internal links from pages Google already visits often move a page faster than a sitemap entry does, and manually requesting indexing in Search Console is the fastest lever we have found.