Google Indexed 1 of Our 43 Pages. DuckDuckGo Sent More Visitors.
Search Console has 43 of our URLs on record. 1 is indexed. Meanwhile DuckDuckGo has sent 38 visits to Google's 10.
Contents
OverviewThe short versionWhat we measured, and how43 URLs checked in Search Console. 1 indexed.Every URL, and what Google says about itOur first search visitors came from DuckDuckGo, not GoogleWhere the visits actually came fromThe traffic is mostly not humanWhat we take from thisBottom lineFAQSourcesAlmost every article about getting indexed is written by someone whose site is already indexed. This one is not. This site has been visible to search since 2026-07-03 — 43 days, about 6 weeks. Google Search Console has 43 of our URLs on record: 1 is in the index.
We are publishing the numbers anyway, because the interesting part is not the failure — it is what happened around it. The pages were crawled. Search visitors did arrive. They just did not arrive from where the advice says they will.
The short version
Search Console, 43 URLs checked: 1 indexed, 17 crawled and not indexed, 25 still unknown to Google. Referrers: DuckDuckGo 38 visits vs Google 10. Bot requests outnumbered human pageviews by about 15.2× last week. Crawl access and search presence are separate problems, and only one of them has a technical fix.
What we measured, and how
The first request in our access log is dated 2026-06-28 — infrastructure and setup traffic, before any article existed. Content went up in the days that followed, and Search Console has data for this property from 2026-07-03 — 43 days ago. It is a small English-language technical site with no backlink campaign, no social push, and no paid promotion. That makes it a clean specimen for one question: what happens to a new site's pages after search engines find them?
Two sources, both first-party. Indexing status comes from Google Search Console for this property, covering 2026-07-03 to 2026-08-14. Traffic comes from our own nginx access logs, processed by the same bot-exclusion rules we use for every number on this site — a user-agent filter, a probe filter, and a behavioural filter, applied identically to every request. Numbers below were generated on 2026-08-15 08:45.
One caveat stated up front: these are small numbers. A few dozen search visits is not a sample you should generalise from. What it is good for is showing the shape of a new site's first months, which is rarely published because it is not flattering.
43 URLs checked in Search Console. 1 indexed.
Search Console has 43 URLs on record for this property — every page we have published, plus a few that have since been retired. 1 is indexed. 17 were crawled and then not indexed. For the remaining 25, Google's own answer is “URL is unknown to Google” — it has not reached them yet. They are all in a submitted sitemap; submitting a sitemap asks Google to discover URLs, it does not oblige it to, and working through one takes as long as it takes.
The direction of travel is the interesting part. In the previous snapshot the crawled-but-not-indexed bucket held 1 URL; it now holds 17. Google is working through the site steadily — reading pages, then declining to keep them. Crawling is not the bottleneck. Judgement is.
This is worth separating from the usual advice, because the usual advice is about the wrong step. Sitemaps, internal links, clean HTML and fast responses all buy you crawling. We have crawling. What we do not have is a verdict in our favour, and there is no technical lever for that.
Every URL, and what Google says about it
| Search Console status | URLs |
|---|---|
| URL is unknown to Google | 25 |
| Crawled - currently not indexed | 17 |
| Submitted and indexed | 1 |
Every URL Search Console currently reports for this property, as of 2026-08-14. Statuses are Google's own wording.
Our first search visitors came from DuckDuckGo, not Google
Here is the result we did not expect. Across the life of the site, DuckDuckGo has sent 38 visits, against 10 from Google. That is roughly 3.8×.
The likely mechanism is visible in the table below: DuckDuckGo and several of the other referrers draw wholly or partly on Bing's index rather than Google's. We cannot see inside either index, so we are not claiming Bing indexed more of this site — only that these visits arrived through a different index with a different threshold, while Google's verdict on the same pages was the one shown above.
It also means the standard framing — get indexed by Google, then traffic follows — describes only one of several doors. The others are smaller, but on a new site they open first.
Where the visits actually came from
| Referrer | Where its results come from | Visits sent |
|---|---|---|
| DuckDuckGo | Bing index + own crawler | 38 |
| Google index | 10 | |
| Bing | Bing index | 9 |
| Kagi | Blended, incl. Bing | 7 |
| Brave Search | Own index | 7 |
| Ecosia | Mixed sources | 2 |
| ChatGPT | Answer engine | 2 |
| Perplexity | Answer engine | 2 |
Referrer hosts recorded in our own access logs across the life of the site, after the same bot-exclusion rules used everywhere on this dashboard. 463 visits arrived with no referrer at all and 6 came from hosts that are not search engines; neither is counted above.
The traffic is mostly not human
Over the last seven days this site served 270 human pageviews and 4,116 requests we classified as bots — about 15.2× more machine than human. Lifetime human pageviews stand at 809.
We mention it because it is the single easiest way to fool yourself with analytics. Raw log lines, or any counter with a weak bot filter, will show a new site growing nicely while the humans stay flat. We learned this the hard way: an earlier count on this site put one article at 198 views; the same article measured with the proper filter was 88. Same log, same day, different filter.
If your traffic graph is going up and your search referrers are not, check what is in the number before you celebrate it.
What we take from this
Crawled-but-not-indexed is the normal state of a new site, not a bug to fix. Google documents that it does not index every page it crawls. Our numbers are what that policy looks like from underneath.
Do not measure progress in crawl volume. This site is crawled constantly and indexed once. The two numbers are barely related, and only one of them sends visitors.
Check the other engines. Most of our search visits came from engines other than Google. That is a channel most new-site advice does not mention, and it costs nothing to verify in your own referrer log.
Publish the shape of your own data. Everything above came from a log file and a free Search Console property. The reason this data is rare is not that it is hard to collect — it is that it looks like failure, so people keep it private.
If you launched a site recently and Search Console is showing you a wall of “Crawled - currently not indexed”, this is what that looks like on another site with the same problem. It is not a misconfiguration, and no sitemap tweak resolves it.
What we would do differently with hindsight is spend less time on crawl plumbing — which was working from week one — and more time checking referrers from engines other than Google, because those were sending visitors while we were staring at the indexing report.
FAQ
Why would Google crawl a page and then not index it?
Google states that crawling and indexing are separate steps and that it does not index every page it crawls. 'Crawled - currently not indexed' means the page was fetched and assessed, and Google chose not to store it. It can change later without any action from you.
Does submitting a URL for indexing help?
It can get a page crawled sooner. It does not commit Google to indexing it — the URL Inspection tool's own wording is that submission requests a crawl, not an index entry. On this site, requesting indexing reliably produced a crawl and did not produce an index entry.
Is DuckDuckGo traffic worth anything?
It is real search traffic with real intent, and on this site it arrived earlier than Google's. Its volume is much smaller than Google's would be at scale, so treat it as an early signal that your content is findable, not as a substitute for Google.
How do you separate bots from humans in these counts?
Three layers, applied to every log line the same way: a user-agent list, a known-probe path filter, and a behavioural check on request patterns. No filter is perfect — a determined crawler with a browser user agent will be counted as human — so read these as best-effort, not exact.