“My pages are indexed and nobody comes”: our 208 impressions, and what they say about intent
Read from our own Search Console on 16 August 2026. Impressions went up tenfold in six days - and almost all of them came from people looking for something other than us.
In this article
There is a very specific disappointment, and it always lands at the same moment: your pages are finally in Google, the impressions curve takes off, and nothing happens. No enquiry, no quote, not one click. You assume you need to rank higher.
This morning we read our own Search Console. That is exactly what is happening to us - and ranking is not the problem.
What our numbers say, unrounded
Waseit has existed in Google since 16 July 2026. Its entire history fits on one screen: 5 clicks, 208 impressions, 2.4% CTR, average position 54.2. Asking for three months instead of one returns exactly the same figures - there is nothing before that.
Before 8 August: 2.6 impressions a day.
The site existed; nobody saw it.
From 8 August: 26 a day.
Our 682 city-and-trade pages had entered the index.
75% of all our impressions landed in the last six days.
A tenfold jump in a week. On paper, an indexing success. In practice, it produced no customers.
The misreading to defuse straight away
Our average position went from about 1 in early August to 54. A dashboard would tell you that you collapsed. That is wrong, and here is why.
On 5 August we had one impression, at position 1. On 10 August, thirty-four, at position 59. The average did not fall because something slipped: it fell because the denominator changed. You suddenly appear on hundreds of queries where you are 60th, and they dilute the handful where you were first.
An average position collapsing while impressions take off is not a penalty. It is arithmetic.
The real problem: they were looking for something else
Here are the queries bringing us the most people: “agence ia lille”, “agence ia nantes”, “audit consulting it lyon”, “ai search agency manchester”, “melhores imobiliárias curitiba”.
Look at them properly. Someone typing “agence ia lille” wants to hire an AI agency. Our page answers the question “which agencies do the AIs recommend in Lille?”. Those are two different people, with two different needs, who will never meet.
And this is where it gets instructive: even at position one, that visitor would not have become a customer. They want a supplier; we sell a measuring instrument. No amount of ranking work fixes that.
Where we are genuinely relevant
The two best non-branded positions in our whole account are long questions, written the way people speak:
- “*lista 10 chaveiros em porto alegre que precisam melhorar sua visibilidade no google*” - position 4
- “*i run a small bakery - how do i get chatgpt to recommend my business when people ask for bakeries near them*” - position 24
- “aeo vs seo”, “seo vs aeo” - positions 29 and 30
Nobody types that into a classic search box. These are conversational sentences - the kind of question people now put to an AI, and that Google increasingly receives too. On those queries we sit three times higher than on our market pages, with content we never optimised for them.
The silliest gap, and the most expensive
Our second query by volume is “free ai audit”: 12 impressions, position 97. We literally sell a free AI audit. Our page is called “Free AI visibility audit”. And we are on page ten.
The cause was not the title - that was already right. It was the page: a heading, one sentence, a form. Nothing to read, so nothing to rank. We fixed it the same day by writing what the audit actually does, what it cannot tell you, and which engines it queries. We will see in a few weeks whether that is enough - and we will say so either way.
What this changes for you
Do not judge a page by its position, judge it by the intent of whoever finds it.
A page ranking first on the wrong query is worth nothing; a page ranking 24th on the right one can be worth a customer.
Look at your queries before your rankings.
Search Console is free, and the Queries tab tells you in thirty seconds whether the people seeing you are looking for what you sell.
Mass-producing city-and-trade pages fills the index, not the order book.
Useful for discovery and internal linking, not an acquisition channel on its own. We have 682 of them and we are saying it out loud.
Write at least one page that answers the question the way it is asked out loud.
That is where search is moving, and it is where we rank best without having tried.
The limits, as always
5 clicks is far too few to conclude anything about a conversion rate, and we are not doing so. One month of data on a young site, with no ads and no emailing, describes a starting point, not a trend. And these figures stop on 13 August: nothing we have fixed since appears in them yet.
We are publishing anyway. A tool that measures other people's visibility while hiding its own would have a worse problem than 5 clicks.