Ask ten agencies how long SEO takes and you will get “three to six months” ten times. It is not wrong. It is just not useful. Nobody shows you what those months actually look like, so you have no way to tell whether your own site is on track or quietly failing.
So I tracked mine.
I launched this website in August 2026. Brand-new domain, no history, no backlinks, one real client. Then I recorded what Google did with it every single day for six weeks. Below are the real numbers, including the ones that make me look bad.
The short answer
Six weeks in: 10,321 impressions, 23 clicks.
That is a click-through rate of 0.22%. If you were expecting a graph that goes up and to the right, this is not that.
But the useful part is not the total. It is the shape.
Week by week
An impression means your page appeared somewhere in Google’s results — even on page 9, where nobody looks. A click means somebody actually came to your site.
| Week | Impressions | Clicks |
|---|---|---|
| Aug 10–16 | 346 | 5 |
| Aug 17–23 | 2,403 | 2 |
| Aug 24–30 | 2,905 | 9 |
| Aug 31–Sep 6 | 2,405 | 0 |
| Sep 7–13 | 1,567 | 4 |
| Sep 14–20 | 695 | 3 |
Look at what happens. Almost nothing in week one. Then a flood. Then the flood drains away, until week six is down 76% from the peak.
If you only watched that column, you would think something broke. I know owners who have fired an agency over a chart that looked exactly like this.
Nothing broke.
The number that actually matters
Here is the same period, measured by average position instead:
- September 11: position 78.2
- September 23: position 47.5
Impressions fell by three quarters. Position improved by thirty places. Those two things happened at the same time, and the second one is the one that pays.
This is the single most misread pattern in new-site SEO, so it is worth being precise about what causes it.
What Google is actually doing
When your site is new, Google has no idea what you are good at. It has no click history, no reputation, nothing. So it does something sensible: it guesses, broadly, and watches what happens.
That is the flood in weeks two to four. Google showed my pages for hundreds of searches across half of British Columbia — mostly at positions 70 to 99. Page 8. Page 10.
Then it checked the results of that experiment. Here is one page of mine:
My Victoria page: 2,915 impressions. Zero clicks. Average position 92.75.
Almost three thousand times Google put that page in front of somebody, and not one person clicked. Google is not sentimental about this. It stopped showing it. That is most of my missing impressions — and losing them cost me exactly nothing, because they were never worth anything.
Meanwhile, the pages people do click kept climbing.
Where a new site ranks fast, and where it does not
The split was brutally clear, and it had nothing to do with how hard I worked on each page.
Cumberland — where I actually am — reached position 7.1.
Victoria — three and a half hours away — sat at position 92.
Same site. Same effort. Same quality of writing. The difference is that Google’s local results weigh three things: relevance, distance, and prominence (Google’s own explanation). Distance is not something you can write your way around. A Victoria customer searching for a local supplier is not looking for someone in Cumberland, and Google knows it.
There was one more useful surprise. I published a page aimed at a single trade — tree services — and in two weeks it went from position 57.5 to 29.0. A narrow page about one specific job beat broad pages about everything, and it beat them quickly.
So what should you actually expect?
Based on watching this closely, here is an honest timeline for a genuinely new site.
Weeks 1–2: almost nothing. Google has to find you first. Google is explicit that it does not guarantee it will crawl or index any page, so this stage is not a formality. Submit your sitemap, set up Search Console, and be patient.
Weeks 2–5: a confusing flood. Lots of impressions, almost no clicks, terrible average position. This feels like progress and mostly is not. Do not celebrate it and do not spend money chasing it.
Weeks 5–8: the drop. Impressions fall, sometimes hard. This is the stage that panics people. Check your average position before you panic — if it is improving, the system is working correctly.
Months 3–6: the real answer. This is when the pages that survived start collecting clicks. I am not there yet, and I am not going to pretend otherwise.
How to tell if yours is on track
Open Search Console and check three things. Ignore everything else for now.
- Is average position improving? This matters more than impressions in the first three months.
- Which pages get clicks, not impressions? A page with 12 impressions and 2 clicks is worth more than one with 3,000 and none. Put your effort there.
- Are you close on anything local? Anything sitting between positions 5 and 20 is within reach. Anything at 90 in a city you do not operate in is not, and no amount of writing will change that.
One caveat on the numbers themselves: Search Console groups and omits some data, so page totals and site totals will not always match exactly (Google explains why). Use it for direction, not accounting.
The honest version
Six weeks of correct, careful work bought me 23 clicks. That is the real number, and anybody quoting you faster results for a brand-new site is either lucky, in an empty market, or guessing.
What six weeks did buy was information: proof that I rank where I actually am, proof that distant cities were never realistic, and proof that narrow, specific pages move faster than broad ones. That is what the first two months are genuinely for.
If your site is new and the graph is scaring you, check your average position before you change anything. The chart that looks like failure is usually just Google finishing its experiment.
The figures above come from this site’s own Google Search Console property, covering 10 August to 25 September 2026. If you want help reading yours, get in touch — or run our free site check to see what a search engine sees when it looks at your website.