How long does SEO take?

Impressions arrive in days. Clicks arrive in months. Those are two different clocks, and almost every argument about how long SEO takes is really two people quoting different ones at each other.

Concretely: four sites launched between 8 and 25 July 2026, checked in Search Console on 10 August 2026, have produced 27,260 impressions and 136 clicks between them. None of them is older than five weeks. The fastest went from zero to 16,500 impressions in about 17 days, peaking near 2,400 impressions a day, and collected 98 clicks doing it - a 0.6% click-through rate at an average position of 22.1.

Phase one - discovery and indexing (days)

Google finds your pages and stores them. On a small new site with a submitted sitemap this takes days. It's the only phase with a binary pass/fail - your URLs are in the index or they aren't - and failures are self-inflicted: a leftover noindex, a blocked path in robots.txt, or content that only exists after JavaScript runs.

Phase two - impressions (days to a few weeks)

Once indexed, Google tests your pages against real queries - mostly long, specific ones nobody is fighting over. That's why the impression curve is steep while the click curve stays flat: you're being shown, at the bottom. Here, the two sites chasing consumer queries went from zero to thousands of impressions inside a fortnight; the two enterprise products settled into a lower, steady daily rate.

Phase three - position (weeks to months)

This is the slow one, and it's slow for a reason that isn't a conspiracy: the signals that move a page from position 25 to position 5 accumulate rather than get configured. Links, people choosing you over the results above you, a track record in a topic - none of it can be done on the afternoon you launch. Average position is also violently noisy on a young site, moving ten places in a week because the query mix underneath changed, not because your page did.

Phase four - clicks, then money (months, then quarters)

Clicks are downstream of position, and revenue is downstream of clicks by however many clicks a sale takes. That last number is wildly different for a site that monetises traffic and a product that monetises a demo request, which is where the question stops having one answer.

First-party Google Search Console figures as of 10 August 2026, unrounded. Two are volume plays that make money from traffic; two are enterprise products that make money from a handful of the right buyers. That distinction is the most important column in the table.

The volume sites did what volume sites do: itscheaper.in peaked near 2,400 impressions a day. The enterprise sites did something that looks like failure and isn't - integri.us has flattened into a steady 100–150 impressions a day and produced five clicks in five weeks. On a volume site that's an emergency. On a product where one qualified demo pays for a year of the effort, it's an unalarming early plateau.

A result at position 22 gets shown to thousands of people and clicked by almost none of them. That isn't a mystery or a penalty. It is the third page.

Published click-curve analyses disagree about the exact percentages and agree about the shape: the top organic result takes a large double-digit share of clicks, position ten takes low single digits, and past position twenty it rounds to nothing. A young site averaging position 22 therefore has a click-through rate near zero by construction. It isn't being ignored - it's being displayed to people who never scroll that far.

So the honest answer to "how long until traffic" is a question about position, not a duration. If you're at 25 and your competitors are at 4, the wait is however long closing that gap takes - a function of your topic's competitiveness, not the calendar.

The caveat most timeline advice skips

Average position is a weak instrument, and our four sites prove it. itsavailable.in has the best average at 12.3 and converts at 0.5%; pharmalogiks.com sits at 48.8 and also converts at 0.5%. Average position is impression-weighted across every query you appeared for, so a site can average 48 while a handful of pages genuinely rank, dragged down by a long tail of hopeless ones. What matters is how many queries sit in your top five, not the sitewide average.

This is the shape our four sites are tracing. Treat it as a reference curve for spotting anomalies, not a promise - the absolute numbers depend entirely on how much search volume exists in your topic.

Patience is the right default and a terrible universal answer. These four symptoms aren't phases you're passing through - they're faults, and waiting makes them worse.

• Not indexed at all after two weeks. If your pages aren't in the index, time changes nothing - the clock hasn't started. Usual culprits: a staging noindex, blocked paths in robots.txt, no sitemap, or content that only appears after JavaScript runs.
• Indexed, but zero impressions after three or four weeks. Indexed-and-invisible means your pages target queries with no volume, or ones so competitive you never enter the shown set. That's a keyword-selection failure disguised as a patience problem.
• Impressions, but entirely the wrong queries. Thousands of impressions no buyer would ever type is a topical mismatch: the machine understood your pages as being about something other than what you sell. Time makes it worse - you're accruing relevance for the wrong topic.
• A flat impression line from day one. A young site should be visibly directional even at tiny numbers. A dead-flat month is a signal, not a plateau - plateaus come after a climb.

You can compress the front of the curve reliably. You can compress the back of it only by choosing an easier fight.

• Be readable without JavaScript. The most common silent failure on modern sites, and it delays or prevents the whole curve. View source with JS disabled; if you can't read your content, nothing downstream matters.
• Publish more pages, not better ones, at the start. Every page is a lottery ticket in a different query space. The fastest site here got there on breadth.
• Pick queries proportionate to a site with no history. Choosing an easier fight is the only lever that genuinely shortens phase three. Everything else is running harder at the same wall.
• Submit a sitemap and check coverage weekly for a month. Minutes of work, and it catches the failures that otherwise cost you a month waiting for a curve that was never going to start.
• Get mentioned somewhere real. Being written about elsewhere is still the strongest accelerant on position - and the hardest, which is why it works.

The Launch Roast is a fixed-price teardown of one live site against the same rubric used on the four above - indexation, impressions, position, and whether the queries you're appearing for are queries anyone would ever click. You get the answer in 48 hours instead of guessing for another quarter.

Or get the 30-day post-build playbook Send it