How Long Does SEO Take to Work? Real Timelines by Site Type
Three to six months is the standard answer and it hides more than it tells. There are four separate clocks running, they move at different speeds, and in 2026 one of them came loose from the others.
How long does SEO take? Ask ten providers and nine will say three to six months, including most pages selling SEO services. It is not a lie, but it is a single number standing in for four different processes, and that is why so many engagements end in an argument about whether anything happened.
Four clocks, not one
When somebody asks when SEO starts working, they usually mean when the phone rings. That is the slowest of four things, each of which has to happen before the next one can.
| Clock | What moves | Typical time |
|---|---|---|
| Indexing | New and fixed pages enter Google | Days to a few weeks |
| Visibility | Impressions rise, positions start to settle | 1 to 3 months |
| Clicks | Visitors actually arrive | 3 to 8 months |
| Revenue | Enquiries close and become money | 5 to 12 months, plus your sales cycle |
Read that table twice, because it explains almost every disappointment in this field. At month three a competent engagement normally has movement on the first two clocks and almost nothing on the fourth. Reporting that leads with revenue at month three makes good work look like failure, and reporting that leads with impressions at month ten hides real failure.
What Google itself says
The most defensible single figure available is Google own: four months to a year before you see impact, given in official guidance by Maile Ohye. It is old and it has aged well, which is unusual in this field. Google has also described a transition period in its patent filings during which a changed page settles into a new position rather than jumping, roughly measured in weeks rather than days.
Supporting evidence from third parties points the same way. Ahrefs research on newly published pages found that only a small fraction reach the first page within a year, and that the large majority of pages currently sitting on page one are several years old. Note that different write-ups of that study quote the figure differently, so treat the direction as solid and the exact percentage as worth checking at source.
Timelines by site type
The averages are almost useless because the starting positions are so different. This is closer to reality.
| Situation | First real signs | Meaningful traffic | What governs it |
|---|---|---|---|
| Brand new domain | 2 to 4 months | 8 to 14 months | No history, no links, nothing to build on |
| Established site, never optimised | 4 to 10 weeks | 4 to 7 months | Often the fastest wins anywhere, because the authority already exists and is being wasted |
| Local service business | 3 to 8 weeks | 2 to 5 months | Map pack and reviews move faster than web rankings do |
| Ecommerce with real category demand | 2 to 4 months | 6 to 10 months | Category page competition, product feed and technical health |
| B2B in a thin niche | 2 to 5 months | Traffic stays small by design | Low volume means you measure enquiries, never sessions |
| After a migration or redesign that went wrong | 2 to 8 weeks | Recovery to previous level in 3 to 6 months | How much was broken and how fast it is unbroken |
| After a core update loss | No quick signs | 6 to 12 months if at all | Site wide quality, not a switch anybody can flip |
Why the clicks clock came loose
Until recently, visibility and clicks moved together. They no longer do, because Google increasingly answers the question on the results page.
Pew Research Center found in July 2025 that the presence of an AI summary cut the share of searches producing a click on a normal result from around 15 per cent to around 8 per cent. Amsive analysed roughly 700,000 keywords in April 2025 and measured an average click rate decline of about 15 per cent where summaries appeared, worse on non branded queries and worse again where a featured snippet sat alongside.
The practical effect on a timeline is specific and worth writing into a contract: it is now normal for an engagement to improve impressions and positions on schedule while clicks lag behind the old benchmarks. That is not an excuse to accept flat traffic forever. It is a reason to insist that your report shows all three numbers, so the difference between a real stall and a market wide shift is visible rather than arguable.
What should have moved by month three
These are the leading indicators. If none of them have moved by the end of month three, something is wrong, and that is true regardless of what the traffic graph says.
- The pages commissioned in month one exist, are indexed, and you can find them with a site search.
- Impressions in Search Console are up against the same period last year, even if clicks are not.
- The number of distinct queries your site appears for has grown.
- Technical problems raised in the audit are closed, with a list of what was fixed and when.
- Something ranks for a term it did not rank for before, however unglamorous.
- You know which pages are meant to produce enquiries, and those pages have tracking on them.
What six months actually looks like
An honest example from our own work, including the part that is unfinished. A health and wellness brand came to us with effectively no organic visibility. Over six months we ran a technical audit and fixes, keyword research built around topical coverage rather than single terms, on page work across content, metadata, internal linking and schema, content written to hold up in a sector Google treats with extra scrutiny, local optimisation, and a steady link building campaign.
After six months: around 12,000 clicks and 2.41 million impressions in Search Console, average position 24.6, roughly 1,200 ranking keywords, and 643 backlinks from 249 referring domains.
Average position 24.6 is page three. That is the honest reading. The foundation is built and the impressions prove the pages are being shown, but most of those 2.41 million impressions are not being clicked, because page three rarely is. Moving those pages from page three to page one is the next six months of work, and it is where the traffic multiplies. Anyone presenting that six month position as a finished result is selling you the first half of a job.
What makes it faster, and what makes it slower
Faster: an existing site with age and links, a market where competitors have neglected their sites, decisions made in days rather than weeks, somebody internal who can answer technical questions, content approved without three rounds of legal review, and a genuine willingness to publish pages about unglamorous specifics rather than brand stories.
Slower: a brand new domain, a sector where the top ten is held by national publishers, a site on a platform that fights every technical change, approval cycles measured in months, a history of cheap link buying that has to be cleaned up first, and a budget that funds one of the three kinds of work while the bottleneck is in another.
When slow is normal and when it is a problem
Normal: flat clicks at month three with rising impressions. Normal: a competitive head term not moving in year one while longer phrases do. Normal: a dip during a core update window, then a recovery.
A problem: month four and the pages promised in month one are not published. A problem: no change in impressions or query count after six months of payment. A problem: reports showing only positions, only rankings for terms nobody searches, or only tasks completed. A problem: every disappointing month explained by an algorithm update, with no corresponding explanation of what will be done differently.
Write a review point into the engagement at month four, on the leading indicators rather than on revenue. If they have not moved, the strategy is wrong and four months is early enough to change it. Waiting twelve months to find out costs you a year and the fee.
Up to a point. More budget buys more pages, more technical capacity and more outreach at once, which does compress the schedule. It does not shorten the time Google takes to trust a new site, and past a certain level the constraint becomes your capacity to approve and publish rather than the money.
Usually the reverse. Improving a page that already has some history tends to pay off faster than publishing a new one, which is why a decent first ninety days often looks like rewriting rather than writing.
Expect two to four months for the first signs and the better part of a year for traffic that matters. If you need customers sooner than that, fund advertising for the gap and treat search as the thing that reduces your dependence on it later.
Then either the strategy was wrong, the work was not done, or the market was never winnable at your budget. All three are findable in an afternoon by someone independent looking at Search Console and the list of what was actually shipped.
Want this checked on your own site?
Send the URL and we will tell you what we would do first, in writing, before anybody talks about a retainer.
- A written audit within 48 hours
- No call required and no obligation
- If search is the wrong spend for you, we will say so
Free site audit
Your URL and what you want more of. That is all we need to start.
Request received!
Thanks! We have your request. Check your inbox within 48 hours for your written proposal. Nothing else needed from you.
