How long does SEO take? What the evidence supports

Ask how long SEO takes and you will get "six to twelve months" from almost everyone. The consistency of that answer is suspicious. Real measurements of complicated systems do not converge that neatly, and the people giving the answer mostly sell the thing being timed.

The number is not a lie. It has an origin, and the origin is more useful than the number. But it is being used to answer a question it was never designed to answer: not "how long does this take in general" but "how long will this take for me".

The question hides three different questions

"How long does SEO take" is usually three questions wearing one coat.

The first is mechanical: how long until Google notices a change I made? That is often days. Google crawls, indexes, and re-evaluates continuously. Fix a page title, and the title in the results can change within a week. Nothing about that part is slow.

The second is competitive: how long until I outrank the pages currently sitting above me? That has no fixed answer, because it depends entirely on who those pages are and what would have to change for the ordering to flip.

The third is commercial: how long until this produces enough enquiries to justify what I am spending? That one is affected by ranking but not determined by it — a first-place ranking on a query nobody searches produces nothing, on any timescale.

Most disappointment comes from asking the third question and receiving an answer to the second. If you want the sequence of what to actually do and when, a realistic timeline for ranking a brand new website sets that out step by step. This article is about interrogating the estimate itself.

Where "six to twelve months" actually comes from

The figure has a traceable source. Google's own guidance for businesses considering whether to hire an SEO has, for years, said that changes take time to have an effect, and puts the typical window at roughly four months to a year before you see benefit. That is a real, checkable statement from Google's documentation, not a study.

Notice what it is: advice to a business owner about setting expectations before signing a contract. It is a calibration device, deliberately wide, deliberately conservative, aimed at the average of everyone who might read it. It was never a measurement of how long anything takes.

That advice then got repeated by every agency that needed a client to be patient, tightened from "four months to a year" to the punchier "six to twelve months", and eventually cited as though it were an experimental result. It is now the most repeated number in the industry and one of the least examined.

The number does have some rough physical support. Building the things that move rankings — pages worth linking to, coverage of a topic deep enough to be credible, links earned rather than bought — takes months of work in the ordinary sense that any body of work takes months. And ranking systems weight signals that accumulate. So a long window is directionally right for a site starting cold. It just is not an answer to your question.

Why an average cannot answer your question

Here is the problem with applying it to yourself: the distribution behind that average is enormously wide, and your site is not sitting near the middle of it.

Consider two situations. A local dance studio wants to rank for "salsa classes in Norwich". Three competitors exist, none of whom have touched their websites in years, and the query is unambiguous. Movement here can happen in weeks, and the ceiling is low but reachable.

Now a personal finance startup wants to rank for "best savings account". The results are held by banks, national newspapers and comparison sites with decades of accumulated authority and full-time editorial teams. Six to twelve months is not an underestimate here — it is a fantasy. The honest answer might be years, or never, and the useful answer is to compete somewhere else.

Averaging those two produces a number that describes neither. It is the same error as saying the average British house costs a certain amount and using it to price a specific street.

An average is only decision-useful when variance is low. Here the variance is the whole story.

Four factors that set your clock

What follows is a reasoning framework, not research. I have not measured a dataset of sites; nobody publishing this kind of estimate has, at least not one you can inspect. What it does is force you to make explicit the four things that dominate the answer, so that your estimate comes from your own situation rather than from a number you inherited.

Score each factor 1, 2 or 3. One means favourable. Three means hostile.

Competition on the specific queries you want. Not your industry — the queries. Open a private browser window and search for three or four terms a customer would actually type. Score 1 if the first page is mostly small businesses, thin directory pages or forum posts. Score 2 if there is a mix of established sites and weaker ones. Score 3 if the whole page is national brands, major publishers, or Google's own surfaces — maps, shopping, AI summaries — leaving little room for a blue link at all. That last case is common and underrated: sometimes the space you are competing for has quietly shrunk.

Your starting position. Score 1 if you have an established domain with a history, existing rankings for related terms, and links you did not have to ask for. Score 2 if the site is a couple of years old with modest signals. Score 3 if the domain is new, or the site has a penalty history, or previous owners did something Google's spam policies would treat badly — buying links, for instance, which Google's documentation classes as link spam regardless of who was running the site at the time.

Whether the constraint is technical or competitive. This is the factor most people misjudge, and it cuts both ways. Score 1 if there is something obviously broken: pages blocked from crawling, no indexable content, a migration that dropped every URL. Broken things get fixed and recover quickly, sometimes within weeks. Score 3 if the site is technically fine and the pages are decent, and you are simply below better pages. That is the slow case, because the only remedy is becoming genuinely more deserving of the position, and there is no shortcut through it. A helpful way to think about why is in how Google ranking works, explained without jargon.

Your realistic execution rate. Not what you intend — what you will actually sustain. Score 1 if someone is working on this most weeks and has budget for it. Score 2 if it gets a few days a month. Score 3 if it is you, in the evenings, after the actual job. This factor is unglamorous and probably the most predictive of the four, because everything else assumes work is happening.

Reading your total

| Total | Rough band before meaningful movement | What that band usually means | |---|---|---| | 4–5 | Weeks to about three months | Low competition, something fixable, real capacity. Move fast; the window may not stay open. | | 6–7 | Three to six months | The genuinely favourable ordinary case. | | 8–9 | Six to twelve months | Where the famous number actually applies. | | 10–12 | A year or more, if ever | Re-examine whether these are the right queries, or the right channel. |

"Meaningful movement" here means visible progress on real queries — appearing on the first two pages for terms you care about, and enquiries starting to arrive — not a single ranking milestone.

Two worked examples, to show the scoring in use.

A two-person plumbing business in one town, four-year-old site, currently invisible because the site has no page for its own services and the Google Business Profile is unclaimed. Competition 1, starting position 2, constraint 1 (things are missing rather than outclassed), execution 3 because it is evenings only. Total 7. Three to six months, and the first useful results probably come from the business profile and a handful of service pages rather than anything sophisticated.

A SaaS with 40 customers targeting a category term that established software companies rank for. Competition 3, starting position 3, constraint 3, execution 2 with a founder who can write one solid piece a fortnight. Total 11. That does not mean give up on search — it means the category term is the wrong target this year, and the realistic route is narrower queries where the score would be lower.

What would make me conclude it is not worth it

We sell paid visibility, so we have no reason to talk anyone into organic work. That also means when we say it is the right call, you can weigh it accordingly — and when we say it is not, here are the conditions.

Your buying window is shorter than your timeline. A seasonal business eight weeks from its peak, or a startup with four months of runway, cannot spend six months waiting. Organic work may still be right as background investment, but it cannot be the plan for this quarter.

The queries do not exist. If you are creating a category, nobody is searching for it yet. Ranking first for a term with no searchers is a well-optimised silence. That is a demand problem, and search is a demand-capture channel.

The results page has no room. If commercial queries in your space are dominated by aggregators, marketplaces and Google's own panels, then even a perfect page is competing for a small residual slice.

You cannot sustain the work. A plan requiring consistent monthly output from someone who does not exist is not a plan.

The arithmetic favours paid. If a paid channel already converts at an acceptable cost, the case for organic is future margin rather than urgent volume — worth doing, wrong to depend on. An honest cost comparison of paid versus organic works through where each one wins.

If several of those apply, buying visibility outright is a legitimate answer — that is what our ranked marketplace listings are, positions priced openly and held by verified spend, with no waiting period because there is nothing to earn. It is also not organic ranking and never becomes it: the honest limits of pay-to-rank says plainly what it cannot do for you. Most sensible answers are a mixture — buy attention now, build the slow asset in parallel.

Questions people ask

Does more spend make organic faster? Somewhat, up to a point. Money buys execution rate — the fourth factor — and can convert a score of 3 into a 1. It does not touch competition or starting position, so it compresses the timeline rather than collapsing it.

Why do agencies quote a timeline before looking at my site? Because the quote is a hedge, not an estimate. Any timeline given before someone has seen your queries, your competitors and your site is a restatement of the industry average. A worthwhile answer names your specific competitors and what would have to change.

How would I know it is working before rankings move? Impressions on new queries in Search Console, pages entering the index, and positions improving from very deep to merely deep. Those move earlier than first-page rankings and are the honest early signal.

Is there any way to go faster than the work allows? Not through search results themselves. Paid channels are faster because you are buying placement instead of earning it, which is a different transaction with different economics — and worth being clear-eyed about rather than treating either as a shortcut to the other.