Skip to content
Skayle Marketing

SEO · 8 min read

How long SEO takes, and what actually changes the answer

There is no single number, and any agency offering one is describing its billing cycle rather than your site. What decides the timeline is which constraint is holding you back — and the leading indicators that tell you it is working long before rankings do.

Written by , FounderUpdated

The honest answer

The single number does not exist

Three to six months is the standard answer, and it is not exactly wrong. It is just not about your site. It is a figure that sits near the middle of a very wide distribution, and it happens to match the shortest contract length an agency can run profitably. Two facts arriving at the same number is not the same as one of them explaining the other.

The reason a single figure cannot be given is that SEO is not one activity with one clock. Correcting a rule that was stopping pages being indexed can show up within a recrawl. Merging four pages that were competing for the same query takes a month or two to settle. Becoming a credible source in a category where three established competitors have been publishing for a decade is a multi-year project, and pretending otherwise is how twelve-month contracts end badly.

So the question worth asking is narrower and much more answerable. Which of those is your situation? That can be determined in the first few weeks by a competent diagnosis, and it should be the first thing any supplier tells you — before the plan, before the pricing, and certainly before the timeline.

The useful question is not how long SEO takes. It is which constraint is binding on your site, because a crawling problem and an authority problem run on completely different clocks.

Different clocks

How quickly different kinds of work tend to show

These are observed patterns, not commitments. What they are useful for is working out which row you are in, because that is what sets your expectation.

The rough pace of different SEO work, and what determines it in each case
DimensionTypical paceWhat decides it
Removing an indexing or crawl blockDays to a few weeks after the correction shipsHow often the site is recrawled, which depends on its size, its history and how frequently it changes
Titles, headings and on-page relevanceTwo to six weeks, where relevance was the actual constraintWhether the page was already being shown — impressions can convert to clicks quickly, but a page with none has further to go
Consolidating pages that compete with each otherRoughly one to three months to settleHow long the duplicates have existed and how many external links point at the version being retired
A new page on an established siteWeeks to be assessed, longer to reach its levelThe authority the site already has, and how contested the query is
A new page on a new domainMonths, and it starts slower than anyone expectsThere is no accumulated history to borrow, so everything has to be established from nothing
Building authority in a competitive marketSix months to well beyond a yearWho holds the positions now, how much they are still investing, and whether anything you publish is genuinely worth citing
Being recognised as an entity in your categoryThe longest of all, and the least controllableConsistent, corroborated references to your brand across sources you do not own or control

The variables

Five things that move the timeline more than effort does

  • What the site has already earned

    An eight-year-old domain with real links and existing impressions is starting a long way up the hill. A domain registered last month is not. This is the single biggest difference between two otherwise identical programmes, and no amount of budget replaces it in the short term.

  • Who you are actually competing with

    Not how difficult a keyword tool says a phrase is, but who currently holds the positions. Three national brands with in-house teams is a different problem from four local firms with brochure sites, even when the search phrase is identical. Look at the results page before you look at the metric.

  • Whether the constraint is technical or authority

    Technical constraints are usually fast and unusually satisfying: something was blocking the site, it stops blocking it, things move. Authority constraints are slow because they depend on other people. Diagnose which you have before agreeing to any timeline, because the two answers differ by a factor of ten.

  • How fast your organisation can approve and publish

    This is the constraint nobody puts in the proposal and the one that most often decides the outcome. If a page takes six weeks to clear legal, compliance and a founder’s inbox, the publishing schedule is set by your internal process rather than by the agency. It is fixable, but only by you.

  • How much has to be undone first

    A site carrying an old migration with broken redirects, three generations of duplicate pages, and content written for a search engine in 2016 spends its first quarter removing liabilities rather than adding assets. That work is necessary and it produces very little visible movement while it is happening.

What to watch instead

Leading indicators at 30, 60 and 90 days

Rankings are a lagging indicator and a terrible early signal. These move first, and they are the ones to agree on before the work starts.

  1. First 30 days: baseline and coverage

    Agree the baseline in writing and confirm the tracking is trustworthy. Then watch indexing: are the pages you care about actually indexed, is the number of valid pages moving in the right direction, and have the crawl errors on the priority templates started to close? At this stage you are looking for eligibility, not performance.

    You get: A written baseline and a clean indexing picture

  2. Days 30 to 60: impressions before clicks

    Being shown precedes being clicked. Watch the count of distinct non-brand queries where you appear at all, and the impression volume on the pages that were worked on. New queries appearing is the earliest genuine signal that relevance has changed, and it typically shows well before any position worth celebrating.

    You get: Query breadth and impression trend on non-brand terms

  3. Days 60 to 90: non-brand clicks and average position

    Now the clicks. Split brand from non-brand rigorously, because brand traffic rises whenever any other marketing runs and will flatter the report. Average position on the worked pages should be improving, and pages sitting just outside the visible range are the ones to watch, since small movements there produce disproportionate click changes.

    You get: Non-brand click trend, split by page template

  4. Months 3 to 6: conversions, not sessions

    Enquiries, calls, bookings or transactions attributable to organic, measured against the agreed definitions rather than whatever the platform defaults to. This is where a programme that is producing traffic without producing customers becomes visible, which is exactly the failure that six months of session reporting conceals.

    You get: Organic conversions against the baseline definitions

  5. Month 6 onward: the number that settles the argument

    Revenue or pipeline from organic, compared to the cost of the programme. Everything above is a proxy that exists because this number is slow to arrive. Once it does, the earlier indicators become diagnostic rather than reassuring, and the conversation moves from whether it works to where to invest next.

    You get: A cost-per-outcome figure you can compare across channels

Questions

Timeline questions people actually ask

How long does SEO take to work?

It depends on which constraint is holding the site back, and that is genuinely knowable within the first month. Indexing and crawl problems can resolve within a recrawl cycle. Relevance and structure problems typically move over weeks. Authority problems in competitive markets take many months, and sometimes longer than a year.

A more useful framing than a date is this: within thirty days you should know what the constraint is, within sixty to ninety days you should see the leading indicators move, and the commercial result follows those.

Why does everyone say three to six months?

Partly because it is roughly true for the middle of the distribution, and partly because it matches the shortest contract term an agency can profitably run. It is a billing shape as much as an observation.

Treat it as a prior, not a forecast. A site with a technical block and a strong brand can move much faster; a new domain in a competitive category will not.

How soon will I know if it is not working?

Sooner than most people expect, provided you agreed what to watch before starting. If new pages are not being indexed, if impressions on non-brand queries are flat after sixty days, and if nothing in the technical fix list has actually shipped, something is wrong with the programme rather than with the timeline.

The failure mode to avoid is agreeing to wait six months and only then asking what should have changed. By then the evidence for what went wrong has usually gone.

Does spending more make it faster?

Up to a point, and then not at all. More budget buys parallelism: more pages written at once, more technical debt cleared in a quarter, more outreach running simultaneously. That genuinely compresses some of the schedule.

What it cannot compress is the time a search engine takes to recrawl and reassess a site, or the accumulation of trust that comes from being cited and referenced over time. Beyond a certain point, extra money buys more work rather than faster results.

Is it faster to fix an old site or start a new one?

Usually faster to fix the old one. An established domain carries links, history and existing impressions, and all of that is an asset a new site starts without.

The exception is a site whose structure or platform makes the necessary changes impossible, or one carrying a serious penalty history. Even then, the right move is normally a rebuild on the same domain with careful redirects, not a new domain.

Find out which clock your site is on

The answer to how long it takes is mostly the answer to what is currently constraining you. That is a diagnosis, not a guess, and it is the first thing we would look at on a call.

Last updated

We use analytics to understand which pages are useful. Nothing runs until you choose, and we do not sell or share what we collect. What we would set.