Expectations

How Long Does SEO Take? Stop Asking for a Date, Start Asking for Indicators

"Three to six months" is the standard answer, and it is useless to a buyer. What you need instead is a set of checkpoints that tell you at 30 and 90 days whether this is working — long before rankings could have moved.

Every guide gives the same answer: three to six months, longer in competitive markets. It is broadly true and almost useless, because it tells you nothing you can act on. You cannot manage an agency relationship against a range that wide, and by the time you can prove the range was wrong, you have spent two quarters of budget.

The question is worth replacing. Not “when will this work” but “what should have changed by 30 days, by 90, and by 180 — and what do I do if it hasn’t?

Why the standard answer is so vague

Because the honest version has a variable in it: how much work is being done.

Timelines get quoted as if SEO has an intrinsic speed. It does not. It has an intrinsic sequence — you cannot rank a page you have not published, and you cannot publish at scale on a site that cannot be crawled — but how fast you move through the sequence is a function of hours.

At the most common rate band in our directory ($100–$149/hr, published by 46.6% of the 1,143 agencies that publish one), a $1,000 monthly retainer is roughly eight hours. A $4,000 retainer is roughly thirty-two. Those two engagements do not run on the same clock, and quoting them the same “three to six months” is where most of the disappointment in this industry is manufactured.

So the first thing to establish is not the timeline. It is the hour count behind the retainer. Then the timeline becomes a reasonable conversation.

The 30-day checkpoint: has the work started?

Nothing about rankings is knowable at 30 days. What is knowable is whether an engagement has begun in earnest or is still circling.

By day 30 you should have:

  • A written statement of which terms and pages this engagement is aimed at, and which it is deliberately ignoring.
  • A technical issue list, prioritised, with the first items already closed rather than only documented.
  • Access sorted — analytics, Search Console, CMS — without anyone still chasing a login.
  • A named person you contact, and a known cadence.

The failure mode at 30 days is an audit and nothing else. Audits are cheap to produce and impressive to receive, and an agency can spend a first month generating one without committing to anything. If month one produced a document and no changes, say so in month two, not month six.

The 90-day checkpoint: are the leading indicators moving?

This is the one that matters, because it is the earliest point at which you can distinguish “working but early” from “not working.”

Rankings are a lagging indicator. These are not:

  • Pages shipped and indexed. Not drafted — live and in the index. A content programme that has produced nothing indexable in a quarter has a production problem, whatever the reporting says.
  • Impressions on target terms. Search Console shows impressions long before it shows useful positions. Rising impressions on terms you care about, even at position 40, is the shape of early progress.
  • New terms appearing at all. Going from “not present” to “present but poorly ranked” is real movement and is easy to miss if the report only tracks a fixed keyword list.
  • Technical debt closed. Count the issues fixed, not the issues found.

If all four are moving at 90 days, the engagement is healthy even with no ranking change. If none of them are, you have a finding — and you have it three months earlier than the standard timeline would have given it to you.

The 180-day checkpoint: is the thesis holding?

Six months in, the question changes from “is work happening” to “was the strategy right.”

By now you should see position improvements on the easier end of the target set, and enough data to test the original thesis. The useful conversation is not “are we ranking yet” but:

  • Which target terms moved, and were they the ones the strategy predicted would move first?
  • Where the work has not moved anything, is the explanation specific — a competitor’s link profile, a site architecture constraint, a category that needs a different format — or is it “SEO takes time”?
  • Given what is now known about this market, would we scope the same engagement again?

That last question is the one to calendar at signing. It is dramatically easier to ask on a date both sides agreed to in advance than to raise cold in month seven.

What actually makes it faster or slower

Roughly in order of impact:

  • Hours per month. The variable nobody quotes and everybody feels.
  • Existing site authority. An established domain with real links moves in weeks on terms a new domain will not touch for a year. This is the single biggest reason two businesses get wildly different timelines from identical work.
  • Competitive intensity of the target terms. Commercial terms with high advertising value are contested by well-funded incumbents, by definition.
  • How much has to be fixed first. A site with crawl, template or migration damage spends its first months on repair, which is real work that produces no visible ranking movement.
  • Your own throughput. Agencies are frequently blocked on client approvals, developer time and subject-matter input. This is the delay clients most often attribute to the agency.

Local and single-location work tends to move fastest, because the surface area is small and a Google Business Profile does a lot of the lifting — which is why local SEO engagements can be genuinely small and still work. Large-catalogue ecommerce work tends to be slowest, because the technical foundation has to be right before content scale means anything.

What to put in the agreement

Three lines, agreed before work starts, remove most of the ambiguity:

  1. The hour count. How many hours a month this retainer represents, and roughly how they split between strategy, production and reporting.
  2. The leading indicators. Which four or five things will be on every report, chosen because they move in weeks.
  3. The review date. A calendared six-month checkpoint with the explicit question: given what we now know, would we start this engagement again on these terms?

None of that shortens the timeline. It just means you find out early, which is worth more than speed — and it is a fair test to apply while you are still choosing, alongside the rest of the shortlisting checklist. Agencies that have thought about their own accountability answer these questions easily. The ones that have not will want to talk about the range again.

Common questions

How long does SEO take to show results?

Movement on low-competition and branded terms often appears within one to three months. Movement on competitive commercial terms — the ones with real budget attached — commonly takes six to twelve months. But the timeline is a consequence of effort rather than a property of SEO. A twenty-hour-a-month engagement and a five-hour-a-month engagement do not run on the same clock, and most disappointing timelines are underfunded rather than slow.

How do I know if my SEO is working before rankings move?

Watch the indicators that move in weeks rather than quarters — pages published and indexed, technical issues closed, impressions on target terms in Search Console, and average position on terms where you were not previously appearing at all. Impressions rising while position is still poor is the normal shape of early progress.

Is it normal to see no results after three months?

It is common, and on its own it is not decisive. What is decisive is whether the leading indicators moved. Three months with no ranking change but with pages shipped, technical debt cleared and impressions climbing is an engagement working as expected. Three months with no ranking change and no leading movement either is an engagement to question now, not at month six.

Should I sign a twelve-month SEO contract?

Long terms are defensible — the work genuinely compounds and an agency that has to re-sell every quarter optimises for the wrong things. What matters is not the length but whether there is a defined review point with real consequences inside it. A twelve-month term with a six-month checkpoint and agreed indicators is fine. A twelve-month term with no checkpoint is a twelve-month bet.

Where to go next

The research behind this post lives in the rankings. These are the lists it draws on.