Hiring an SEO Agency · Guide

How Long Should You Commit to an SEO Agency?

This work genuinely needs months to show anything. That is also the argument used to justify agreements which mainly protect the agency. Both of those are true at once and this page holds them both rather than picking whichever suits us.

Updated: August 2026
Written by: Andrew Odgers, Managing Director
Reading time: 11 minutes
Neither side of this is the whole answer

Two True Things That Pull Against Each Other

The work needs time. That statement is accurate. It is also the sentence every agency reaches for when asking a business to sign away its ability to leave. You have to hold both.

The first true thing. Nothing happens quickly.

A business expecting visible change within weeks will cancel before anything could have worked, having concluded from an accurate observation that nothing was happening. Block two covers why.

The second true thing. That argument is convenient.

It is the perfect justification, because it is unfalsifiable in the short term. Any complaint before the term expires can be answered with patience. A firm doing nothing sounds identical to a firm doing the early work properly.

Why most pages on this get it wrong. They pick one.

Agency-written pages tell you to commit for a long period because the work is slow. Sceptical pages tell you never to sign a term at all, which ignores why the term exists. Both are simpler than the situation.

What holding both looks like. A different question.

Not how long should you commit. What should be true by which point, plus what should override the term entirely. Those are answerable and the rest of this page answers them.

The case, made properly

Why It Genuinely Takes Time

Four reasons the early period produces little that is visible. Each is a fact about how the work operates rather than an appeal for patience.

The first period is research and planning. By necessity.

Understanding the business, auditing what exists and deciding what to do have to happen before anything is produced. None of that generates output a client can look at.

Content has to exist before it can do anything. Then wait.

A page must be written, published, found and assessed before it can contribute. That sequence has a floor beneath it that no amount of money removes.

The effect compounds. Slowly, then faster.

Early work makes later work more effective, so the shape is flat at the start and steeper afterwards. Judging the whole by the flat part is the commonest misreading of this work.

You are moving relative to others. Who are also working.

Improvement means overtaking somebody who is not standing still. That takes longer than improving against a fixed standard would.

What this does not license. Indefinite patience.

Everything above explains why results are slow. None of it explains why work would not be visible. Block six turns on exactly that distinction.

A principle rather than a number

What A Reasonable Term Looks Like

We will not tell you how many months is correct, because it depends on your situation and any figure we gave would be presented as a standard. What can be described is what makes a term reasonable.

It should be justified by the work. Specifically.

An agency should be able to say what happens during that period and why a shorter one would not allow it. That answer is checkable against what actually arrives.

You should be told why it is that length. Before signing.

Nobody should have to ask. A term presented as a formality, else explained as industry standard, has not been justified. It has been asserted.

It should be proportionate to the front-loading. A fair exchange.

Where an agency does substantial early work on the expectation of a longer engagement, a term covering that exposure is fair. Where the work is level throughout, the argument weakens considerably.

It should not survive non-delivery. The crucial part.

A reasonable term binds you while they work. One that binds you whether or not they work is protecting income, which the contract page covers alongside the four things we commit to.

The question that settles it. Ask what happens if.

Ask what the term means if they stop producing work. The answer separates the two kinds of term faster than reading the clause does.

Achievable, also worth asking for

The Rolling Alternative

An initial period followed by a monthly rolling arrangement is entirely workable. Firms confident in their work offer it. It resolves most of the tension in block one.

Why it resolves the tension. It separates the two claims.

The initial period covers the front-loaded work, which is the legitimate argument. The rolling arrangement afterwards means retention depends on continuing to deserve it, which removes the illegitimate one.

Why agencies resist it. Predictability.

Rolling arrangements are harder to plan a business around, which is a real cost to the agency rather than a pretext. Knowing that is fair before you ask for one.

What it signals when offered. Confidence.

A firm content for you to be able to leave each month is a firm expecting you not to want to. That is a stronger statement about their work than any assurance.

Our own position. Stated on the contract page.

We commit to no tie-in that exists to protect us rather than to make the work possible, which is deliberately not the same as promising no minimum term. The distinction is whose interest it serves.

How to ask for it. Directly.

Ask whether the term can become rolling after an initial period. The response tells you which kind of term you were being offered.

A checkpoint for judging the agency

Judge It At Three Months, Not Six Weeks

Six weeks is too early to assess anything meaningfully. By around three months you can assess the agency, which is a separate matter from assessing the results.

What you are judging. Them, not the outcome.

This is the distinction that makes an early checkpoint possible at all. Results may be nowhere in sight while the agency is performing perfectly. The four things below are about their conduct rather than your position.

Work produced. Something exists.

Pages written, fixes applied, listings addressed. Not a quantity we can specify. There should be a body of work you can point at.

A coherent plan. That you could describe.

You should be able to say what they are doing and roughly why. If you cannot, either it was not explained or there is nothing specific to explain.

Early movement in visibility. On easier terms first.

Appearing for a wider range of less competitive searches usually precedes anything on the searches you care about. That is the earliest genuine sign. Its absence at this stage is not yet a verdict.

Understanding of your business. Demonstrated.

They should now know things about your customers and your margins that they did not at the start. What to expect in the first three months goes through it month by month.

The counterweight to everything above

What Should Make You Leave Early

A term is not a reason to stay through non-delivery. Four things override any commitment. None of them requires you to understand the work.

No work produced. The plainest.

If months pass with nothing written, fixed or changed, the argument about results taking time does not apply. That argument explains slow outcomes rather than absent activity.

No reporting. Or reporting you cannot verify.

Nothing arriving, else documents you cannot connect to any actual work. Either way you have no way of knowing what you are paying for.

No contact. Silence.

An agency you cannot reach, else one that only appears when invoicing. Whatever the term says, that relationship has already ended in practice.

Practices that worry you. Trust your judgement.

Something described that sounds wrong, else an unwillingness to explain what is being done. You do not need to prove anything to act on it.

Why the term does not protect them here. It was justified by the work.

Every argument for a minimum term rests on the work being done during it. Where it is not, the justification has gone. When to change your SEO agency covers doing it properly.

The other side of leaving

Stopping Is Not Neutral

Ending an arrangement has a cost even when it is the right decision. The work decays rather than freezing. Leaving and rejoining is more expensive than continuing would have been.

What does not happen. An immediate collapse.

Nothing switches off. What was built remains, so the first weeks after stopping look reassuring and tell you very little.

What does happen. Gradual erosion.

Competitors continue, content ages and nothing adapts. The position slips slowly, which means the cost of stopping arrives long after the saving and rarely gets connected to it.

What restarting costs. Ground and time.

A new arrangement means somebody learning the account again and recovering what was lost meanwhile. That expenditure produces nothing new, it returns you to where you were.

How to weigh it. Against the reason.

None of this argues against leaving a firm that is not delivering. It argues against leaving casually, plus against pausing to save money without counting what the pause costs.

The middle option. Reduce rather than stop.

Where money is the issue rather than performance, a smaller ongoing arrangement keeps continuity. The full series is on the hiring an SEO agency guide.

Website migrations

Ask what the
term means if
we stop working.

That question separates a term protecting the work from a term protecting income, faster than reading the clause does. Ask it of us and of anybody else you are considering, then ask whether the arrangement can become rolling after an initial period.

What we commit to in writing:

No setup charge No protective tie-in You own the website Updates every three weeks Work you can inspect Reporting you can question Accounts in your name Handover if you leave

Local £350 a month fixed. National £1,550 a month. Published rather than quoted on request.

The full guide series

Every guide.
One practice.

What an agency is, whether to buy at all, which model suits you, how to vet one, what things cost, what should be in a contract and how to work with them afterwards.

Questions people ask

Commitment, Briefly

How long should we sign up for?
We will not give you a number, because it depends on your situation and any figure would be treated as a standard. What matters is whether the term is justified by the work: an agency should be able to say what happens during that period and why a shorter one would not allow it. A term explained as industry standard has been asserted rather than justified.
Is it reasonable for an agency to want a minimum term?
Yes, where they do substantial early work on the expectation of a longer engagement. A term covering that exposure is a fair exchange. Where the work is level throughout, the argument weakens considerably. The test is what the term means if they stop producing work: one that binds you regardless is protecting income.
Can we get a rolling monthly arrangement?
Often, so it is worth asking directly whether the term can become rolling after an initial period. It separates the two arguments cleanly: the initial period covers the front-loaded work, then rolling afterwards means retention depends on continuing to deserve it. Agencies resist because rolling arrangements are harder to plan a business around, which is a real cost rather than a pretext.
When can we tell whether it is working?
Around three months you can assess the agency, which is different from assessing the results. Four things should be visible: work actually produced, a plan you could describe, early movement for less competitive searches, then evidence they now understand your business. Results may be nowhere in sight while the agency is performing perfectly.
We are in a term but nothing is happening. Do we have to stay?
A term is not a reason to stay through non-delivery. If no work is produced, no reporting arrives or you cannot reach anybody, the argument about results taking time does not apply, since it explains slow outcomes rather than absent activity. Every argument for a minimum term rests on the work being done during it.
Could we pause for a few months to save money?
You can, though it is not a neutral act. Nothing switches off immediately, so the first weeks look reassuring and tell you little. What follows is gradual erosion as competitors continue and content ages, so the cost arrives long after the saving and rarely gets connected to it. Where money rather than performance is the issue, reducing beats stopping.