Agency or SEO consultant? A decision matrix

The question "should we work with an agency or a consultant" is usually asked in the wrong order: proposals are collected first, and only then does anyone try to understand what the two models actually are. Yet the decision can be made before requesting a single proposal, by looking at the organisation's own situation. This article defines the two models neutrally and proposes a five-criterion decision matrix. No competitor is named; these are the notes of a team that has seen both models work well in the right context.

Our own model is described on the approach page and our scope on the services page; this article was written to help you find which model fits you, and the decision may not land with us.

The real difference between the two models

An agency is a structure that gathers several specialisms under one roof: strategy, technical, content, design, sometimes advertising and development. A consultant is usually one person or a very small team; deep in expertise, narrow in scope, and usually leaving implementation to your team or a third party.

This definition says nothing about quality. There are bad agencies and good consultants, good agencies and bad consultants. The decision is not about which is better but about which one completes the piece you are missing.

Criterion 1: do you have implementation capacity

This is the most decisive question. If you have a development team to ship technical findings, a writer to produce content and an internal owner to coordinate them, what you need is diagnosis and direction; that is where the consultant model is strong. If that capacity is missing, the consultant's list of findings is still a list three months later, and an agency that also takes on implementation is the better choice.

In practice we test it like this: in the last six months, at whose request and how quickly were changes made to the website. If the answer is "within two weeks, by the internal team", a consultant works; if the answer is "still waiting", implementation capacity is the first thing to buy.

Criterion 2: is the problem concentrated in one area

If the problem sits in a single area, for example indexing and hreflang on a multilingual site, or AI search visibility, a consultant with deep expertise usually delivers faster. If the problem spans technical, content, advertising and conversion at the same time, a structure is needed to coordinate those areas, and building that structure inside the organisation is hard.

The common mistake here is keeping the scope wide and choosing a consultant, or keeping the scope narrow and buying the agency's full service package. Both produce the same result: part of the money paid goes to unused capacity.

Criterion 3: speed and decision distance

With a consultant the decision distance is short: the person you talk to is the person doing the work. In an agency that distance can stretch across an account manager, specialist teams and sometimes subcontractors. In return, the agency's capacity for parallel work is higher; technical fixes and content production can run in the same week.

Your question should be: is our bottleneck decision speed or production speed. If it is decision speed, the consultant has the advantage; if production speed, the agency model does.

There is also the matter of availability. Organisations working with a consultant notice that work stops in the weeks the consultant is on holiday or on another project; in an agency another specialist fills that gap. On the other hand, staff turnover is more frequent in agencies, and every change means a loss of context. Which weighs more depends on the organisation's own publishing rhythm: a site publishing weekly prioritises continuity, an organisation working to a quarterly plan prioritises depth of context.

Criterion 4: dependency and knowledge transfer

The consultant model, set up correctly, accumulates knowledge inside the organisation; because your team does the work, the method stays with you. In the agency model implementation is external, so knowledge can stay external too, and when the contract ends the organisation may return to where it started.

This is a tendency, not a rule. There are agencies that write knowledge transfer into the contract, and consultants who keep everything in their own hands. The question to ask is: what will we have when the contract ends, and whose responsibility is that.

Criterion 5: cost structure

A consultant usually works for a lower monthly fee but leaves implementation cost outside; total cost has to be calculated together with the internal team's time. The agency fee is higher but includes implementation. When comparing the two, put side by side not the monthly invoice but the three-month total cost and the amount of work shipped in that period. Where the pricing models break was covered in our pricing models article.

Reading the matrix

When you put the five criteria into a table, the result for most organisations approaches one of two profiles.

• Internal development and content capacity exists, the problem is concentrated in one or two areas, decision speed matters, knowledge should stay in-house: the consultant model.
• Implementation capacity is missing or busy, the problem is multi-area, production speed matters, handing coordination outside is acceptable: the agency model.
• In between: a consultant for diagnosis and direction, a limited-scope agency or freelancers for implementation. This hybrid gives the most efficient result when managed well and the messiest when managed badly.

Do not weight the criteria equally when filling in the matrix. Implementation capacity is decisive on its own in most organisations; the remaining four confirm or qualify the direction that answer gives. There is also a time dimension: an organisation that needs the agency model today can move to the consultant model within a year once it has built an internal team. That is why setting up the contract to be re-evaluated after a year is healthier than binding yourself permanently to either model.

What applies in both models

Whichever model you choose, the warning in Google's documentation on the subject is the same: no one can guarantee a number one ranking on Google. A proposal that guarantees results, claims a special relationship with Google or does not explain its method should be eliminated, agency or consultant. The same document lists the questions to ask any provider: can they show previous work, do they adhere to Google's essentials, what results do they anticipate within what timeframe and how will success be measured, do they have experience in your industry and your geography, will they explain every change they make.

These questions are model-independent. For a more detailed list, see our 12 questions to ask before hiring a consultant; every one of them can be put to an agency as well.

Conclusion

The difference between an agency and a consultant is not a difference in quality but a difference in structure. The decision starts with your implementation capacity, continues with the scope of the problem and is completed by speed, dependency and cost structure. Collecting proposals after filling in the matrix is far cheaper than collecting proposals and filling in the matrix afterwards. Whichever model you settle on, buy the method, not the guarantee.

Sources

• Do you need an SEO? — Google Search Central