SEO agency, freelancer or in-house: how to choose

Most buyers frame this decision as agency versus freelancer versus internal hire, as if the three were personality types. They are not. They are three ways of buying the same two things: senior judgement about what to do, and execution capacity to do it. An agency sells you breadth of skill with limited attention. A freelancer sells you a large share of attention with limited breadth. An internal hire sells you business context and permanent availability, at the price of one point of view. Everything else follows from that distinction.

The right answer also changes over a company's life, and that is normal. A five-person business with no content operation needs a generalist who can ship. A hundred-person business needs someone internal to own the roadmap and external specialists to fill the gaps. Treating the choice as permanent produces the two classic mistakes: hiring an agency to compensate for having no internal owner, or hiring an owner and giving them no budget to execute anything. The discipline described in our guide to choosing an SEO agency applies to all three.

What follows is a decision framework rather than a verdict. We look at what each model really costs once everything is counted, at the trade-off that sits underneath the headline price, at the situations where each model clearly wins, and at the hybrid arrangements most mature teams drift towards. One principle holds throughout: judge candidates on verified performance data rather than on the polish of the pitch, whether you are reading an agency directory, shortlisting independent SEO consultants or interviewing a candidate for an internal role in your own team.

Agency retainers in Europe in 2026 typically run from 500 to 5,000 euros per month, with the useful band for genuine strategy sitting between 1,500 and 3,000. Below roughly 1,000 euros the scope is narrow: local visibility for one location, basic technical hygiene, a modest content cadence. Above 3,000 you start buying dedicated senior time, competitive national campaigns and real coordination with your developers. The pricing guide breaks these tiers down, but the headline retainer is never the whole cost, and treating it as such is how budgets quietly overrun.

Freelancers price differently. Ongoing arrangements land between 300 and 1,500 euros per month, and day rates between 300 and 800 euros, with recognised specialists well above that. The structural advantage: nearly the whole fee buys practitioner hours. An agency must cover account management, sales, tooling and payroll overhead, so a smaller share of your retainer reaches the person touching your site. That is not a criticism, it is what pays for the breadth agencies offer. But it explains why a freelancer at 1,000 euros can out-produce an agency at that price.

In-house is where budgets are most often underestimated. A mid-level SEO manager in Western Europe costs between 45,000 and 70,000 euros in gross salary, and the fully loaded figure (employer contributions, equipment, holiday, management time, recruitment amortised over the first year) is usually 1.3 to 1.6 times that. Add a tool stack at 200 to 800 euros per month and the true annual cost of one internal person lands between 75,000 and 120,000 euros. That is several thousand euros a month before a single link is earned or a page rewritten.

Every SEO programme needs four capabilities: technical diagnosis, content strategy and production, authority building, and measurement. Very few are strong at all four. A good agency covers all four because it is several people. A good freelancer is excellent at one or two and competent at the rest. An internal generalist covers whatever their previous jobs taught them. The question is therefore not who is better in the abstract, but which of the four your site needs most over the next twelve months, and whether your model can supply it reliably.

Bandwidth is the second axis, and the more brutal. A freelancer working alone has perhaps fifteen to eighteen productive days a month across every client. If your programme needs a migration reviewed, eight articles briefed, outreach run and a monthly report built, no single person absorbs that. An agency can, because it parallelises across specialists. An internal hire has the best bandwidth per euro on paper, but only in one direction: they are entirely available to you, and unable to learn from the twenty other sites an agency sees each quarter.

There is a third and quieter axis: continuity. An agency survives an individual leaving without missing a month. A freelancer's holiday, illness or better-paying client pauses your entire programme. An internal hire walks out with two years of accumulated context about why the site is structured the way it is. None of these is disqualifying, but each should appear somewhere in writing: named substitutes in an agency contract, a documented handover clause with a freelancer, and internal playbooks so a resignation is a setback rather than a restart from zero.

A freelancer is usually the better buy when the work is deep rather than wide. A technical audit of a complex site, an international structure with hreflang problems, a migration plan, a recovery after a traffic drop: these are single-expert problems where one senior brain beats four juniors. The same logic applies whenever your budget sits below roughly 1,500 euros per month: at that level an agency can only give you junior hours, whereas an independent gives you their own. The freelancer leaderboard is a faster filter than reading portfolios.

A freelancer also wins when you already have an internal owner. If someone internally can write briefs, chase developers and approve content, you do not need the coordination layer an agency provides, and buying it means paying twice for the same function. In that configuration the freelancer works as an on-demand specialist: two days a month of senior review, a quarterly audit, a standing channel for questions. It is often the highest return per euro in search marketing, and stays underused because it does not arrive with a proposal deck.

The agency wins under the opposite conditions: broad scope, high pace, and nobody internally with time to run the programme. Multi-country rollouts, catalogues with tens of thousands of URLs and companies publishing several pieces of content a week all need parallel execution and a project manager who is not you. Agencies also carry insurance, redundancy and process, which matters more than founders expect once the site becomes revenue-critical. Country catalogues such as SEO agencies in France or agencies in Germany make the local market easier to compare on like-for-like terms.

Building in-house makes sense when SEO stops being a project and becomes a permanent function. The trigger is usually one of three things: organic search already drives a large share of revenue, your product or catalogue changes weekly so external partners are permanently out of date, or your market is technical enough that briefing an outsider costs more than doing the work. If none of those is true, an internal hire mostly buys availability you do not yet need, at a cost you will feel every month for two years.

The second condition is that you can give the hire something to do besides strategy. A single in-house specialist with no content budget, no developer capacity and no authority to change templates will spend a year producing recommendations nobody implements, then leave. Before opening the role, check that developer time is reserved each quarter, that content production is funded separately, and that the person will report somewhere with the standing to get changes shipped. The scope of a serious SEO programme is considerably wider than what a single job description covers.

Finally, be honest with yourself about the hiring risk you are taking on. Search marketing titles are unusually noisy: many candidates have run programmes that grew because the company grew, and cannot separate their own contribution from the underlying trend. Ask for a specific site, a specific period, and the Search Console evidence behind every claim, exactly as you would ask an agency. The questions used to interview an agency work almost unchanged in an interview loop for an internal role, and they filter candidates far better than a portfolio.

The arrangement that quietly works best for companies between roughly twenty and three hundred employees is not a pure model at all. It is one internal owner, junior to mid-level and therefore affordable, plus external specialists bought in for the parts they cannot do. The internal person owns priorities, keeps the roadmap alive between engagements and makes sure recommendations actually ship. The externals bring depth on demand: technical audits, digital PR, link acquisition, measurement design. Nobody pays a full agency retainer for work that is genuinely needed one week per quarter.

The hybrid also fixes an accountability problem. When a single agency owns strategy, execution and reporting, it is grading its own homework. When an internal owner sets objectives and external partners deliver against them, the reporting finally has a reader with reason to question it. Keep analytics and Search Console properties in your own accounts rather than the agency's, so the record survives any change of partner. The measurement discipline in our guide to SEO KPIs and reporting is what makes this arrangement work rather than merely look tidy on paper.

Two failure modes are worth naming before you commit to a hybrid. The first is the internal owner who becomes a message relay: if they cannot judge the quality of the work they receive, the hybrid degrades into an expensive agency with an extra layer of delay. The second is fragmentation, where four freelancers each own a slice and nobody owns the outcome. Both are solved the same way, by writing down who decides what, who reviews what and who reports on what, before anybody starts work on anything at all.

Under 1,000 euros per month, the answer is almost always a freelancer, or nothing. Spread across an agency, that budget buys a handful of junior hours and a report. Concentrated on one experienced independent, it buys a real audit in month one and steady work afterwards. Between 1,000 and 3,000 euros, the deciding factor is whether you have an internal owner: if you do, take the freelancer; if you do not, take the agency, because somebody has to run the programme and at that budget it will not be you.

Between 3,000 and 8,000 euros per month you can afford either model, so the question becomes one of breadth. Multiple markets, multiple templates, high content velocity: agency. Single market with a deep technical or authority problem: a senior freelancer plus a separate content budget. Above 8,000, a first internal hire usually starts to pay for itself and external spend shifts towards specialists. Company stage matters too: before product-market fit buy flexibility, after it buy continuity. Compare candidates on verified Search Console performance rather than on positioning, awards or proposal polish.

Whatever you choose, run the same evaluation. Ask for verified data covering twelve months, not a screenshot of a good quarter. Give the two strongest candidates temporary read-only access to your Search Console and judge the specificity of what comes back. Structure a three-month pilot with defined leading indicators and no automatic renewal, and read the contract with the standard red flags in mind. The model matters less than the discipline: a well-run freelance engagement beats a badly run agency retainer, and both beat an internal hire with no budget.