SEO Agency vs Freelancer vs In House: Honest Cost and Risk Comparison
Three models, and the published comparisons all leave out the two costs that decide it: the hours you spend managing the arrangement, and what it costs you when it ends.
Nearly every SEO agency vs freelancer comparison is published by one of the three, concludes in its own favour, and compares only the invoice. The invoice for SEO work is the smallest part of the decision. What actually separates these models is how many of your own hours each one consumes, what happens when the person leaves, and who carries the risk when something goes wrong.
What the published numbers say, and why they disagree
Three reasonably credible datasets exist and they do not agree, which is useful rather than annoying once you see why.
Ahrefs surveyed 439 providers in 2024: average monthly retainer around 2,900 US dollars, agencies around 3,200, freelancers around 1,350, and providers with under two years of experience averaging roughly half what those with five to ten years charged. Clutch directory data puts the average retainer around 3,200 a month, with typical engagements running about a year. SE Ranking surveyed 260 providers in December 2024, most of them very small operations, and found 64 per cent charging under a thousand a month.
The contradiction resolves cleanly. The higher figures describe several people doing content, technical and authority work in parallel. The lower figure largely describes one person doing a slice of it. Both are real products. They are not the same product at different prices, and the most common buying mistake is treating them as if they were.
The total cost, including the parts nobody invoices
This is the table the comparisons leave out. Figures in US dollars per month, for a small or mid sized business.
| Cost | Freelancer | Agency | In house |
|---|---|---|---|
| Fee or salary | 500 to 2,500 | 1,500 to 8,000 | Mid level salary spread over the year, plus employer costs |
| Tools | Often yours to buy | Included | Yours, 300 to 1,000 or more |
| Your management hours | 4 to 10 a month | 2 to 5 a month | 6 to 15 a month, more at first |
| Work they cannot cover | Usually one or two of the three disciplines | Usually all three | One person covers one or two; the rest gets contracted out anyway |
| Time to first real output | 2 to 4 weeks | 3 to 6 weeks | 2 to 4 months including hiring |
| Cost when it ends | Low, if you hold the accounts | Low to moderate, depending on the contract | High: recruitment again, plus months of lost momentum |
Two rows deserve attention. The management hours row is the one that catches people out: a cheap freelancer who needs ten hours a month of your time is not cheap if your time has a value. And the last row is why in house looks better on a spreadsheet than it behaves in practice, because a single hire who leaves after fourteen months takes the context with them.
The risk each model carries
Not advantages and disadvantages. Specific failure modes.
Freelancer: key person risk. One person means no cover for illness, holidays or a better offer. It also means the scope is capped by what one human can do, and the scope of this work has widened considerably in the last three years. Someone genuinely strong at technical work is rarely also a strong writer and an outreach specialist. The upside is real: you talk to the person doing the work, every time, which is worth more than most agencies admit.
Agency: dilution risk. You are buying a share of several specialists, and the question is how large a share. A specialist carrying twelve accounts gives yours a few hours a month, which may be plenty or may be nothing. The other agency risk is that the person selling you is not the person doing it, so the quality you were shown and the quality you receive can differ. The upside is capacity and continuity: somebody covers when somebody else is away.
In house: scope risk. One employee cannot cover content, technical work and authority building, so you end up contracting part of it out and now you are managing both. Ramp up is also slow, typically six to twelve months before an internal hire is producing at their level. The upside is substantial and often decisive: an employee learns your business properly, which is the input no external provider ever fully gets.
Who should pick which
A freelancer, when you know which specific problem you have and want that fixed well. A technical audit and remediation. A local profile and review programme. A content writer who understands your trade. The model works best with a clear brief and worst as a general retainer for everything.
An agency, when you do not know where the bottleneck is, or you know it is in more than one place at once. Also when continuity matters more than price, or when an engagement failing quietly for three months would cost you more than the fee difference.
In house, when search is a primary channel rather than one of several, when you have enough volume to keep a specialist busy, and when the domain knowledge is hard to transfer. Below a certain scale this is the most expensive option by a wide margin once you count ramp up and turnover.
A combination, when you can afford it, which more businesses can than realise. One internal person who owns the strategy and knows the business, plus contracted specialists for technical work and outreach, is often the best value arrangement available and almost never appears in these comparisons because nobody sells it.
The threshold worth knowing
A rough rule that holds up: below roughly 2,500 US dollars a month, an agency is usually the wrong shape. At that level the fee is split across several people, each of whom gives you a few hours, and nothing receives enough attention to move. The same money spent on one capable freelancer with a narrow brief tends to produce more.
Above that, the agency case strengthens quickly, because the binding constraint stops being money and becomes how many kinds of work can happen at once.
What has changed since 2024
One development genuinely affects this choice. The scope of the work has widened. Three years ago a competent generalist could cover most of it. Now the same brief includes technical health, content, authority, and making sure your business gets named when an assistant answers a question in your field.
That stretches the freelancer model. It is worth asking any single operator directly which parts they do not do, and taking a clear answer as a good sign rather than a weakness.
There is a second change, less comfortable. Cheap content is now often generated rather than written, and Google has an explicit policy on content produced at scale with no added value. A low fee that includes a high page count is a combination worth asking hard questions about, whichever of the three models is quoting it, because the penalty risk lands on your domain.
Five questions that apply to all three
- Which of content, technical and authority will you personally cover, and which will not get done?
- How many hours of my time per month does this need, and from whom?
- Will every account be in my name with me as owner?
- If this ends, what do I keep, and what do I have to rebuild?
- Who covers the work when you are unavailable for three weeks?
Question five is the one that separates the models most sharply, and the one least often asked.
Different rather than riskier. A freelancer concentrates the risk in one person and removes the risk of being handed to a junior. An agency spreads cover and introduces the possibility of your account being the least interesting one on somebody list of twelve.
Ask for two clients you can contact, ask to see one Search Console account with permission, ask what they do not do, and ask about a project that went badly. The last question is the most useful one, in this and every hiring decision.
Occasionally, and usually as a hybrid rather than a full hire. Someone who already works for you, given a day a week and a budget for contracted specialists, often beats all three pure models, because the hardest input to buy is knowledge of your own business.
Yes, and it is a sensible order if the accounts are in your name and the work is documented. Insist on both from the first week, and the transition costs you a handover meeting rather than a restart.
Want this checked on your own site?
Send the URL and we will tell you what we would do first, in writing, before anybody talks about a retainer.
- A written audit within 48 hours
- No call required and no obligation
- If search is the wrong spend for you, we will say so
Free site audit
Your URL and what you want more of. That is all we need to start.
Request received!
Thanks! We have your request. Check your inbox within 48 hours for your written proposal. Nothing else needed from you.
