Every founder eventually runs the same decision.
Hire a freelance Framer expert. Hire an agency. Or hire someone in-house and keep it forever.
The advice you get depends entirely on who you ask, and everyone answering has a stake. So let me argue all three sides properly, including against myself.
The freelance expert
What you are buying: one person's judgment, applied directly.
The pitch is speed. No account manager, no weekly status ritual, no game of telephone between the person who understood your business and the person building the page. You explain the problem once. The person who heard you is the person who ships it.
That compression is the whole value. Most of the waste in agency work is not laziness — it is translation loss. Every handoff between strategist, designer, and developer costs a little context, and by the third one the page is a reasonable interpretation of a misremembering of what you said.
Where it breaks: capacity and continuity.
One person has a ceiling. If you need a site, a brand system, motion work, illustration, and copy simultaneously by Friday, one person cannot do it, and any freelancer who says otherwise is going to disappoint you politely in about four weeks.
There is also the bus problem. If your only builder gets sick, has a family emergency, or takes a full-time job, you stop. Good freelancers mitigate this by building things anyone can pick up — clean structure, documented components, no clever tricks that only make sense to the author. Ask about that explicitly. It is the single best predictor of whether you will be fine without them.
Best when: the scope is clear, the timeline is real, and you want the person who understands the problem to be the person solving it.
The agency
What you are buying: capacity, coverage, and someone else's project management.
Agencies are genuinely better at some things and it is dishonest to pretend otherwise. Parallel workstreams. Deep specialization — a real motion designer, a real copywriter, a real SEO person, not one generalist doing a passable version of each. Continuity when someone leaves. And a contract with a company rather than a person, which some boards and procurement departments simply require.
If you are running a rebrand across a site, a product, an app store presence, and a conference booth at the same time, an agency is the right call. That is a logistics problem as much as a design one, and logistics is what they sell.
Where it breaks: cost structure and distance.
You pay for the layer that coordinates the work, and that layer is real money. Depending on the shop it can be a third or more of the invoice before anyone opens a design file.
The more common problem is seniority drift. You are sold by the person whose taste you liked and delivered by whoever is available. This is not malice, it is scheduling — but it means the work you saw in the pitch and the work you receive may come from different hands. Ask who is actually building it and get the name in writing.
Best when: the scope is broad, the deadlines are simultaneous, or your organization needs a vendor rather than a person.
In-house
What you are buying: permanent context.
The strongest case for in-house is memory. A designer who has been with you two years knows why the pricing page is worded carefully, which customer complaint caused the onboarding rewrite, and what the CEO will reject before it gets drawn. That knowledge compounds and no external hire can shortcut it.
You also get availability. Small changes stop being a scoping conversation and start being a Slack message.
Where it breaks: utilization and range.
Websites are lumpy work. Intense for two months, then quiet. A full-time designer costs the same in the quiet months, and if there is not enough real work they will either leave or start redesigning things that were fine.
Range is the other issue. Someone excellent at product design is often mediocre at marketing sites — different craft, different objective, different instincts about persuasion. Hiring one person for both usually means getting one of them well.
And the true cost is not the salary. It is salary plus employer contributions plus tooling plus recruiting plus the three months before they are productive plus the risk that the fit is wrong.
Best when: design work is continuous, the domain is complicated, and you are building something long enough that institutional memory beats raw speed.
The version that usually works
Most companies I see do best with a combination, and they arrive at it accidentally after getting it wrong once.
Early on, before you know what your site should say, an external specialist is right. The problem is not capacity, it is clarity, and clarity comes from judgment rather than headcount.
Once the site is alive and changing weekly, a retainer beats a project. You stop paying to re-explain your business every time something needs to move.
Once design work is genuinely constant across product and marketing, hire in-house — and keep a specialist on call for the spikes. That is the arrangement I end up in with most long-term clients, and it works because nobody is pretending one model covers everything.
The costs nobody quotes
Three things sit outside every proposal and they decide whether this goes well.
Your time. The cheapest option on paper is the one that needs the most supervision. If you are reviewing four rounds of work that missed the point, you paid in the currency you had least of.
The rebuild. A site built without structure gets replaced, not extended. Figure out at the start whether you are buying an asset or a placeholder, and be honest about which one the budget supports.
Editability. If every copy change requires a contractor, you will stop making copy changes. Then your site slowly becomes a description of a company you used to be. This is the most expensive failure of the three and it never shows up as a line item.
How to actually decide
Ask what your constraint really is.
If the constraint is clarity — you do not yet know what the site should say — hire the sharpest individual you can find. This is a judgment problem.
If the constraint is volume — you know exactly what you need and there is a lot of it — hire the agency. This is a throughput problem.
If the constraint is continuity — the work never ends and context keeps getting lost — hire in-house. This is a memory problem.
Most people misdiagnose this.
They have a clarity problem, buy throughput, and end up with a large quantity of confident work pointed in the wrong direction.
FAQ
Is a freelancer risky for an important launch?
Somewhat, and you manage it the same way you manage any single point of failure: check that the work is structured so someone else could pick it up, and agree on what happens if a date slips. A freelancer who has an honest answer to that question is safer than an agency that avoids it.
What does each option cost?
Freelance project work and retainers land in fairly predictable ranges — I break them down in what it actually costs to hire a Framer expert. Agencies typically run two to four times higher for comparable scope. In-house costs the salary plus roughly thirty to forty percent on top, before tooling.
Can I start freelance and move in-house later?
Yes, and it is the sensible sequence. Get the system built by someone who has built many, then hire someone to run it. Just make sure the handover is a real deliverable in the contract, not a favor you ask for at the end.
How do I evaluate any of them?
Same filter regardless of model: ask why their best work is shaped the way it is. Answers about the reader's decision sequence are good. Answers about what looks current are not. More on that in How to Hire a Framer Expert Without Paying for Pretty Screenshots.
If you want a straight opinion on which of the three fits your situation — including when the answer is not me — tell me what you are working on.
