Hiring

Freelancer or Agency: Who Should Build Your Website?

The freelancer-versus-agency debate is usually argued by people with an obvious stake in the answer. The useful framing is not which is better — neither is — but what you are actually buying with the price difference. An agency costs more because it is selling coordination. If your project needs coordinating, that is money well spent. If it doesn't, you are paying for a project manager to attend meetings about your five-page website.

What are you actually buying in each case?

A website needs several distinct skills: structure and strategy, visual design, writing, building, and search setup. The question is only ever how those skills get assembled.

A solo freelancer does all of it, usually well in two or three of those areas and adequately in the rest. You get one head holding the whole picture, which is a real advantage — the person choosing the layout is the person who heard you describe your customers. You also get one person's limits.

A small studio of two to five people splits the work — often a designer-builder plus someone who writes, with a specialist pulled in occasionally. You keep most of the directness while covering more skills. This is where most competent South African small-business web work happens.

An agency of ten or more adds a layer whose job is coordination: account managers, project managers, process. That layer exists because at a certain size, four specialists working on one project without coordination produces chaos. It is genuinely valuable — on projects big enough to need it.

So the honest question is not "who is better." It is: how many people need to touch this project at once? If the answer is one or two, coordination is pure cost.

How do they compare on the things that matter?

This is the same trade-off table from our guide to hiring a web designer in South Africa, extended with what actually goes wrong in each case.

Notice the failure modes are different in kind, not severity. A freelancer's risk is that they disappear. An agency's risk is that you are deprioritised behind clients paying ten times more — which does not feel like a disaster, it just feels like everything taking three weeks.

When should you choose a freelancer?

Your site is under about ten pages and mostly informational. Services, about, contact, a few proof pages. One capable person builds this comfortably.

You want to talk to the person doing the work. Something is lost in translation every time your brief passes through another human. If you know your customers well and can describe what you need, that directness is worth real money.

You need it soon. A freelancer can start on Monday. An agency starts after onboarding, a kick-off meeting and a slot in the schedule.

Budget is the binding constraint. Not because cheap is good, but because a finished freelancer site beats an unfinished agency one. Sites that never launch earn nothing.

Your business is a trade or local service. The things that make these sites work — a phone number that is impossible to miss, WhatsApp on every page, fast loading on mobile data, honest photos of real work — are execution details, not coordination problems. It is why a good electrician's site often comes from one person who understood the job, not a team.

The freelancer risk is real, though, and it is worth managing rather than ignoring: own your domain, own your hosting, get the files at handover, and pay in stages. Those four habits reduce "my designer disappeared" from a catastrophe to an inconvenience. More on that in red flags when hiring a web designer.

When is an agency genuinely worth it?

When several specialists must work at once. An online store with stock management, payment gateway, delivery rules and a product photographer is four jobs running in parallel. Somebody has to hold that together, and that somebody is what you are paying for.

When the website is one channel of several. If the same campaign has to land on the site, in email, on paid ads and in print simultaneously, coordination stops being overhead and starts being the product.

When continuity is a requirement. Some businesses genuinely cannot accept the risk of one person being unavailable — a medical practice with online bookings, say, or anything where downtime costs money hourly. An agency gives you an organisation to hold accountable rather than a person.

When you need someone else to run it. If you have no time to make decisions, an account manager chasing you is worth paying for. This is an honest and underrated reason.

What is not a good reason: believing agency work is inherently better designed. Much agency output for small clients is built by a junior from the same templates a freelancer would use, with more people billing hours against it. If you go this route, ask directly who will build your site and ask to see their work, not the agency showreel.

How do you hire each one well?

The checks differ, because the risks differ.

Hiring a freelancer, ask: How many projects are you running right now? What happens if you get sick mid-build? Will the domain and hosting be in my name? Can I speak to a client from last year? Do you write the copy, or do I? A freelancer who answers all five clearly is usually a safe bet — and the capacity question is the one that predicts the most.

Hiring an agency, ask: Who specifically will build this — can I see their work? Who is my day-to-day contact, and what happens when they go on leave? What is the minimum billing increment for small changes after launch? How many clients does my account manager handle? Is there a monthly retainer and can I leave it? The last one matters: some agency pricing looks reasonable until you find the site is tied to a contract you cannot exit without losing hosting.

For both: get scope, ownership, revisions, timeline and support in writing, and compare quotes on identical line items rather than headline price. The method is in how to compare website quotes without getting burned, and the full question list is in 12 questions to ask before you pay a deposit.

One last thing worth saying plainly. The single best predictor of a good website is not the size of who built it. It is whether whoever built it understood what a visitor is supposed to do on the page, and made that thing obvious. That understanding is not correlated with headcount at all — which is why the choice matters less than most of the advice on this topic pretends.