Every supplier will tell you they combine the best of all the options. It is more useful to understand what each shape is structurally good and bad at, because those tendencies are baked into how the business makes money and no amount of good intent overrides them.
Three shapes, not two
The usual framing is freelancer versus agency, which misses the middle. There are three common shapes and they behave quite differently: an individual working alone, a small studio of a few people who all build, and an agency with separate account, design and delivery functions.
We are the middle one, so read the middle section with appropriate suspicion. The other two are described as fairly as we can manage.
The freelancer
Good at: price, directness and speed on well-defined work. You talk to the person doing the job, decisions happen in one message, and there is no overhead to fund. For a small, clearly-scoped site this is frequently the best value available anywhere.
The risk is capacity, not competence. Freelancers are usually good at what they do. What they cannot do is be in two places at once, and your project's timeline is exposed to their illness, their holidays and their other clients. If they stop replying, there is no organisation behind them to escalate to.
Ask about: what happens if they are unavailable mid-project, who else could pick the work up, and whether the code and accounts are somewhere you could hand to a successor.
The agency
Good at: continuity, capacity and process. People leave and the work carries on. They can run several strands at once, they have seen your problem before, and there is a contract with a company behind it rather than a person.
The risk is distance. The person who sold you the project is usually not the person building it, and information degrades on the way through. Small changes acquire a process. And an agency's overhead has to be covered, which for a small local business often means you are the smallest client in the building — where the attention goes when something urgent lands elsewhere is not a mystery.
Ask about: who specifically is building it, whether you can speak to them directly, and whether any of the work is subcontracted.
The small studio
Good at: the middle ground — you speak to whoever is building, and there is more than one person. Scoping is quick because the person estimating is the person doing it, which is also why the estimates tend to hold up.
The risk is scale. A studio has real limits on how much it can take at once, and a genuinely large project may be better placed elsewhere. Ours is the shape we chose deliberately, but it is a shape with a ceiling and it would be dishonest to describe it as strictly better than the other two.
Ask about: current capacity, what else is in flight during your build, and what happens to your site after launch.
Questions to ask before you sign
These work on all three, and the quality of the answers separates suppliers faster than a portfolio does.
Who owns the code and the accounts when this is finished? The answer should be you, unambiguously, including the domain, the hosting account and the source code. Anything else is a lock-in you would be agreeing to on purpose.
Who am I speaking to when I need a change in six months? You want a name and a route, not a general assurance about support.
What is not included in this price? Domain, hosting, third-party services, content, photography, ongoing changes. A supplier who has thought about your project can list these without hesitating.
What would you not build for me? A good answer here is the strongest single signal you will get. Anyone who wants everything you mentioned is selling, not scoping.
Can I see something you built that is still live and still being used? Portfolios show launch-day screenshots. A site that has been running for two years and still works is a different claim entirely.
What happens if it takes longer than you estimated? The honest answer involves who absorbs the cost. Ask it before you need to know.
Answers that should end it
"We guarantee you'll rank first on Google." Nobody controls ranking. Google's own guidance on hiring an SEO lists a guaranteed number one position as a reason to walk away, and it is the clearest single disqualifier in this whole process.
"We have a special relationship with Google." No such thing exists, and offering it as a selling point means either they do not know that or they assume you do not.
Vagueness about ownership. If it takes more than one sentence to establish that you own your own domain and code at the end, you have your answer.
A quote with no written scope. A price without a page list and a definition of done is not a quote, it is an opening position. Everything after that is a negotiation you are structurally set up to lose.
Pressure to decide today. Websites are not perishable. Any urgency in the sales process is manufactured.
The short version
For a small, clearly-defined site, a good freelancer is usually the best value on the market and the main thing you are buying is their availability. For a large programme with several strands, an agency's continuity is worth its overhead. In between — a site that has to do a specific job, where you want to talk to whoever is building it — a small studio fits.
Whichever you pick, the questions above matter more than the category. A freelancer who answers all six well is a safer bet than an agency that answers four.
Put those questions to us
Ask all six. We publish our prices and our answer to the ranking question is on every relevant page, so you can check most of them before you get in touch.
Start a conversation