The right answer depends on whether the work ends
Most businesses pick a web vendor the way they pick a restaurant after food poisoning: by avoiding the last one. Burned by an agency, they hire a freelancer. Burned by a freelancer who disappeared, they try to hire in-house. Each model fails in its own predictable way, and the useful question is not which is best but which matches the shape of your work. Here is the honest version, including where we are the wrong choice.
Ask first whether the work ends
Some web work has a finish line: a site gets built, launched, and then changes a few times a year. Other work never ends: a store with weekly promotions, a content operation, a product that ships features. The first shape is a project. The second is a function.
Projects suit vendors, because you are buying a defined outcome and the relationship can go quiet without anything breaking. Functions suit either an ongoing retainer or an internal hire, because the work arrives every week whether or not anyone is available. Businesses get into trouble when they buy a project and actually needed a function, or hire a person for a function that turns out to be a project with a long tail.
- Project: defined outcome, ends, resumes occasionally
- Function: arrives continuously, has no natural end
- Buying the wrong shape is more expensive than picking the wrong vendor
- Ask how many times per month the site actually needs to change
What each model is genuinely good at
A freelancer gives you the lowest cost per hour, direct communication with the person doing the work, and speed on well-defined tasks. That is a real advantage and it is why so much good work gets done this way. An agency or studio costs more per hour and buys you continuity, a range of skills under one contract, and someone still reachable when the individual who wrote the code is unavailable. An in-house hire gives you availability and accumulated context nobody else will have, at the cost of payroll, benefits, and the burden of keeping one person current across design, development and marketing.
Note the pattern: you are mostly paying for coverage and continuity, not for talent. The best freelancer and the best agency developer are the same kind of person. What differs is what happens on the day they are not there.
The failure mode of each, stated plainly
The freelancer failure is availability. They take a full-time role, get busy with a bigger client, or simply move on, and your site becomes unmaintainable at the worst possible moment. The mitigation is not trust; it is documentation and accounts in your own name from day one.
The agency failure is distance and pace. You are one account among many, the person who understood your business rotates off, and small changes take longer than they should. Ours included: a remote international studio is not the right fit for a business that wants someone in the room on Thursday. The mitigation is asking who specifically does the work and what the turnaround is for a small change, in writing.
The in-house failure is scope. One person is expected to cover design, development, SEO, analytics and content. They will be genuinely good at one or two of those, and the rest quietly does not happen. The mitigation is to hire for the part you need most and buy the rest.
- Freelancer: availability risk, mitigated by documentation and account ownership
- Agency: distance and pace, mitigated by naming who does the work and the response time
- In-house: scope too wide for one person, mitigated by hiring narrow and buying the rest
- Ask any vendor, including us, what happens if the person doing your work is unavailable for two weeks
The clause that matters more than the model
Whichever you choose, the thing that determines whether you are stuck is ownership. The domain, the hosting account, the code repository, the analytics property and the ad accounts should be registered to your business, with the vendor holding access rather than title. This costs nothing to set up at the start and is the difference between changing vendors in a week and rebuilding from scratch.
Ask for it explicitly before signing, from an agency, a freelancer, or a contractor working alongside your in-house hire. A vendor who resists is telling you something useful. And if you are hiring in the US as an employee rather than a contractor, the classification question is a payroll and tax matter worth putting to your accountant or employment counsel rather than resolving from a blog post.
- Domain, hosting, repository, analytics and ad accounts in your business name
- Vendor gets access, not ownership
- Written turnaround for small changes, not just for the build
- Employee versus contractor classification is a question for your accountant or counsel
Frequently asked questions
We had a bad experience with an agency. Does that mean we should go freelance?
Not by itself. Look at what actually failed: if it was pace and access, a freelancer may well be better. If it was continuity or breadth, going freelance makes the same problem worse. Diagnose the failure before switching models.
Is it cheaper to hire someone in-house?
Rarely at first, once you count payroll taxes, benefits, equipment and the ramp-up period. It becomes cheaper when the work is genuinely continuous and would otherwise be a large ongoing retainer. Compare against a year of actual work, not against a single project.
Can we mix models?
Yes, and it is common: an internal person who owns day-to-day content and a vendor for the build and the technical work. It works when the split is written down. It fails when both assume the other is handling something.
When are you the wrong choice?
When you want someone physically present, when the work is a few hours a month of small edits, or when you need same-day turnaround on a schedule. We work remotely and internationally, and we would rather say so than sell you a fit that is not there.
Not sure whether your next step is a vendor, a hire, or both? Tell us how often your site actually needs to change and we will tell you honestly which model fits.