Agency, Freelancer or DIY: Who Should Build Your Business Website

We're an agency, so read this with that in mind. We've also told people to build their own site on Squarespace and meant it, and we've recommended freelancers we know over ourselves when the brief was a better fit for one person. The three routes are different products, and the useful question isn't which one is best. It's which one is right for the business you have this year.

What you're actually choosing between

DIY means a builder: Squarespace, Wix, Shopify if you're selling, WordPress.com at the hosted end. You pick a template, you pour your words and photos into it, and you pay a subscription for as long as the site exists. The tools are good now, and a careful person with taste can produce something that passes.

A freelancer is one person, usually a designer who codes or a developer who designs. In the GTA that person bills somewhere between $60 and $120 an hour if they're experienced, and often quotes a flat figure instead.

An agency is a team with a process. Senior time in Toronto runs $100 to $185 an hour, and you're paying for more than the hours: project management, a design and a development specialist rather than one person doing both, and the ability to survive somebody being on holiday. The rates are the ones from what a website costs in the GTA.

Here's the whole comparison in one place, as we read the market in 2026. Everything below it is the argument for each row.

Cost, including the cost nobody writes down

The builder route reads as nearly free and it isn't. The subscription is small, but a founder building their own eight-page site spends somewhere between forty and a hundred hours on it in our experience, spread across evenings, and most of those hours go into writing copy and fighting with a layout rather than anything that would show on a proposal. If your time is worth anything at all, the cheap route costs a few thousand dollars in the currency you have the least of.

That's often still the right trade. The point is only to run the comparison with the hours in it.

Freelancer pricing is the widest band, because "freelancer" covers a student with a Wix account and a twenty-year developer who left an agency to work alone. The second kind is often the best value in the whole market: senior work without senior overhead. The first kind is fine for what it is, provided you know which one you've hired.

Agency pricing carries the overhead, and it's worth being honest about what the overhead buys. Some of it is genuinely useful: two sets of eyes, a process that catches the redirects nobody thought about, a company that will still exist in three years. Some of it is an account manager relaying your message to the person who could have just read it. The size of the shop decides which one you're paying for.

Speed

DIY is the fastest way to something live and the slowest way to something good. A site can be up in a weekend. Whether it's finished in a month depends on whether you have the words and the photos, same as every other route.

A freelancer's timeline is usually three to eight weeks, and the honest caveat is the queue. One person can only work one project at a time in any real sense, so the start date matters as much as the duration.

Agencies take longer per project because more people touch it and each handoff costs a few days. Six to twelve weeks is typical for a custom small-business site. The trade is that the work runs in parallel: design and copy and technical setup happen in the same fortnight rather than one after another.

The quality ceiling, and why it's not the same as the floor

The ceiling on a builder is the template. You can make it yours, and you can make it good, but you can't make it do something the template doesn't do. Where that bites isn't the look. It's the structure: whether your service filters are real pages with their own addresses, whether the content is in the HTML when a crawler arrives, whether the page carries the structured data that tells Google and the AI assistants what kind of business this is. Builders have improved on all three and are still behind on all three.

A freelancer's ceiling is that person's range. A genuinely good one beats a mediocre agency every single time, at half the price. The problem is variance. The portfolio tells you about the design; it tells you almost nothing about whether they think about search structure, performance, or what happens to your old URLs. Ask them. The answer is usually clear within a sentence.

An agency has a higher floor because someone reviews the work, and a higher ceiling on paper because it has specialists. The catch is that the specialists appear in the pitch and the junior appears in the build. Ask who will actually do the work, and whether the person scoping the project is the person building it. At a large shop those are three different people. At ours they're the same one, and that's most of what we mean when we say we're small on purpose.

The risk you're taking with each

The freelancer risk is the one everyone knows. The person gets a full-time job, moves, gets busy, or stops answering. Plan for it rather than hope against it: your domain in your registrar account, your hosting in your name, the code in a repository you can log into, and a short document saying how the site is deployed. If those four things are true, a disappearing freelancer is an inconvenience. If they're not, it's a rebuild.

The builder risk is quieter. You can't take a Squarespace site anywhere. You can export the blog posts and the images, and then you start again somewhere else with them. The platform also sets the price, changes the price, and decides which features move to the higher tier. None of that is malicious, and none of it is negotiable either.

The agency risk is overhead and attention. You may be a small account at a shop whose real revenue is elsewhere, and small accounts wait. Agencies also close and get acquired, which is the freelancer risk wearing a suit, so the same four questions about ownership apply.

Who owns what at the end

The domain is yours in every case, or it should be. If a freelancer or an agency registers it in their own account "to make it easier", ask them to transfer it before launch.

Hosting on a builder is the builder, inseparably. With a freelancer it's often their reseller account, which is fine while they're around and a problem the day they're not. With an agency it's usually theirs too, sometimes with a monthly line item. Either way the contract should say the site can be moved and what that costs.

Code is the interesting one. On a builder there isn't any to own. On a custom build there is, and whether it's yours at handover is a line in the contract that almost nobody reads. Our position is on the pricing page: code, content and accounts are handed over at the end, because a price that doesn't include ownership isn't a purchase price.

What happens at year two

Year two is where the three routes stop looking alike.

The DIY site is still there and you're still paying for it. It's also drifted. The services have changed, the team photo has two people who left, and nobody has touched the blog since the launch burst. You've become the webmaster of a business you were supposed to be running.

The freelancer-built site is cheap and quick to change, provided the freelancer is still doing this kind of work and still returns your emails. When they do, a small update is a same-week job at a fair rate. When they don't, you're finding someone new to learn a stranger's build, and that person's first suggestion is usually to start over.

The agency-built site is on a maintenance plan, which costs more and is predictable. Whether that's worth it depends on how often your site changes. A restaurant that updates its menu monthly and a law firm that adds a page a year are getting very different value from the same retainer.

And across all three, the rebuild is the number to watch. The cheapest route is the one most likely to be replaced at year two, and the replacement is where the money that looked saved gets spent, often with a ranking drop on top. That's the argument in how to redesign without losing your rankings, and it's the strongest case for spending a little more once.

The decision, by situation

If you're pre-revenue or still working out what the business is, build it yourself. Genuinely. Spend a weekend on a builder, get the phone number and the map and a paragraph that says what you do, and put the money into finding out whether anyone wants it.

If you're a local service business with five to eight pages, one town, and customers who mostly come from referrals and your Google listing, a builder or a freelancer will serve you well. The gap between that and an agency build is real but it's not where your growth is coming from.

If search is where your customers come from, the structure matters, and that's where the builder column starts to cost you. A trade covering fifteen municipalities, a clinic with twelve services, a firm with nine practice areas: each of those needs a page per thing, built so the pages are found, and that's the work a good freelancer or an agency does and a template doesn't. Pick whichever one can show you a site they built that ranks for something specific.

If you're selling online with a modest catalogue, Shopify on your own is a reasonable place to start and a lot of businesses never need to leave it. Past a hundred or so products, or once filtering and search start to matter, you're in custom territory whoever builds it.

If you already have a site that ranks and you're replacing it, don't do it yourself. The design is the easy half. The migration is where rankings are lost, and it's the half a builder gives you no tools for.

If you need a login, a dashboard, a booking flow with real logic, or anything that behaves like software, you need a developer or an agency with developers. A builder plugin is somebody else's product on somebody else's roadmap.

And if you're sending paid traffic anywhere, the page it lands on matters more than the rest of the site combined. That's a small, bounded piece of work, and the kind of thing that fits a fixed-price sprint rather than a full build.

Where we sit in this

We're the agency column with most of the overhead removed. The person who scopes your site builds it, there's no account manager in between, and the sites on the projects page are hand-built in Next.js by the same small team. Landing pages and other bounded work are sprints at $1,499. Full sites, stores and apps are quoted individually, with the scope, the timeline and the number back within 24 hours.

If you're in the first situation above, a twenty-minute call will end with us telling you to use a builder, and we'd rather say so than sell you a site you'll outgrow in the wrong direction. If you're in the third or fourth, web design and development is what the page is for.