I run an agency that sells custom builds, so you should read what follows with appropriate suspicion. I am going to argue that a template is the right answer for a meaningful share of the businesses that ask me for a custom site, and that the real decision has almost nothing to do with design.
The pitch you usually hear is that templates are cheap and generic and custom is premium and unique. That framing is useless, because it compares the wrong things. A Squarespace site can look better than a bad custom build, and often does. Uniqueness is not what you are buying when you pay for custom.
You are not buying a look. You are buying the ability to change things later without permission.
What you actually buy with a custom build
Every website is a set of constraints. A template hands you a good set of constraints for free. That is genuinely valuable, and for a lot of businesses those constraints never bind. The cost of custom is paid up front. The cost of a template is paid later, and only if you hit one of its walls.
The walls that actually get hit, in roughly the order I see them:
- Conversion structure. You want the page to work differently than the template's layout allows, because your buying process is not the one the template author imagined. This is the most common wall and the most expensive one, because it silently costs you leads and nobody attributes it to the platform.
- Integration. Booking system, CRM, ERP, inventory, payment flow, or an internal tool needs to talk to the site with real data in both directions. Templates offer a fixed set of app integrations. If yours is not on the list, you are done.
- Performance at scale. A page builder ships every feature it supports whether you use it or not. On a five page brochure site nobody notices. On a two hundred page site with real traffic and Core Web Vitals mattering, the ceiling is hard.
- Accessibility. If you are held to WCAG 2.1 AA, Section 508, or a state procurement standard, you need control of the markup. Most builders generate div soup you cannot rewrite.
- Content model. You need a structured content type the builder does not have, like a product with variant specs, a property with fifteen fields, or a treatment with a price, downtime, and aftercare block. Faking it with rich text works right up until you have forty of them.
The four questions that actually decide it
Skip the design conversation entirely and answer these. They will tell you the answer faster than any agency call.
- 01Does the website produce revenue, or describe it? A site that is a brochure people check after they already decided to call you is a brochure. Buy a good template and spend the difference on the thing that actually drives the call. A site that is the primary path from stranger to booked appointment is infrastructure, and infrastructure is worth engineering.
- 02Does anything have to talk to anything? One real bidirectional integration usually justifies custom on its own, because the workaround cost of faking it compounds every month.
- 03How many unique templates do you need, and how many pages share each one? Twenty pages sharing three templates is cheap to build custom. Six pages that are each bespoke is expensive. People consistently estimate this backwards.
- 04Who changes the site in twelve months, and how often? If the answer is a non-technical person editing copy weekly, that requirement shapes the build. If it is nobody, you do not need a CMS and should not pay for one.
The honest heuristic
If three or more of those answers point to brochure, no integrations, few unique templates, and rare edits, buy a template. I will tell a prospect this on a first call and I have talked people out of projects on exactly these grounds.
The costs nobody puts in the comparison
| Cost | Template | Custom |
|---|---|---|
| Up front build | Low | High |
| Monthly platform fee | Ongoing, rises over time, non-negotiable | Hosting only, usually trivial |
| Plugin and app stack | Accumulates, each one a dependency and a breakage risk | None by default |
| Cost to change structure | Low until it is impossible | Consistent and predictable |
| Migration cost later | High, and paid at the worst possible moment | Low, you own the code |
| Performance ceiling | Fixed by the platform | Whatever you engineer |
The line item that gets people is the last one in the first column. Template migrations happen under duress, usually because the business finally hit a wall it cannot work around, and now it is paying for a custom build anyway plus a content migration plus redirect mapping plus whatever SEO equity gets lost in the move. The template was not cheaper. It was financed.
The worst option is the middle
There is a third category that is worse than either extreme: a heavily customized page builder site. A theme, plus a builder plugin, plus fifteen plugins to add the things the builder cannot do, plus custom CSS overriding all of it, plus a developer who left. This is the most common site I am asked to fix.
It has the cost profile of a custom build and the constraints of a template. Nobody chooses it deliberately. It is what a template becomes after two years of being asked to do things it was not built for. If your current site is in this state, the honest assessment is usually that fixing it costs more than replacing it, and I will say so rather than bill for the archaeology.
How we handle this in practice
Every build we ship is hand-coded, typically a React front end with semantic markup and no page builder in the stack. That is our position because most of the businesses that reach us are past the walls above. But the first conversation is diagnostic, not a pitch.
- If the four questions point to a template, we say so. We do not have a way to profit from that answer, which is exactly why it is worth asking us.
- If they point to custom, you get a written scope with a fixed price, defined exclusions, and acceptance criteria before anything starts.
- If your team specifically requires WordPress because that is what they can maintain, we build it properly with a custom theme rather than assembling it from a builder.
- You own the code either way. That is in every contract we write.
If you want the diagnostic without the sales call, the free website audit covers the same ground in writing.