How to price a website

Price the scope, not the website

There is no market rate for "a website" because no two are the same job. A five-page brochure site and a headless Shopify build with a subscription engine are both "websites" and are worlds apart in cost. So the first rule of pricing is: you can't price what you haven't scoped. If you're quoting from a vibe, you're quoting your own risk.

That means pricing starts with the scope of work. Once the deliverables, assumptions and exclusions are clear, the number almost falls out of them. Price the defined thing, and the defined thing protects your price.

Estimate from the section tree

The reliable way to reach a figure is bottom-up: break the build into sections, estimate hours for each, and let the total build itself. Guessing a round number top-down ("about £8k feels right") hides where the cost actually is and gives you nothing to defend if the client pushes back.

For each section, estimate the hours honestly across the real phases, not just build, but design, content fill, testing and revisions, which are where estimates quietly blow out.

• Design, layouts, states, responsive behaviour.
• Build, templates, components, integrations, data.
• Content and testing, population, cross-browser checks, accessibility, launch.
• Project overhead, communication, project management, sign-off cycles. Real hours; charge for

them.

Multiply hours by your rate, and you have a cost per section that rolls up into a total you can explain line by line. This is exactly how ScopeDeck's live pricing works, each section carries a cost and internal hours, and the running total updates as you build the scope, so the estimate is a by-product of scoping rather than a separate spreadsheet.

Choose a pricing model

Once you have an internal cost, decide how to present it. The two main models suit different work:

• Fixed price, one number for a defined scope. Clients prefer it because it caps their risk.

You should only offer it when the scope is genuinely clear, because you're the one absorbing the overrun if it isn't.

• Hourly or day rate, you bill for time spent. Fairer on genuinely open-ended work, but

clients dislike the open meter and it rewards slowness.

Most website projects are best sold fixed price on a tight scope, with anything uncertain handled as optional extras or a separate discovery phase. For the full trade-off, read fixed price vs hourly.

Add contingency before you're forced to

Every project has unknowns: a fiddly integration, a migration with dirty data, a client who needs an extra round. A contingency is a small, deliberate buffer, often 10–20% depending on risk, added to cover them. Add it while you're calm and pricing, not later when you're over budget and having an awkward conversation.

Contingency isn't padding you hide. On a fixed price it can sit inside the number; on a more open engagement it can be a named line the client understands. Either way, it's the difference between a profitable project and one you finish resenting. See contingency in a quote for how to size and word it.

Present a range, then a fixed figure

Early on, before the scope is nailed, an honest range is better than a false precision. "£18,000 to £24,000 depending on the integrations we confirm in discovery" tells the truth and anchors the client, without committing you to a number you'll regret. As the scope firms up, the range collapses to a fixed figure.

This mirrors how ScopeDeck handles it: the Quote can show a rounded lower/upper bracket while the Specification locks to a single agreed figure, early uncertainty becomes one fixed build number without you re-quoting from scratch. For the technique of quoting a range that still commits you, read how to give a quote range.

One more presentation rule: don't itemise every line. A quote broken into forty micro-charges invites the client to line-item-veto their way to a cheaper, worse project. Group into meaningful blocks. See why not to itemise every line.

What to avoid

• Quoting before scoping. The most expensive mistake. Scope first, always.
• Charging for build but not for management, testing or revisions. Those hours are real. Price

them or absorb them.

• A single number with no buffer. One surprise and your margin is gone.
• Racing to the lowest price. You rarely win the good clients on price, and you always lose on

the bad ones. Compete on clarity and confidence instead.