How to Write a Website Brief Your Developer Can Actually Use
You asked three firms for a website quote and got back three numbers that can’t be compared, built on three different assumptions about what you want. That isn’t the developers being evasive. It’s what happens when the brief is a paragraph of email and everything else is inferred.
A website brief is the cheapest document you’ll ever write and the one that most reliably prevents a bad project. It doesn’t need to be long, and it should not contain design instructions. It needs to state what the site is for, what it must do, and what you’ll accept as finished.
Why does a vague brief cost you money?
When a brief leaves a gap, somebody fills it. Either the developer makes an assumption and prices it, or they price the smallest plausible reading and discover the gap in week three. Both routes end in a change request, and change requests mid-build cost more than scope agreed up front because the work already done has to be unpicked.
The second cost is comparison. If one firm assumes you’re writing the copy and another assumes they are, their quotes differ by thousands and neither is wrong. You can’t evaluate proposals that answer different questions, so you end up choosing on price or on rapport rather than on fit.
The third cost is the one nobody invoices for: a site that launches and doesn’t do the commercial job, because the job was never written down.
What does a developer actually need from you?
A usable brief answers six questions. If yours covers these, most competent firms can quote it accurately.
- What is the site for, commercially? One sentence. “Generate qualified enquiries from manufacturers in southern Ontario” is a brief. “Modernise our online presence” is a wish.
- Who is reading it, and what are they deciding? Name the two or three audiences and what each needs before they’ll contact you. A procurement officer and a founder want different proof.
- What does a visitor do when convinced? Call, submit a form, book a slot, request a quote, download something. This single answer determines half the architecture.
- What pages and content types exist? Not a design, a list. Services, case studies, team, blog, resource downloads, careers. Say roughly how many of each and whether they grow.
- What has to integrate? CRM, booking tool, payment processor, email platform, accounting software, an existing member database. Name the products.
- Who owns content, and when does it arrive? The honest answer here is the single best predictor of whether the project lands on time.
What should you deliberately leave out?
Briefs get worse when owners write solutions instead of problems. Resist specifying the mechanism when what you actually have is a requirement.
- Don’t prescribe the technology unless you have a real constraint, like an existing team that only knows one platform, or a corporate standard. Otherwise let the firm propose and justify.
- Don’t specify design details. “A full-width hero video, then three icon columns” is a layout you saw somewhere. Describe what the page has to achieve and let the designers solve it.
- Don’t send only competitor links. One or two references are useful if you say what you like about them and why it matters to your buyer. A folder of screenshots with no reasoning produces imitation.
- Don’t hide the budget. Withholding it doesn’t get you a better price, it gets you proposals aimed at the wrong tier. Our published pricing starts at CAD 500 for a Starter Website of up to five pages and CAD 1,500 for a Business Website of up to fifteen pages, with web applications quoted after a call — a range you can position yourself against before you write a word.
How do you write the requirements so they’re testable?
The difference between a brief that protects you and one that doesn’t is testability. A requirement you can’t check at handover is a requirement you’ll argue about.
Convert each intention into something observable. “The site should be fast” becomes “service pages load the main content in under two seconds on a mid-range Android phone on a typical mobile connection.” “It should be accessible” becomes “templates meet WCAG 2.0 Level AA, and the enquiry form is completable by keyboard alone with visible focus.” “The team should be able to update it” becomes “a named non-technical staff member can publish a new service page, with images, in under thirty minutes without developer help, demonstrated at handover.”
That last pattern — a demonstration at handover rather than an assurance during the pitch — is worth applying to anything you care about. It costs the developer nothing if the work was done properly, and it’s the cheapest leverage in the document.
What about search, content and the parts after launch?
Two things routinely fall out of briefs and then become disputes. The first is content. State explicitly who writes the copy, who sources images, and whether existing pages are being migrated, rewritten or dropped. If you’re migrating an existing site, say so plainly and ask how redirects will be handled, because that’s where search traffic gets lost.
The second is what happens on day two. Name who holds the hosting account, the domain registrar and any licences — you should, in your own name. Say whether you want a maintenance arrangement, what response time you expect when something breaks, and who’s responsible for updates. A site is an operating asset, not a delivery.
It’s also worth stating how you want the site to behave for machines, not just people. If being quotable by AI answer engines matters to your category, say it in the brief so structure and schema are scoped rather than retrofitted. That’s the thinking behind our AI visibility work, and it’s far cheaper designed in than bolted on.
What does a good brief look like on the page?
Two pages. Headings, short paragraphs, bullet lists. No design mood board required. A workable skeleton:
- Business context — what you do, who buys, what’s changing.
- Objective — the commercial job, in one sentence, with how you’ll measure it.
- Audiences and their decisions — two or three, with what each needs to see.
- Primary action — the one thing a convinced visitor does.
- Scope — page and content types, approximate counts, what grows.
- Integrations — named products.
- Requirements — performance, accessibility, editability, each stated testably.
- Content ownership and dates — who, what, when.
- Budget range and target launch — honestly.
- Post-launch — ownership of accounts, maintenance expectations, support response.
Send the same document to every firm you’re considering. Unequal information produces unequal proposals, and you lose the only real benefit of asking more than one. When the responses come back, the useful signal isn’t the price — it’s which firm asked the sharpest questions about the parts you left thin.
If you’d rather not write it cold, our web development and web design pages set out how we scope, and a short call usually produces a better brief than a blank page does. Start a project and we’ll work through the six questions with you.
Frequently asked questions
How long should a website brief be?
Two pages is usually right for a small or mid-sized business site. Long enough to answer the six core questions and state your requirements testably, short enough that everyone quoting actually reads it. Length beyond that tends to be design preference, which belongs in the design phase, not the brief.
Should I include my budget in the brief?
Yes, as a range. A budget doesn’t invite overcharging so much as it lets firms propose the right scope rather than guessing at a tier. Without it you’ll receive proposals built on different assumptions about ambition, which you can’t compare meaningfully.
Do I need wireframes or a sitemap before I brief a developer?
No. A list of page and content types with rough counts is enough, and that’s genuinely useful. Wireframes drawn before anyone has thought about the conversion path tend to constrain the design without improving it, and a good team will produce structure as part of the work.
What if I don’t know what pages I need?
Then describe what your buyer needs to believe before they’ll contact you, and let the structure follow from that. The page list is an output of the audience question, not an input to it. If you can answer who’s reading and what they’re deciding, the architecture is largely determined.
Get the next one in your inbox.
Practical insights — no fluff, straight to your inbox.
Or follow us on LinkedIn:
Follow StrategyPeeps






