The short version
- A good brief describes the problem and the constraints. It does not describe the solution.
- The most useful section is the one most briefs leave out: what success looks like, in numbers.
- Naming your budget range gets you better proposals, not worse prices. Withholding it gets you three quotes for three different projects.
- Half a day writing the brief saves weeks of ambiguity later.
Most briefs we receive fall into one of two categories. Either they are two paragraphs asking for a price, which forces the agency to guess, or they are a forty page specification written by somebody who has already designed the solution and now wants a quote for building it.
The useful version sits in between. It gives an agency enough to think properly and leaves room for them to tell you something you had not considered. Here is what to put in it.
Start with the problem, not the project
The opening should explain why you are doing this at all, in plain commercial terms.
Not: “We require a new website with a modern responsive design.” That says nothing.
Instead: “Our enquiry volume has fallen for two years and we cannot tell which marketing produces customers. The site is difficult to update so it has not changed since 2019, and our sales team does not use the enquiries it generates because the quality is poor.”
The second version tells an agency what to solve. It also tells them, indirectly, that this is not really a design project. Some of the responses you get back will be more useful than anything you would have specified yourself.
Say what success looks like
This is the section most briefs omit and the one that changes proposals the most. Be specific and be numerical where you can.
- What should be measurably different twelve months after launch?
- Which number, if it moved, would make this project obviously worth doing?
- What would count as a failure, other than being late?
- Who in the business decides whether it worked?
If you cannot answer these, that is worth knowing before you commission anything. It usually means the measurement needs building before the project does.
Give it to a colleague who was not in the planning meetings. If they can explain back what problem is being solved and how you will know it worked, the brief is doing its job.
The sections worth including
About the business
What you sell, who buys it, how the sale actually happens, and what your sales cycle looks like. Include the awkward parts. An agency that knows your busy season is April, or that your best customers come from one industry, will propose something better.
Audience
Who needs to use this and what they are trying to do. Distinguish between the person who buys and the person who uses. For a website, distinguish between prospects, existing customers and, if relevant, recruitment.
What exists now
Current platform, current hosting, current CRM, current everything. Include versions and include what is broken. Agencies pricing a project without knowing what they are inheriting will either pad the number or discover a problem halfway through, and you pay for it either way.
Scope, in and out
List what is definitely in and what is definitely out. The out list is more useful. “Content will be supplied by us” and “no migration of the old blog” are the sentences that prevent arguments in month three.
Integrations
Name every system this needs to connect to and say whether it has a modern interface. This single section moves the price more than any other, and getting it wrong is the most common cause of overruns.
Content
State plainly who is writing the copy, who is providing photography, and who is doing migration. If the answer is “we assume you will”, say so, and expect it in the price. If the answer is “we will”, be realistic about your team’s capacity, because this is the most common cause of delay.
Constraints
Brand guidelines, accessibility requirements, security or procurement standards, regulatory obligations, technology your IT team will not accept. Anything that limits the solution belongs here rather than emerging in week six.
Timescale
Give real dates and say what is driving them. “We would like it soon” produces a different proposal from “our exhibition is in March and the site must be live before it”.
Budget
Yes, include it. More on this below.
Decision process
Who decides, on what criteria, and by when. Agencies deprioritise briefs where this is vague, because vague decision processes usually mean no decision.
Writing a brief and want a second opinion? Send us a draft. We will tell you what is missing and what will cause problems later, whether or not you end up working with us.
Why you should name your budget
The fear is that naming a budget means an agency will spend all of it. Occasionally true. What certainly happens when you withhold it is worse.
Without a budget, three agencies will each guess at a different project size. One proposes something modest, one proposes something ambitious, one proposes what they think you can afford. You then compare three proposals for three different projects and conclude that agency pricing is arbitrary.
With a range, all three respond to the same question: what is the best thing you can do for this money? That is a comparison worth making, and it is the only way to see how differently agencies would prioritise the same budget.
A range is enough. You do not have to name a precise figure to be useful.
What not to put in
- A page-by-page specification, unless it is genuinely fixed. You are buying expertise. Prescribing the structure discards it.
- A list of features copied from a competitor. You do not know why they built those, or whether they work.
- Technology choices, unless you have a real constraint. If your team must maintain it in a particular language, say so and say why. Otherwise leave it open.
- A demand for fixed price on an undefined scope. You will get a padded number or a fight about change requests. Better to fix the scope, or accept a discovery phase first.
Discovery is not a delaying tactic
If your project has real uncertainty, particularly around integration or bespoke functionality, a paid discovery phase is the cheapest risk you will ever buy. A few weeks of proper investigation produces a scope, an architecture and a price everyone can stand behind.
The alternative is a fixed price built on assumptions, which becomes a series of change requests, which becomes an uncomfortable relationship. We would rather charge for the thinking up front and give you a number we can hold to.
Common questions
How long should a brief be?
Three to six pages covers most website and software projects. Longer than that and it is probably a specification, which is a different document with a different job.
Should I send the same brief to every agency?
Yes. Identical inputs are the only way to compare responses meaningfully. Answer follow-up questions individually, but keep the base document the same.
How many agencies should I approach?
Three is usually right. More than four and you will not give any of them proper attention, and good agencies can tell when they are one of eight.
What if I genuinely do not know what I need?
Then write the problem section, the success section and the budget range, and say openly that you want help defining the rest. That is a perfectly respectable brief and often the most honest one we receive. It is also how most of our lead generation engagements begin.
