How do we write a redesign brief so the proposals we get back are comparable?

When one proposal says $8,000 and the next says $34,000, the usual cause is that the brief did not say how many unique page designs were needed, who was writing the copy, or what happened to the old blog. Each studio filled those gaps with its own assumption. You are not comparing prices. You are comparing guesses.

In our experience, two studios handed the same six numbers land within roughly twenty percent of each other, and the remaining gap is genuinely about them: seniority, speed, how much of the build is bespoke. That gap is the one you want. It is a choice. The gap caused by missing information is not a choice, it is noise.

So the job of the brief is narrower than most people think. You are not describing the website. You are removing the guesses.

Each item below changes a quote by thousands. State them and studios price the same project. Leave them out and each studio quietly picks a number, and you never find out which one until the invoice or the change request arrives.

The pattern is worth noticing: a missing number pushes a studio one of two ways. Something visible and bounded, like template count, gets guessed high, because guessing low and being wrong costs them the project’s margin. Something invisible, like a content migration, gets guessed low or excluded silently, and you pay for it later as extra scope.

• **Unique templates, not pages.** List the page designs you think you need: home, product, pricing, blog index, blog post, contact. A template is a design, not a URL, and six templates can drive hundreds of pages through the CMS. Missing, a studio reads "40 pages" as 40 designs and prices accordingly.
• **Who writes the copy, and by what date.** Say whether it exists, who is writing it, and when it lands. Most studios do not write copy, and the ones that do charge for it. Missing, half the proposals assume you supply final text and the other half quietly assume they will be waiting six weeks, which they price as risk.
• **How many old pages move across, who moves them, and who does the redirects.** Give the number. "About 300 blog posts, our team pastes them in, you set up the redirect map." Missing, migration is the single most commonly excluded item we see, and it reappears as a mid-project cost.
• **Which page types marketers must build without a developer.** Name them. "Marketing needs to launch new landing pages and case studies alone. Everything else can come to you." Missing, one studio builds a flexible block editor and another hardcodes every page, and those are different amounts of work.
• **Reusable components with documentation, or one-off pages.** A documented set of components costs more up front and less every time you add a page. Say which you are buying. Missing, you get both priced as if they were the same thing.
• **Budget range and launch date, said out loud.** Both. A date without a budget produces a proposal that ignores the date, and a budget without a date produces one that ignores the budget.

The first is the page count that is not a page count. A brief arrives saying “roughly 14 pages”. We ask for the list, and it comes back as home, about, a services page repeated four times with different text, pricing, contact, and a blog. That is five or six designs, not fourteen. Anyone who quoted from the original sentence quoted for more than double the work, and the client concluded that redesigns are expensive.

The second is the blog nobody mentioned. The brief covers the marketing pages in detail and says nothing about the 200 existing posts, because to the person writing the brief those posts already exist and feel finished. They still need a template, a CMS structure, an import, and redirects for every URL that changes shape. When the brief is silent, most studios assume the client handles it, and the client assumed the studio did. That argument happens in week three.

Both are the same failure. The brief described the site as the client experiences it, not as the work is actually counted.

Refusing to name a budget gets you a proposal priced for the studio’s uncertainty, which is always higher than a proposal priced for your constraint. There is no version where the studio guesses low out of politeness.

Here is what a studio actually does with the number. It does not raise the price to meet it. It decides what fits inside it, then writes a proposal for that. Tell a studio $12,000 and it proposes six templates, a component set and a CMS your marketers can use. Tell it nothing and it proposes the safest thing it knows how to deliver, padded, because it cannot tell whether you will be shocked at $8,000 or at $80,000.

Phrase it as a ceiling with a purpose, not a target. “We have $15,000 to $20,000 approved and we would rather spend the lower end and keep budget for photography” is a constraint. “Our budget is around $20,000” is a target, and every proposal will arrive at $19,800.

If you genuinely do not know the number, say what you can. “This is a first website spend and we do not have a benchmark, so tell us what you would do at $5,000 and what changes at $15,000” is a fair question and a useful one. Published prices help you calibrate before you ask. Ours start at $1,000 for a single landing page and $5,000 for a full marketing site, and those are floors, not quotes.

Four sections show up in most briefs, cost you a day to write, and either get skimmed or actively damage what comes back.

Adjectives without an example. “Clean, modern, premium, human” describes every site ever launched. Three URLs you like and one sentence each on why beats a paragraph of feel. If you cannot name a site, name the thing you hate about your current one.

Prescribed page structures. A wireframe of the homepage in the brief tells four studios to price your layout instead of showing you a better one. You lose the only free thinking in the process. Say what the page must achieve, then let them argue.

Technology mandates with nobody to maintain them. Naming a framework is reasonable when you have engineers who will own the code. When you do not, it narrows your field for a decision you cannot evaluate. Say what matters instead: the client’s team must be able to edit content safely, the site must be deployed to accounts you own, and you must get the repository. Those are ownership terms, not technology terms.

Company history. Two sentences. Nobody pricing the work needs the founding story, and it is the section that makes a brief look like it was written to impress rather than to be answered.

Nine headings, one to three lines each. It fits on two pages and it is enough to price against.

Send the same document to every studio, on the same day, with the same deadline. Different versions of a brief produce different projects, which is the problem you started with.

1. **What we do, in two sentences.** Who buys from you and what the site is for.
2. **Why we are redesigning.** The problem, not the aspiration. "Sales cannot send prospects to our pricing page" is a brief. "Modernise our brand" is not.
3. **Templates we think we need.** The list. Say it is your best guess and you expect to be corrected.
4. **Content.** Who writes the copy, when it lands, who supplies photography, and whether either exists yet.
5. **Migration.** How many existing pages move, who moves them, who owns redirects.
6. **CMS.** Which page types marketers must be able to build alone, and who else needs access.
7. **Budget and dates.** A range, a launch date, and what the date is tied to.
8. **How we will decide.** Who signs off, how many studios you are asking, and when you will reply.
9. **What we need in your proposal.** Fixed price or range, milestones and payment points, what is excluded, and what you would cut if the budget dropped thirty percent.

Three proposals cost you real time: writing the brief, three calls, three read-throughs, three polite rejections. Below roughly $5,000 of work, that time costs more than the price spread you are trying to discover. For a single landing page, ask one studio you already rate and read its published price.

If you already trust someone who has done good work for you, a competitive brief is theatre with a cost. Ask them for a scope and a price. Compare it to a published floor if you want a sanity check.

If you cannot answer the six numbers, the honest brief is a paid discovery engagement, not a request for proposals. Pay one studio for two weeks to audit the current site, count the templates, decide which of the 400 pages survive, and write the scope. Then you can either continue with them or take that scope out to competition, and this time the replies will be comparable. A brief written to hide the fact that nobody has decided the scope produces four proposals that each decide it differently.

And say in the brief what would make you cancel. “If it turns out our traffic problem is a content problem, we will spend this budget on writers instead.” It is not a weak thing to write. It tells studios which problem you are actually buying a solution to, and a studio worth hiring may tell you a redesign is not the cheapest correct answer. That is a good outcome. You keep the money.

Do three things, in this order, before you look at the totals.

First, normalise on template count. Divide each price by the number of unique templates that studio has actually proposed, not the number of pages. Two of the three quotes usually converge once you do this, and the third is either pricing something extra or has misread the scope. Ask which.

Second, find the three items each studio excluded silently. Read for what is absent, not what is present. The usual absences are content migration and redirects, analytics and Search Console setup, and post-launch fixes. Write the three names next to each proposal and ask each studio to price them. Prices reorder themselves at this step more often than at any other.

Third, ask everyone the same question: “If the budget were thirty percent lower, what would you cut?” The answer tells you what they consider load-bearing. A studio that says it would cut a template has understood the site. A studio that says it would cut the design system has told you the system was decoration. A studio that says nothing can be cut has not thought about your budget as a real constraint.

Also compare how the money is released. Milestones tied to approvals, where each stage is invoiced only after you sign it off, are easier to stop than a fifty percent deposit against a single delivery date. We run four, foundation, content system, page deployment, launch ready, and the reason to prefer that shape has nothing to do with us: it means an unhappy project can end after milestone two rather than after the whole spend.

This is written for someone emailing three or four studios directly. Enterprise procurement and public sector RFPs work differently, and the six numbers will not save you there. Those processes have mandatory scoring matrices, security questionnaires, accessibility conformance statements, insurance thresholds and rules that forbid you from discussing scope with bidders once the tender is open. The main lever there is the scoring weights, which are set before the tender publishes, and if you can influence anything, influence those. Ask your procurement team what they are before you write a word.

Everything above also assumes you have authority to state a budget. If you do not, get it before you send the brief. A brief sent by someone who cannot confirm the number produces proposals that will all need rewriting.

You started with three documents pricing three different projects. The fix is not a longer brief, it is a brief that closes the six gaps a studio would otherwise fill on your behalf: templates, copy, migration, CMS, reuse, and money.

Do that and the remaining differences are the ones you actually wanted to see. A brief is not a description of the site you want. It is a list of the guesses you refuse to pay for.