Website Redesign RFP: What to Include and a Template You Can Copy

A website redesign RFP is a structured procurement document sent to shortlisted digital agencies that defines your business objectives, page template scope, non-functional technical constraints, budget corridor, and weighted evaluation criteria so every vendor prices the exact same deliverables and submits directly comparable proposals.

That last part is where most procurement rounds fail. Five agencies read the same twelve pages, form five different pictures of the engineering and content effort, and send back numbers between €18,000 and €140,000. Those figures cannot be compared because they are not priced against the same scope.

The fix is not a longer document. It is one that pins down the four variables that actually move a quote: how many unique page templates exist, who writes and migrates the content, which business systems must integrate via API, and what objective verification counts as done. Pairing your RFP with an executable website redesign checklist and transparent website pricing benchmarks keeps every vendor aligned from day one.

An RFP costs real time on both sides. Two to four weeks to write it, three weeks for agency responses, and one to two weeks to score them puts eight weeks between the decision to redesign and the first line of production work. Below roughly €25,000 of project value that procurement overhead exceeds the savings it produces, and a direct brief to two or three vetted agencies gets you further faster.

In the 2027 European market, standard corporate websites with ten to twenty pages typically fall into the €4,500 to €9,000 investment band, while custom web portals, multi-locale enterprise builds, and deep CRM or ERP integrations run from €12,000 to €25,000+ and upwards into six figures. Matching your procurement process to the actual budget band prevents wasted effort.

Write one when at least two of these are true:

• The budget is above €25,000, where a wrong vendor choice is expensive to unwind.
• Procurement or a board requires a documented competitive selection.
• More than one department has to sign off, and you need their functional requirements written down before anyone starts designing.
• The site connects to systems that agencies must integrate cleanly: a CRM, an ERP, a booking engine, or an authenticated customer portal.

Skip it and brief two or three agencies directly when:

• You have a straightforward corporate site under twenty pages in the €4,500 to €9,000 band.
• You already know which agency you want and the RFP would be theatre.
• You need to start in under a month, because formal RFP cycles do not compress well.

In the second case, a shorter website design brief does the same alignment work with a quarter of the effort. The brief describes the project; the RFP describes the project and the procurement rules around it.

It also helps to study real corporate websites and why their layouts work before you lock the scope.

Agencies read an RFP in a predictable order, and it is rarely the order most RFPs are written in. Scope, budget, and deadline get read first, because together they decide whether the agency has the right team and commercial fit to respond at all. Content ownership gets read fourth and quietly determines whether the launch schedule can hold. Company history gets read last, if it gets read at all.

Write for that triage order. Put the decision-relevant material in the first four pages and the corporate background after it. When technical leads review those opening pages, they immediately check whether the CMS architecture is dictated or open to recommendation, whether third-party APIs have documented endpoints, and whether mandatory launch features are separated from phase-two wish lists.

Copy the eleven headings below into a document and fill them in. A website redesign RFP template that works tends to run eight to fifteen pages. Past twenty, agencies skim, and skimming produces the vague quotes you were trying to avoid.

1. Project summary. Three sentences. What you are redesigning, why now, and when it needs to be live. If a reader cannot repeat this back after one pass, rewrite it.

2. About the organisation. What you sell, to whom, in which markets and languages. Two paragraphs.

3. Current site and what is wrong with it. URL, platform, approximate page count, monthly traffic, and the specific problems. "Dated design" is not a problem statement. "Mobile conversion is 0.4% against 1.9% on desktop" is.

4. Goals and how you will measure them. Two or three business metrics, each with a current baseline value, a target, and a date. If you cannot state the current value, say so; that itself tells the agency you will need analytics configuration.

5. Audience. Who visits, what they came for, which devices, and which languages. Include the internal editorial audience if staff use the CMS daily to publish pages.

6. Scope. The section that decides the price. List unique page templates rather than individual pages: home, service detail, case study, blog index, blog post, contact, careers, legal. Ten templates and two hundred pages cost far less than thirty templates and forty pages. Then list functional modules, split strictly into mandatory for launch and desirable for phase two.

7. Content. State plainly who writes the new copy, who supplies photography, and whether existing content is migrating as is. If the answer is "the agency", say so and expect the quote to rise by roughly a third.

8. Technical and non-functional requirements. Hosting architecture, systems to integrate, single sign-on, accessibility level, performance targets, data residency, and SEO migration ownership. Name the specific systems rather than general categories, and write verifiable engineering thresholds rather than aspirations. Specify WCAG 2.2 Level AA compliance as required under the European Accessibility Act, including full keyboard operability and 4.5:1 text contrast. For speed, require passing Google's Core Web Vitals on 4G mobile connections: Largest Contentful Paint (LCP) under 2.5 seconds, Interaction to Next Paint (INP) under 200 milliseconds, and Cumulative Layout Shift (CLS) below 0.1. Include strict GDPR data residency on European Union servers and 1:1 301 redirect mapping.

9. Budget and timeline. A realistic disclosure band and a target launch date, with any fixed external dates that cannot move.

10. Selection process. Response deadline, format, page limit, presentation round, decision date, and the evaluation criteria with their percentage weights. Give one named contact for questions and one deadline for asking them.

11. Terms and 3-year TCO. Full intellectual property ownership of design files and source code, milestone payment schedule, post-launch warranty period, and 3-year Total Cost of Ownership (TCO) separating one-time build fees from annual hosting, licensing, and maintenance retainers. Attach your contract template if you have one, because a surprise legal review after selection can add a month.

The most common instruction we see is a blank where the budget should be, on the theory that naming a number invites agencies to spend all of it. The reasoning is understandable and the result is the opposite of what you wanted. Every respondent guesses instead, the guesses scatter, and you end up comparing quotes that were never priced against the same assumptions.

In practice that means one response at €22,000 and one at €95,000 for the same brief. You learn nothing from the spread and spend two weeks in follow-up calls establishing the range you could have written on page one.

A disclosed budget corridor solves it without giving anything away. Whether your project sits in the €4,500 to €9,000 band for a focused corporate website, €12,000 to €25,000+ for a custom multi-locale B2B portal, or €40,000 to €60,000 for an enterprise integration build, stating the band lets agencies propose an appropriate technical architecture. When a vendor comes back at €110,000 against a €40,000 to €60,000 corridor, you immediately know they either spotted unstated integration complexity or misread the brief.

If you genuinely do not know what the work costs, published market figures are a reliable starting point. Our breakdown of website redesign costs details European agency bands by project complexity.

Decide the evaluation weights before the proposals arrive, ideally before the RFP goes out, and publish them inside the document itself. Scoring after you have read the responses is how the flashiest pitch deck wins rather than the strongest engineering fit, because by then you are rationalising a preference you already formed. Publishing the weights also raises the quality of what you receive: agencies write directly to the criteria they can see.

A balanced B2B procurement scorecard allocates 30% to technical architecture, CMS fit, and Core Web Vitals commitments; 25% to relevant B2B portfolio cases with verified conversion rate optimization (CRO) uplift; 20% to transparent phase pricing and 3-year TCO clarity; 15% to 301 SEO migration, WCAG 2.2 AA accessibility, and GDPR compliance; and 10% to post-launch SLA coverage and named team seniority.

One qualitative signal is worth more than the whole slide deck: whether the proposal pushes back on anything. An agency that accepts every requirement without comment has either not read the RFP closely or has decided to charge for change requests after signing. The proposals that question an unrealistic deadline, flag an undocumented API, or point out that your content schedule is optimistic are written by practitioners who have delivered complex relaunches before.

Ask for two client references from the last twelve months and call them. Ask one specific question: what went wrong during the build, and how did the agency handle it? Every real project encounters friction; a reference who cannot name a single obstacle did not work with them on anything demanding.

The two words get used interchangeably in conversation, and they sit at opposite ends of the same exchange. Knowing which one you are writing settles most arguments about what belongs in it, because the RFP is the question and the proposal is the answer. One is identical for every recipient; the other should be different every time.

You write the RFP. It states the problem, the constraints and the rules of the selection. It is the same document for every recipient, which is what makes the responses comparable.

The agency writes the proposal. It states the solution, the process, the team and the price. A website redesign proposal that arrives without a scope table or an assumptions list is not finished, and asking for both is reasonable.

A middle form is worth knowing about. If your requirements are still forming, an RFI, a request for information, asks a short set of questions of a wider group without demanding a full costed response. It costs agencies an hour instead of two days, so more of them reply, and you can write a sharper RFP afterwards. This works well when you are unsure whether the project needs a platform migration at all, a question our website redesign guide covers in more depth.

Six patterns account for most of the disappointing response rounds we see, and all six are decided by a sentence or two in the RFP rather than by anything the agencies do. They are cheap to fix while the document is still a draft and expensive to fix afterwards, because the only remedy once responses are in is to reissue and lose three weeks.

Sending it to ten agencies. Response quality drops as the recipient list grows, because good agencies triage. Four to six is the range where you get considered answers. Above that, you get templates.

Listing features without priority. When everything is required, nothing is, and the quote comes back at the top of every assumption. Split the list into required and desirable and be honest about which is which.

Leaving the content owner unnamed. Content is the most frequent reason redesigns slip, and it is decided in one sentence of the RFP. Name the owner and the deadline for the first draft.

Asking for free concept work. Speculative design in a proposal round tells you which agency had a junior available that week, not which one will do the strongest work. It also quietly excludes the ones with a full pipeline, which are usually the ones you want.

No question window. Without a stated deadline for questions and a shared answer document, you spend two weeks answering the same thing five times and the responses still differ.

Skipping the redirect plan. Almost no RFP mentions it, and it is the costliest omission after launch. According to Google Search Central, a site move with URL changes requires every old address to be mapped to its new one before the switch, and ranking recovery still takes weeks afterwards. Decide in the RFP whether that mapping is your job or theirs, or read what happens when nobody owns it in our guide to SEO during a redesign.

A good website redesign RFP is short, specific about scope, honest about budget, and clear about who writes the content. Those four properties produce quotes you can lay side by side. Everything else in the document is context that helps an agency write a warmer proposal but changes the number very little.

If you are about to send one, do a last pass with one question in mind: could a stranger price this without calling you? If the answer is no, the missing information is what your next paragraph should contain. Then send it to four agencies rather than ten, give them three weeks, and score against weights you wrote down before the first response arrived.