Start with pass-fail gates
Keep nonnegotiable requirements out of the weighted score. If an agency cannot meet one of them, a strong creative presentation should not compensate for the gap.
Common pass-fail gates include:
- Budget range or approved commercial model
- Required launch window
- Security, privacy, or procurement requirements
- Critical CRM, analytics, or product integrations
- Accessibility obligations
- Named senior-team involvement
- Contract, insurance, or data-processing terms
- Support coverage during and after launch
Write each gate so an agency can answer it with a clear commitment, exception, or assumption. “Strong accessibility experience” is too loose. “Agency will design and develop against the stated WCAG target and include testing responsibilities in scope” gives the committee something it can verify.
Do not score a bidder that fails a true gate. If your team keeps an agency in the process despite a failed item, the requirement was negotiable and belongs in the weighted matrix.
Use a 100-point website redesign RFP scoring matrix
This matrix suits a B2B company hiring an agency for strategy, design, development, CMS work, integrations, analytics, migration, and launch. It puts more weight on the quality of the proposed system than on the sales presentation.
| Criterion | Weight | What the committee should assess |
|---|---|---|
| Business and website strategy | 20 | Understanding of buyers, positioning, conversion paths, content priorities, and business goals |
| Proposed solution and technical fit | 20 | Architecture, CMS model, integrations, performance, accessibility, analytics, security, and fit with internal constraints |
| Delivery plan and risk control | 15 | Discovery, decision points, dependencies, migration, QA, launch planning, ownership, and risk handling |
| Team and relevant experience | 15 | Named team, senior involvement, role clarity, comparable complexity, and evidence of sound judgment |
| Maintainability and client ownership | 15 | Editorial workflow, component system, documentation, training, code ownership, and support model |
| Commercial fit | 15 | Price, assumptions, payment structure, change control, exclusions, and total cost implications |
| Total | 100 |
The weights make a point: your company will live with the website system long after the pitch meeting. The agency’s discovery method, architecture, component model, and handoff deserve more scrutiny than presentation style.
1. Business and website strategy: 20 points
Look for a diagnosis of the current website and a clear connection between the redesign and the buying process. The agency should address positioning, information architecture, conversion paths, proof, content needs, and sales handoff where those issues fit the brief.
A high score needs proposal evidence. The bidder may identify a conflict between the requested sitemap and the way buyers evaluate the offer. It may propose research to resolve an unclear audience hierarchy. It may explain how the discovery phase will turn commercial goals into page priorities and measurement requirements.
Give a lower score when the proposal repeats your RFP language, treats design style as strategy, or promises conversion gains without explaining the work behind them.
2. Proposed solution and technical fit: 20 points
Ask whether the proposed platform and architecture fit your operating model. Your marketing team, developers, sales systems, content volume, localization needs, analytics stack, and security constraints should shape the recommendation.
Score the proposal on points such as:
- CMS roles, workflows, preview, and structured content
- Component flexibility and design-system control
- CRM, marketing automation, product, and data integrations
- Analytics events, consent handling, and tag governance
- Performance budgets and technical SEO
- Accessibility requirements and test coverage
- Hosting, deployment, monitoring, and rollback
- Source-code access and ownership
Watch for a platform recommendation that appears before the agency understands the requirements. Familiarity helps delivery, but tool preference cannot replace fit.
3. Delivery plan and risk control: 15 points
A useful plan shows how the agency will move from open questions to approved work. It names phases, decisions, client responsibilities, review windows, and dependencies.
Look for treatment of the work that causes redesign schedules to slip: stakeholder approvals, content production, migration volume, integration access, legal review, analytics specifications, browser testing, redirect mapping, and launch ownership.
A list of weeks and deliverables earns a middle score at best. Give higher scores when the agency explains how it will control risk, surface decisions, and adjust the plan when your team misses a dependency.
4. Team and relevant experience: 15 points
Score the people who will do the work. Ask for names, roles, time commitments, and the point at which senior staff leave the project. A strong agency brand does not tell you who will lead discovery, make architecture decisions, or review the build.
Relevant experience means comparable problems, not a logo collage. A project with similar CMS governance, migration complexity, stakeholder structure, or integration risk may matter more than work in the same industry.
Use references and interviews to test the proposal claims. Ask how the team handled a difficult scope decision, a delayed client dependency, or a technical recommendation the client challenged.
5. Maintainability and client ownership: 15 points
Your team should understand what it can change after launch, what requires a developer, and who owns the underlying assets.
Score these areas:
- Reusable component and page-building model
- Content-editor guardrails
- Documentation and training
- Design files, source code, and account ownership
- Dependency and package maintenance
- Warranty, support, and enhancement process
- Handoff to an internal or third-party development team
A maintainable site balances editorial flexibility with design and technical control. Unlimited page-builder freedom can create inconsistency. A rigid system can turn routine campaign work into a development ticket. The proposal should explain where the agency will set that boundary.
6. Commercial fit: 15 points
Compare the price with the work and assumptions behind it. A low total can hide missing discovery, thin QA, client-owned migration work, limited revision cycles, or an aggressive estimate for unresolved integrations.
Review:
- Fixed fees, ranges, or time-and-materials portions
- Assumptions and exclusions
- Allowances for content, migration, and integrations
- Revision and approval rules
- Change-control process
- Payment schedule
- Ongoing hosting, software, support, and maintenance costs
Give the highest score to the clearest commercial model that fits your risk tolerance. Do not reward a false sense of certainty when the RFP leaves major questions open.
Define the scoring scale before proposals arrive
Use one scoring scale across every criterion:
| Score | Meaning | Evidence standard |
|---|---|---|
| 0 | Missing or unacceptable | Proposal does not address the criterion or conflicts with a requirement |
| 1 | Weak | Generic response, major gaps, or unsupported claims |
| 2 | Partial | Covers part of the requirement but leaves material questions |
| 3 | Credible | Meets the requirement with a workable approach and clear assumptions |
| 4 | Strong | Shows specific insight, useful evidence, and sound risk control |
| 5 | Exceptional | Fits the requirement with specific evidence and anticipates material issues your brief missed |
Convert each criterion score into weighted points with this formula:
Weighted points = (criterion score ÷ 5) × criterion weight
A score of 4 on a 20-point strategy criterion produces 16 weighted points. Keep one decimal place if your spreadsheet needs it. More precision will not make a subjective judgment more accurate.
Require evaluators to cite a page, section, interview answer, or reference call for scores of 1, 2, 4, or 5. A score of 3 means the agency met the stated requirement. Scores above or below that baseline need a reason.
Score proposals before the committee meeting
Send the same proposal package, matrix, and instructions to each evaluator. Ask everyone to score alone before the group discussion.
Independent scoring protects the process from predictable committee problems. The first speaker can anchor the room. A senior executive can turn a preference into the apparent consensus. A strong presenter can receive credit for work the written proposal never committed to.
Collect the scores before the meeting. Then flag criteria where evaluators differ by two or more points on the five-point scale. Use the committee session to discuss those gaps rather than reading every score aloud.
The discussion should answer four questions:
- Which proposal evidence supports each score?
- Did evaluators interpret the criterion in the same way?
- Does the difference expose an unanswered risk or stakeholder priority?
- Do all finalists need the same follow-up question?
Record the reason for any score the committee changes. Procurement gets a cleaner decision trail, and the project team keeps the assumptions that influenced selection.
Separate proposal scoring from interviews
The written proposal and finalist interview test different things. Score them in separate columns or run the interview as a second round with its own criteria.
A finalist interview can assess:
- Quality of the named working team
- Ability to explain tradeoffs without sales language
- Response to a realistic scope or stakeholder scenario
- Clarity about assumptions and unresolved questions
- Working style with your internal project owner
Give every finalist the same core questions. Add bidder-specific questions for proposal gaps, but do not let one agency receive an easier interview.
Avoid requesting unpaid strategy concepts or speculative design work. Those exercises reward pitch production and encourage agencies to make decisions without discovery. Use a paid workshop if you need to test collaboration on a real problem.
Test the matrix before you issue the RFP
Run the draft matrix against two imaginary proposals. One bidder offers a lower price and thin discovery. The other costs more, assigns a senior team, and gives marketing stronger CMS ownership.
If your weights produce the same result regardless of those differences, the matrix does not reflect your priorities. Adjust it now.
Ask each stakeholder to name the criterion they would protect if the budget tightened. Marketing may prioritize content operations. IT may protect security and integration fit. Sales may care most about buyer journeys and CRM handoff. Resolve those tensions before agencies invest time in proposals.
Share the criteria and weights in the RFP. Agencies can then spend their response time on the evidence your team needs. Hidden criteria do not improve selection; they produce generic proposals and surprise bidders during interviews.
Use the score as decision support
The final total should narrow the field and expose tradeoffs. It should not replace judgment.
Set a minimum score for each high-risk criterion. An agency with the highest total may still be the wrong choice if it fails technical fit or maintainability. Record that rule before scoring so the committee does not invent it to favor a bidder.
Review close totals by looking at the source of the points. A one-point lead driven by price carries a different risk than a lead driven by solution fit and delivery control. The committee should decide which risk it wants to own.
The matrix earns its keep when it improves the questions, documents assumptions, and makes the selection logic visible. The contract and discovery plan should then carry those assumptions into the project.
Frequently asked questions
How should we weight price in a website redesign RFP?
Use price as one weighted criterion after each agency passes your budget and commercial gates. A 10 to 20 percent weight often keeps cost visible without letting the cheapest proposal override strategy, delivery, and long-term ownership.
Should every evaluator use the same RFP scoring matrix?
Yes. Give each evaluator the same criteria, weights, scoring scale, and evidence rules. Ask them to score proposals alone before the committee discussion so seniority and group dynamics do not set the first score.
Can we change the scoring criteria after proposals arrive?
Avoid changing weights after you read the proposals. A late change can favor one bidder and weaken the integrity of the process. If a proposal exposes a missing requirement, document the issue and ask every finalist the same follow-up question.
What should disqualify a website redesign proposal before scoring?
Use pass-fail gates for requirements your team cannot negotiate, such as budget range, launch window, security review, required integrations, contract terms, accessibility obligations, or named-team availability.
Turn the matrix into a stronger agency brief
A scoring matrix works best when the RFP gives agencies enough context to respond with evidence. Define the commercial problem, project scope, required systems, internal responsibilities, decision process, and known constraints. State what remains open for paid discovery.
Virdis helps B2B teams plan and deliver custom website redesigns across strategy, design, development, CMS architecture, integrations, analytics, migration, and launch. If your team needs to turn an RFP into a buildable brief, start with the risks and decisions the scoring matrix exposed.
