By Mehdi HassanWhat is an agency evaluation scorecard?
An agency evaluation scorecard is a simple matrix: rows are the criteria you care about, columns are the agencies, and each cell gets a 1–5 score multiplied by a weight. The total gives you a ranking you can defend to a cofounder or a board, instead of a vague sense that one pitch felt better.
It works best as the last step of a structured process. You write the brief, send the same request to every agency, and then score what comes back. If you haven't sent a request yet, start with our website redesign RFP template so every proposal answers the same questions. A scorecard can't compare proposals that were written for different briefs.
Decide what matters before anyone pitches you. After the pitches, everything matters a little.
Which criteria belong on a web design agency scorecard?
Seven criteria cover almost every website project: relevant fit, scope clarity, copy and content ownership, the team doing the work, process and communication, post-launch support, and total cost. The weights below are a starting point we suggest for a startup marketing site. Change them to match your situation, but change them before you read any proposal.
Criterion | Suggested weight | What a 5 looks like |
|---|---|---|
Relevant fit | 15% | Similar stage, audience and complexity in their portfolio, and they can explain why it worked |
Scope clarity | 20% | Pages, templates, CMS, integrations and revisions listed, plus what's excluded |
Copy and content ownership | 15% | Named person writes the copy, or the proposal says plainly that you supply it |
Team doing the work | 10% | You know who designs, writes and builds, and you've met them |
Process and communication | 10% | Clear stages, review points, one channel, a named contact |
Post-launch support | 10% | Defined fix window, handover, training and an option for ongoing work |
Total cost and terms | 20% | Fixed price or a tight range, payment schedule, change-request rules |
Notice that price is one row, not the whole decision. Cost still carries a heavy weight, because budgets are real, but it only gets compared once the scopes match. A useful companion is our guide to what a good web design proposal looks like, which shows what each of these rows should read like on paper.
What do buyers usually forget to score?
The two rows buyers skip most are copy ownership and post-launch support, and both tend to surface as surprise costs a month in. A third, quieter one is who owns the files and accounts at the end.
Who writes the copy
Many design proposals assume you'll supply final text. If you don't have it, the project stalls on lorem ipsum, or you hire a copywriter halfway through and rework the layouts around real words. Research we reviewed on landing-page pricing found that two quotes can differ tenfold depending on whether strategy and writing are included. Score this row on what the proposal actually says, not what you assume. Our list of design agencies that also write your copy explains what to look for.
What happens after launch
Ask how long bugs are fixed for free, who you call when the form stops sending, and what it costs to add a page in month three. A proposal that ends at go-live scores a 1 or 2 here, however good the design is.
Who owns what at the end
You should end up owning the domain, hosting account, CMS, design files and code repository in your own name. If the proposal is silent on this, score it low and ask.
How do you score agency proposals objectively?
You score proposals objectively by fixing the rules first, scoring independently, and only then talking. The process takes an afternoon for three to five proposals.
Lock the criteria and weights
Agree on the rows and percentages with whoever signs off, before you open a proposal. Write them down.
Define what a 1, 3 and 5 mean
For each row, write one sentence for a low, middle and high score. This stops a 4 from meaning different things to different people.
Score independently
Two people score every proposal alone, using only what's written in the proposal and said on the call.
Reconcile the gaps
Where your scores differ by two or more points, discuss the evidence and settle on one number. Gaps usually reveal an unclear proposal.
Multiply and rank
Multiply each score by its weight and add the row totals. The top one or two agencies become finalists.
Test the finalist
Before signing the full project, run a short call on a real problem or a small paid piece of work to check the score holds up.
How do you compare agency proposals with different prices?
Compare agency proposals with different prices by normalizing scope first: list what each price includes, line by line, and fill the gaps with what you'd pay elsewhere. Only then is the cost row fair.
Build a short checklist: number of designed pages, CMS and its setup, copywriting, animation, responsive QA, SEO basics, analytics, redirects if it's a redesign, and post-launch support. A cheap quote that leaves out copy and a CMS often lands near a more expensive one once you add those back. For a sense of market ranges, the research we use puts Webflow agency projects at about $5,000 to $80,000+, Framer builds at about $3,000 to $28,000 and Next.js builds at about $20,000 to $120,000. These are typical ranges we see quoted, and they vary a lot.
Hypothetical example: scoring three proposals
For example, suppose a seed-stage startup compares three proposals for a six-page site. Agency A is cheapest but excludes copy and offers no post-launch window. Agency B includes copy, a CMS and 30 days of fixes at a mid-range price. Agency C is the most expensive, with a strong portfolio but a vague scope.
Scored on the weights above, A does well on cost but scores 1 on copy ownership and 1 on post-launch support. B scores 4 or 5 on most rows and 3 on cost. C scores 5 on fit and 2 on scope clarity. B comes out on top, even though it isn't the cheapest. This is an illustration of the method, not a real comparison.
What red flags should lower an agency's score?
Lower the score when a proposal is vague about scope, silent about copy, or priced only by the hour with no estimate. These are signs the real cost will be set later, after you've committed.
- No exclusions listed. Every honest scope says what isn't included.
- Unlimited revisions with no process. It sounds generous and usually means the project has no clear end.
- A portfolio you can't discuss. If they can't explain why a past site worked, the next one is luck.
- The team changes after signing. Ask who does the work and whether you'll meet them.
- No payment schedule or change-request rule. You'll negotiate both mid-project, at your weakest point.
Once you have a finalist, turn their proposal into a contract with clear deliverables. Our website statement of work template shows the scope language to ask for.
Should the highest score always win?
No. The highest score should earn the shortlist, and a small test should earn the contract. A scorecard measures what's written down, and some of what matters, like how an agency handles feedback, only shows up once you work together.
If two agencies finish within a few points of each other, pick the one whose team you'd rather spend three months with, or ask both for a short paid piece of work. If you're early and still building a shortlist, our roundup of website agencies for early-stage startups and our explainer on what a full-service agency really delivers can help you decide who to invite.
