Portfolios are the least reliable signal
Most agency selection starts and ends with portfolios. That is understandable, because visual work is easy to judge quickly. It is also misleading. A portfolio shows the final look of past projects, not whether they met their goals, how difficult the agency was to work with, or whether the site still performs today.
What predicts a good outcome is the agency's thinking: how it diagnoses problems, how it handles content and structure, how it measures success and how it behaves when things go wrong. Those qualities can be assessed, but only if you design the selection process to reveal them.
Hire the agency that asks you the hardest questions, not the one that agrees with you fastest.
Step 1: Write a brief that makes comparison possible
A vague brief produces proposals that cannot be compared. Before contacting anyone, write down your business goals and how success will be measured, your audiences and their key tasks, the scope as you understand it (page types, integrations, content responsibilities), constraints such as platform, timing and compliance, and what you expect after launch.
Be honest about budget, at least as a range you can share privately. Agencies design very different solutions for different budgets; withholding it often produces proposals that are either unaffordable or underpowered. What drives website cost explains the factors that shape a quote.
Step 2: Shortlist on relevance
Look for agencies with experience relevant to your situation: your type of business (B2B, ecommerce, public sector), your platform, your scale and the kind of problem you have, such as a redesign, a re-platform or conversion improvement. Relevant experience shortens learning time and reduces risk.
Consider also what kind of partner you need. Some agencies are strongest at brand and visual design; others at technical build; others at performance and conversion. A specialist may be stronger in its area; a broader agency may coordinate more easily. The specialist versus full-service agency guide sets out the trade-offs.
Step 3: Run a selection process that reveals thinking
Fig. 01 · Process
Tap to explore
A selection process in six steps
Be cautious about asking for speculative design work. Free creative produced without any research reveals what an agency can make in a hurry, not how it would solve your problem, and it rewards agencies willing to give away work. A short paid discovery phase or a working session on your actual challenge is usually more revealing.
Step 4: Evaluate against clear criteria
Fig. 02 · Scorecard
Tap to explore
What to weigh when comparing agencies
Bars show relative emphasis, not measured data
Score each agency independently with two or three colleagues before discussing. This reduces the influence of the most persuasive presentation and of whoever speaks first in the debrief.
Step 5: Ask the questions that matter
Checklist
0/10Questions to ask every shortlisted agency
Pay attention to the questions agencies ask you during these conversations. Questions about your customers, sales process, analytics and constraints show an agency already thinking about outcomes. Questions only about pages, features and visual preferences suggest an agency that will deliver what you ask for, which is not always what you need.
Step 6: Compare proposals on substance
Lay proposals side by side and compare scope, not totals. Check what each includes and excludes: discovery, content, number of templates, integrations, testing, accessibility, training, post-launch support. Note the assumptions, especially about content and approvals. A lower price that excludes content and assumes one round of feedback is not cheaper; it moves cost and risk to you.
| Compare | Why it matters | Watch for |
|---|---|---|
| Scope by stage | Shows what you are actually buying | Vague phases such as 'design and build' |
| Assumptions | Reveals hidden cost and risk | Content, revisions and approvals assumed on your side |
| Team and time | Determines quality and pace | Senior people at pitch, juniors on delivery |
| Measurement | Links the project to outcomes | No mention of analytics or success metrics |
| Ownership and exit | Protects you later | Agency-owned hosting, code or accounts |
Reading references properly
Reference calls are often wasted on polite questions that produce polite answers. Ask instead about the moments that test a relationship: what went wrong, how the agency responded, whether the budget and timeline held, and how the site has performed since launch. Ask whether they would hire the same team again for a similar project, and listen for hesitation.
Choose references yourself where you can. Agencies will naturally offer their happiest clients; ask for a project similar to yours in scope and sector, and if possible one completed more than a year ago, so you learn how the work has aged and whether the client still uses the agency.
- Did the agency challenge the brief, and was that useful?
- How accurate were the estimates of time and cost?
- Who did the work day to day, and were they the people introduced at the pitch?
- How was the handover: documentation, training, access to accounts?
Red flags
Myth vs reality
Assumptions that lead to poor choices
Other warning signs include reluctance to provide references, unclear answers about who will do the work, no questions about your customers, and contracts that leave the agency owning your domain, hosting or analytics accounts.
Set the relationship up to succeed
Once chosen, agree success measures, decision rights, communication rhythm and how changes will be handled. Name one decision-maker on your side and give the agency access to the people and data it needs, including customers and sales teams. Agencies do their best work when clients treat them as partners in a shared goal. The in-house versus agency guide covers how to split responsibilities over the long term, and the website design process shows what good delivery looks like.
Key takeaways
- 01Portfolios show past looks, not outcomes; evaluate how an agency thinks about your problem.
- 02A clear brief with goals, scope, constraints and a budget range makes proposals comparable.
- 03Meet the people who will actually do the work and speak to recent clients.
- 04Compare proposals on scope and assumptions, not totals, and make sure you own all assets.
- 05Treat guaranteed rankings or uplifts and free speculative design as warning signs.
Frequently asked
- How do I choose the right web design agency?
- Write a clear brief, shortlist agencies with relevant experience, and run a process that reveals how they think: discovery conversations, proposals with explicit scope and assumptions, and meetings with the actual delivery team. Score agencies against agreed criteria, speak to recent clients, and confirm ownership of all code and accounts.
- What questions should I ask a web design agency?
- Ask how they would define and measure success, who will work on the project, what their discovery involves, how they handle content, SEO migration, accessibility and performance, how scope changes are managed, what happens after launch, who owns the outputs, and whether you can speak to recent clients with similar projects.
- Should I ask agencies for free design concepts?
- Generally not. Speculative designs are produced without research into your customers, so they show style rather than problem-solving, and they disadvantage agencies that do not give work away. A paid discovery phase or a working session on your real challenge is a better way to judge fit.
- What are red flags when hiring a web design agency?
- Guarantees of rankings or specific conversion uplifts, vague scopes, reluctance to provide references, senior people who disappear after the pitch, no questions about your customers or goals, and contracts in which the agency owns your domain, hosting, code or analytics accounts are all reasons for caution.
- Should I choose a local or remote web design agency?
- Location matters less than relevance, process and communication. Remote collaboration works well for most web projects. Local presence can help for workshops or where cultural and market understanding is important, for example when serving specific regional audiences. Judge on capability and fit first, then consider practicalities such as time zones.
Published by Fabulous.Media, a network of specialist marketing agencies. Updated 9 October 2026. Platform features change often; check current official documentation before acting on platform-specific detail.





