What if the most expensive design mistake isn’t a bad color palette or a clunky menu, but the questions you forgot to ask before signing? Hiring a UI/UX design company is a lot like hiring a builder for a house. The portfolio looks great, the pitch is smooth, and then six months later you discover the foundation was never poured. The difference between a partnership that lifts your product and one that drains your budget usually comes down to what you asked in the first meeting.
Quick Answer: What questions should I ask an UI UX design company before hiring?
Ask about accessibility standards, testing methods, how success gets measured, and who actually does the work. A good agency should explain how it designs and tests against recognized standards, how it runs usability research, and which metrics it uses to prove results. These answers reveal whether you are buying a repeatable process or just a pretty mockup.
Some questions are about capability, others about accountability. A studio may excel at visual polish but skip research entirely, while another may be strong on research but slow at delivery. Knowing which type you are talking to changes how you scope the contract.
Before you sign, compare accessibility conformance, research methods, success metrics, design-system handoff, data-handling practices, IP ownership, and performance testing on low-end devices.
Why These Questions Matter More Than the Portfolio

Portfolios are curated. They show the final screens, the happy path, the version that survived three rounds of internal review. What a portfolio rarely shows is the messy middle: the research that got scrapped, the feature that tested badly, the developer who could not build what was designed.
That gap is where budgets go to die. A design that looks stunning in a Figma file can fall apart the moment it meets real users, slow networks, or a screen reader. The questions in this guide are designed to surface that gap before money changes hands. They focus on process, evidence, and accountability rather than aesthetics.
There is also a practical angle. Design work often sits upstream of legal and technical obligations. Accessibility is required in many markets, and privacy rules shape how user research can be conducted at all. Asking the right questions early keeps those obligations from becoming expensive surprises later.
Accessibility: The Question Most Buyers Forget
Start here, because it is the question buyers skip most often and regret most.
Ask: Do you design and test against WCAG 2.2, and at which conformance level? WCAG 2.2 became a W3C Recommendation on 5 October 2023. It defines success criteria across levels A, AA, and AAA. A credible agency should name the level it targets and explain how it verifies conformance, rather than only claiming it.
Then widen the lens depending on where you operate. In the EU, the European Accessibility Act (Directive (EU) 2019/2102) had a transposition deadline of 28 June 2022 and applies from 28 June 2025, though its scope covers specific products and services rather than every website. In the US public sector, Section 508 sets standards for federal agencies’ information and communication technology, and the Revised 508 Standards incorporate WCAG 2.0 AA.
A useful follow-up: Can you show a recent accessibility audit, including what failed? Agencies that test honestly will have found problems. The ones that claim a perfect score on the first pass are usually not testing at all.
Research Methods and Sample Sizes
Ask: How do you decide how many users to test with, and do you include people who use assistive technology?
The famous “five users” heuristic comes from Nielsen and Landauer’s 1993 work. It is widely cited, but it is contested when used to support quantitative claims. Treat it as a rule of thumb for catching major usability problems, not as a statistical guarantee. A good agency will explain its reasoning rather than recite the number.
Dig into the method itself. Are sessions moderated or unmoderated? Remote or in person? Do participants match your actual audience, including older users, users with disabilities, and people on low-end devices? If the answer is “we test with whoever is available,” that is a warning sign.
This is also where privacy enters the conversation. If the agency will handle participant data in the EU or UK, GDPR (Regulation (EU) 2016/679) applies from 25 May 2018. Ask how consent is collected, where recordings are stored, and how long they are kept. For teams evaluating tooling alongside services, the same discipline applies to platforms, and a structured set of AI chatbot security questions to ask before buying shows how far that diligence should extend.
How Will Success Be Measured?
Ask: Which metrics will you baseline, and how will you report them?
Vague answers like “improved user satisfaction” are not measurable. Push for named instruments. The System Usability Scale (SUS), a 10-item questionnaire published by John Brooke in 1996, is one of the most established tools for measuring perceived usability. Ask whether the agency uses validated instruments like SUS, and how it establishes a baseline before redesigning anything.
Pair that with behavioral metrics. Task completion rate, time on task, error rate, and support ticket volume all tell a story. If the work touches the live site, performance metrics matter too. Core Web Vitals are page-experience signals used by Google, introduced in 2020, with Interaction to Next Paint replacing First Input Delay in March 2024. An agency that designs without knowing these thresholds is designing blind.
A practical tip from teams that have been through this: agree on the measurement plan in writing before design starts. Retrofitting metrics after launch is how “we improved the experience” becomes an argument instead of a result.
Process, Handoff, and Design Systems
Ask: What do you hand off, in what format, and to whom?
Design systems are where good intentions meet engineering reality. Ask whether the agency delivers to a documented design system and how it handles developer handoff and design tokens. The W3C Design Tokens Community Group publishes a format specification, which is a useful maturity signal, though its status has been evolving, so confirm the current state before citing it in a contract.
Then ask the uncomfortable question: Who actually does the work? It is common for senior designers to run the pitch and junior staff to execute the project. That is not automatically bad, but you should know. Ask for the named team, their roles, and whether any work is subcontracted.
This is where a portfolio review becomes an interview. Ask for case studies that show problem framing, research inputs, iteration, and measured outcomes. A case study that starts with a beautiful screen and ends with a beautiful screen tells you nothing about process.
Contract Details Buyers Overlook
Ask: Who owns the research artifacts, source files, and design system if we part ways?
This is a common contract gap. Many agreements cover final deliverables but stay silent on research recordings, component libraries, and working files. Clarify ownership on termination as well as on completion.
A few more contract questions worth asking out loud:
- What happens if key team members leave mid-project?
- Are revisions capped, and what counts as a revision?
- Who is responsible for accessibility fixes after handoff?
- What are the payment milestones tied to?
If the agency also sells managed services, the diligence bar rises. A managed detection and response checklist of questions to ask before committing is a useful model for how to structure that conversation, because the same principle applies: define scope, ownership, and exit terms before signing.
Inclusive Design and Performance Constraints
Ask: Do you design to performance budgets and test on low-end devices and slow networks?
This question separates agencies that design for their own fast laptops from those that design for real users. Ask what device and network profiles they test against, and whether performance budgets are part of the brief.
Inclusive design goes beyond screen readers. It covers contrast, touch target size, motion sensitivity, cognitive load, and language complexity. Ask how the team accounts for these in the design process, rather than only in a final checklist.
For buyers comparing vendors across categories, the pattern repeats: the questions that matter are about evidence, not promises. A buyer’s review framework like this Rillabox review of what buyers should know before ordering illustrates how structured questioning beats marketing copy every time.
Red Flags and Green Flags in the Answers
Some answers should make you pause. “We test with five users and that’s science” is a yellow flag. “Accessibility is handled by the developers” is a red one. “We’ll figure out metrics after launch” is another.
Green flags look different. The agency names a specific conformance level. It shows an audit that found real problems. It explains sample-size reasoning instead of reciting a number. It names the people who will do the work. It puts measurement and IP ownership in writing.
A practical habit that works well: send your questions in advance and ask for written answers. Verbal answers in a sales call are shaped by the room. Written answers reveal how the team actually thinks. If an agency resists putting process commitments in writing, that itself is information.
Questions to Ask, Organized
| Area | Question to ask | What a strong answer looks like |
|---|---|---|
| Accessibility | Which WCAG 2.2 level do you target, and how do you verify it? | Names a level, shows a recent audit |
| Research | How do you set sample size and include assistive-tech users? | Explains reasoning, shows recruitment plan |
| Metrics | Which validated instruments and behavioral metrics do you use? | Names SUS or similar, sets a baseline |
| Handoff | How do you deliver design tokens and documentation? | References a documented design system |
| Team | Who does the work, and is anything subcontracted? | Names roles, discloses subcontractors |
| Contract | Who owns research artifacts and source files? | Clear ownership terms on termination |
| Performance | Do you design to performance budgets? | Names device and network profiles |
Conclusion
The best time to ask hard questions is before the contract, not after the first deliverable disappoints. Accessibility conformance, research method, measurement, handoff, team composition, IP ownership, and performance testing are the seven areas that separate a design partner from a design vendor. None of them require technical expertise to ask. They require only the willingness to slow down for one meeting.
Agencies that welcome these questions tend to be the ones worth hiring. The ones that deflect them are telling you something useful too. Either way, you leave the room with better information than a portfolio alone could ever provide.
Frequently Asked Questions
How many design companies should I talk to before choosing one?
Three is a practical number. It is enough to compare approaches and pricing without turning the search into a full-time job. Ask each the same written questions so the answers are genuinely comparable.
Is a bigger agency always the safer choice?
No. Larger agencies offer more capacity but often assign junior staff to execution. Smaller studios may give you senior attention but have less bench depth. Ask who specifically will do the work and how continuity is protected if someone leaves.
What should I expect to pay for a UX audit?
Pricing varies widely by scope, number of screens, and whether usability testing is included. Rather than shopping on price alone, ask what the audit covers, whether it includes testing with real users, and what deliverable you receive at the end.
Do I need user research if I already know my customers?
Usually yes, but the scope can be modest. Internal knowledge tends to reflect the loudest customers and the most familiar workflows. A small round of testing often surfaces problems that no amount of internal discussion would have caught.
How do I verify accessibility claims?
Ask for a recent audit report and check whether it names specific success criteria and conformance level. Then ask how issues were prioritized and fixed. A credible agency will show failures alongside fixes rather than claiming a clean pass.
What if the agency’s process does not match how we work?
Say so early. Good agencies adapt their process to client constraints, whether that means working inside an existing design system or joining an established sprint cadence. If an agency insists there is only one way to work, that rigidity will show up again mid-project.


