UX Research on a Shoestring: Five Methods Under $500

'We don't have a research budget' usually means 'we've never tried the cheap methods.' Five users, a laptop and your support inbox contain most of the truth expensive studies find: moderated hallway tests, replay analysis, ticket mining, five-second tests and exit questions cover discovery through validation for almost nothing.

Taste is a process

Beautiful outcomes come from boring loops: references argued over, directions killed early, details iterated after everyone is tired. The myth of the inspired genius mainly produces missed deadlines.

Our reviews run on questions, not moods: does this serve the user's next decision, does hierarchy match importance, would this still work with real ugly content. Opinions survive only if they improve answers.

A consistent process beats occasional brilliance — for clients, and honestly for designers too.

The design QA ritual before handoff

1. Real content in every state: empty, error, loading, translated-long
2. Squint test per screen: exactly one obvious next action
3. Contrast pairs measured, not squinted at
4. Touch targets verified on the smallest supported device
5. Every interactive element has hover, focus, active, disabled
6. Loom walkthrough recorded for anything with motion or state

Hierarchy is the product

Every page secretly asks the user to do one thing next. Design's job is making that next thing obvious to a tired stranger holding a phone in bad light. Everything else is garnish with ambition.

We audit hierarchy with the squint test and the five-second test, not with debates: what reads first, second, third — and does that order match the business's priority list.

When everything shouts, the user leaves. One voice, clearly louder, converts.

Taste debates end where user questions begin.Studio wall print

Delight must earn its budget

Motion, illustration and charm are budget line-items like anything else; they belong exactly where doubt and boredom cost money — onboarding, empty states, waiting moments — and nowhere they're stealing attention from decisions.

The test is whether the delight carries information: progress, confirmation, orientation. Animation that answers a question pays; animation that asks for applause should audition elsewhere.

Design debt compounds too

Buttons that drifted, greys that multiplied, spacing that means nothing twice — interface entropy is real, and it taxes every feature after it. Small teams feel it as 'why does every screen take longer now.'

The fix is periodic consolidation masquerading as housekeeping: tokenize the greys, kill button variants to three, formalize spacing on one scale. Unglamorous weeks that speed up the year.

Users may not name inconsistency, but they feel it as untrustworthiness. That's a conversion metric wearing a trench coat.

The fold isn't dead; indifference is.Mahnoor Qureshi, head of design

Where to start this week

This week: watch five session replays end-to-end and read your last twenty support tickets. Write down every time reality disagreed with your design assumptions.

Then keep it honest with a short list:

• 'Make it pop' accepted as actionable feedback
• References from three industries and zero analysis of why
• Approval by committee poll instead of one accountable owner
• Put the date on the calendar — playbooks without Fridays are just reading

And when the scope outgrows the spreadsheet, that's precisely what our team is for.

Straight answers

How many design concepts do we get?

One direction, done properly — after a shared strategy phase so the direction isn't a guess. Scattergun concepts look generous and are actually the agency hedging; we'd rather spend those hours making one direction undeniable.

How do you handle feedback rounds?

Two structured revision rounds per milestone with consolidated comments from one decision-maker. Async by default, calls for anything subjective — taste discussions die in email threads.

Can you redesign without us losing brand recognition?

Evolution is our default unless strategy says revolution. Equity audit first: what's recognized and must survive the refresh, then invention everywhere else.