Find out what people actually want.
Clients rarely know what they want, and users rarely say what they need. Research is how designers stop guessing. Here's how professionals uncover the colours, styles and feelings a project needs before drawing anything.
From what people say to what you design
- ListenTalk to 5 real customers. Write down their exact words, not what you hoped they'd say.
- Dump it allEvery quote becomes a sticky note. No judging yet.
- Find patternsCluster similar notes. Two themes appear: time and trust. Designers call this affinity mapping.
- Write the insightOne sentence that explains the behaviour. This is what the design must solve.
- Turn it into designFast → one big button. Not cheap → calm colours and quality type. Research just chose your palette.
Design research has three targets
The business
What does the client need to achieve? More sales, a new audience, looking more premium?
The people
Who is it for? What do they care about, fear and already love? Where will they see it?
The market
What are competitors doing? What's overused? Where's the empty space to stand in?
Good design sits where all three overlap. A gorgeous logo that ignores the audience fails. A safe logo that looks like every competitor fails too.
Before you buy a birthday present, you find out what your friend likes, what you can afford, and what everyone else is already giving them. That's market research!
People are bad at telling you what they want. Watch instead.
Say ≠ do
Economists call it stated vs revealed preference. People say they want "clean and minimal" and then click the loud, colourful one. Trust behaviour over opinions.
Jobs to be done
Clayton Christensen's idea: people "hire" products to do a job. His team found many morning milkshakes were bought to make a boring commute bearable. Ask what job your design is hired for.
Faster horses
"If I'd asked people what they wanted, they'd have said faster horses." It's usually credited to Henry Ford, though there's no proof he said it. The point stands: ask about problems, not solutions.
The Kano model: three kinds of wants
Professor Noriaki Kano (1984) showed that features affect happiness in different ways:
Basic needs
Nobody thanks you for them, but everyone's angry if they're missing. Readable text, working buttons, correct logo.
Performance
The better you do, the happier people get. Faster, clearer, easier to find.
Delighters
Unexpected extras: a lovely animation, a witty line, a hidden detail. These make people talk about you. Today's delighter is tomorrow's basic.
Research methods you can do this week
| Method | What you do | What you learn | Effort |
|---|---|---|---|
| Stakeholder interview | A 45-minute chat with the client: goals, audience, competitors, what they hate. | The real problem behind "make a logo". | Low |
| User interviews | Talk to 5–8 real customers about their lives and habits, not your design. | Motivations, language, pain points. | Medium |
| Competitive audit | Collect 10–20 competitor logos, sites and packs on one board. Note colours, fonts and tone. | Clichés to avoid and gaps to own. | Low |
| Moodboard | Collect images, colours and type that capture the feeling. Make 2–3 different directions. | Visual agreement before you design. | Low |
| Five-second test | Show a design for 5 seconds, hide it, ask what they remember. | Whether your hierarchy works. | Low |
| Preference test | Show 2–3 options and ask which fits "trustworthy" (or another word), and why. | Which direction matches the brand words. | Low |
| Usability test | Watch people try to complete a task with your design. Don't help! | Where they get confused. | Medium |
| A/B test | Show two live versions to real traffic and measure clicks or sales. | Hard numbers on what performs. | High |
| Surveys | Short questionnaires to many people. | The "what" and "how many", rarely the "why". | Medium |
Nielsen & Landauer's model suggests testing with about 5 users uncovers most usability problems (roughly 85% in their model). Several small rounds beat one big test.
Nielsen, J. & Landauer, T. K. (1993). A mathematical model of the finding of usability problems. Proceedings of ACM INTERCHI'93."This, not that": turn feelings into decisions
Branding agencies run workshops with sliders like these. The client places each one, and the answers become colour and type decisions. Try it: move the sliders and watch the direction change.
Every project starts with a brief
A brief is a one-page agreement on what you're making and why. No brief means endless revisions. Copy this template and use it for every project, including personal ones.
Where the work is heading
What this means for you: pure print-only graphic design is shrinking. Design that lives on screens, in systems and in motion is growing. Learn the fundamentals in this guide (they never expire), then aim them at digital work.
Brand identity
Logos, colour and type systems, guidelines. Highest creative status, very competitive.
UI / product design
Apps and websites. Best pay and growth. Needs graphic skills plus UX thinking.
Marketing & social
Ads, social posts, campaigns. Lots of jobs and fast-paced. A great first job.
Motion design
Animated logos, explainer videos, UI animation. Rising demand, fewer people do it well.
Packaging
Physical products on shelves. Combines branding, print and 3D thinking.
Illustration
Custom drawings for brands, editorial and products. The antidote to stock-photo sameness.
Adobe's analysis of BLS wage-survey data shows counted graphic-design employment fell from about 214k (May 2024) to about 198k (May 2025), while the median wage rose. Routine production work (resizing, simple social posts) is being automated. Taste, strategy, research and systems thinking are what clients still pay humans for. That is exactly what this guide trains.
Sources: U.S. Bureau of Labor Statistics, Occupational Outlook Handbook & OEWS (May 2024); Adobe, "The Economic State of Creative Professions" (2026).Homework
Pick a real local business
A chai stall, gym, bakery or tutor. Fill in the brief template as if they hired you.
Competitive audit
Collect 10 competitors on one Figma board. Write 5 observations about colour, type and tone.
Interview 3 customers
Ask why they go there, what they'd change and which other brands they love. Write down their exact words.
Two moodboards
Make two opposite directions. Run a preference test with 5 people using the personality words.