you can get in touch with us
UX research isn’t just “a stage in building your e-commerce site”. It’s something that genuinely helps you build better sites, ones people want to use and know how to use.
If you understand how your users behave, what frustrates them, what they need and what motivates them, you have an enormous advantage. That knowledge can’t be guessed. It has to be researched. And that’s exactly what UX research methods are for.
In this article we’ll go through the basic research approaches; I’ll explain the different ways of dividing them up and how to understand these methods. I’ll show you when it’s worth using them and how to match them to the stage your product’s life cycle is at, or to the kind of data you want to get.
This piece is for you if you want to quickly get a handle on the fundamentals of UX research and understand how to use them in practice. We won’t get into specific tools or the details of data analysis here, we’ll focus on what genuinely helps you make better design decisions, and when and how to choose a given method.
UX research is a set of different methods. You don’t need to know all of them right away; here’s a quick rundown of the most important ones
Usability testing + Guerilla testing– users carry out specific tasks in the product, and you observe where things work and where they fall apart. These can be moderated or unmoderated, remote or 1:1.
Usability benchmarking – a more “numbers-based” version of usability testing. The same tasks, specific metrics, and comparing results over time or across products.
Field studies / ethnography– instead of inviting the user to you, you go to them. You observe how they really use the product in their own environment.
Contextual inquiry– something between an interview and observation. You sit with the user and go through their real tasks together.
Participatory design + co-design – you hand some of the control to users. You give them materials and ask them to design a solution themselves.
Focus groups– several people, one moderator, and a discussion. Good for gathering opinions, but not always for observing real behavior.
Interviews– 1:1 conversations that let you dig deeper into a user’s needs, motivations and experience.
Eye tracking– you check where a user actually looks on screen, not where they think they looked.
Concept testing– you show an idea, even a very early one, and check whether it makes sense at all from the user’s perspective.
Heuristic evaluation – a pragmatic assessment of a product by an expert; quality here is tied directly to the expert’s own experience.
Diary studies– users document their experience over time (notes, photos, apps, say).
User feedback– everything users leave behind “along the way”: forms, emails, feedback buttons.
Desirability studies– you check how users respond to different visual versions and which one they prefer.
Card sorting– users group content in whatever way is logical to them. Excellent for information architecture.
Tree testing– you check whether users can find something in the structure, with no visual layer involved.
Analytics (Google Analytics 4 or heatmaps, say)– data on what users actually do: clicks, paths, drop-offs.
Clickstream analytics– you look at a user’s full sequence of steps in the system.
A/B testing– you compare two versions and check which performs better based on real behavior.
Surveys– a fast way to gather opinions and declared data from a larger sample.
To get a better grip on these research methods, it helps to look at them a bit more broadly, not as a list, but as a system.
At first glance, UX methods can look like a random list of techniques. A bit of interviewing, a bit of testing, a bit of analytics… and it’s not entirely clear what to pick when.
But there’s a simple way to think about it.
Instead of memorizing methods, it’s better to look at them through three questions that help you choose the right one.
- What are you actually studying? Are you observing the user in their natural environment and trying to understand a situation, or are you already testing a specific solutionyou’ve designed?
- What kind of data do you want to gather? Are you interested in numbers and measurable results (quantitative), or deeper understanding and context (qualitative)?
- )? What do you base your data on?What users say (declared), or what they actually do ()?
behavioral
1. C)? This model shows that UX methods aren’t a random grab bag of techniques. Each one answers a slightly different question, is suited to a different goal, and fits a different stage of a product’s life. That’s exactly why choosing a method should never be random, it should be matched to what you actually want to learn.
What are you actually studying?
Studying the situation vs the solution
Are you trying to understand a problem, or assessing a solution? UX research methods are used throughout your product’s whole life cycle. The X axis splits methods by what they typically study: either the situation (the problem) or the solution being built. At the start of a project you usually don’t have anything to test yet. You want to understand the
situation : how people act, what constraints they face, where problems show up. Field studies, diary studies or interviews work well here. But once you have a
solution
| (a prototype or product), the perspective shifts. You start checking whether it works, whether it’s understandable, whether it does its job. That’s the moment for A/B tests, usability testing, concept testing or even surveys. | In short: first you understand reality, then you test your ideas for improving it. | Area |
|---|---|---|
| Studying the situation | Studying the solution | What they assess |
| Analyzes the context of a solution | Analyzes a solution in context | Findings |
| Identifies opportunities and constraints | Helps set priorities and build benchmarks | Use |
| Used to build a picture of the current situation | Used to refine working prototypes | Example methods |
e.g. literature review, diary studies e.g. concept testing, usability testing
2. What data do you want to gather?
Qualitative vs quantitative research
What kind of answers are you looking for?The Y axis splits methods by the type of data they typically deliver: more qualitative or more quantitative.If you want to understand why something is happening,.
to dig deeper into needs and context, you go for qualitative research. Interviews, usability testing, focus groups all help you seewhat’s behind the behavior. If instead you care about
the scale of a phenomenon, how many people have a given problem, how often something happens, then quantitative methods come in. Surveys, A/B tests, analytics.
| Solid research should combine both approaches, because they give different perspectives on the user experience | (What UX research involves and why it matters for CRO). | Area |
|---|---|---|
| More qualitative | More quantitative | What they assess |
| Assess the quality of the experience | Assess the scale/volume of a phenomenon | Findings |
| Identify needs and inspire solutions | Identify patterns and support KPIs | Use |
| Most often: exploration and validation | Most often: observation and iteration | Example methods |
3. NUser interviews, heuristic evaluationUser surveys, behavioral analytics
What do you base your data on?
Studying declarations vs behaviorDo you need to hear it, or observe it?This distinction can be boiled down to:
“what people say” vs “what people do”, and those things often differ.Studies of
declarationsare excellent for understanding a user’s perceptions, beliefs and language, but they’re limited by what people are aware of and willing to reveal. Examples: card sorting (mental model), surveys (opinions), focus groups (perception of a brand/concept).Studies of
behavior.
| focus on how users actually use the product: A/B tests (the effect of changes on behavior), eye tracking (visual interaction). | If your project’s stage allows for it, it’s worth mixing these methods, because the gap between what people “say” and what they “do” gives you the most interesting insights | Area |
|---|---|---|
| Declarations (Attitudinal) | Behavior (Behavioral) | What they assess |
| What users say and declare | What users actually do | Findings |
| Opinions, beliefs, mental models | Real behavior and interactions | Limitations / strengths |
| Limited by awareness and honesty of responses | More objective, grounded in real actions | Example methods |
Surveys, card sorting, focus groups
A/B tests, eye tracking
To sum up, UX research is a toolkit you draw on depending on the stage you’re at, what you want to learn, and the balance between what users declare and how they behave. In the next section I’ll tell a story that shows, through an example, how these methods can be applied.
OK, theory is theory, but what do you actually do in practice?
I’ve told you a lot already, but the model alone isn’t quite enough. So instead of another table, let’s go through this with a concrete example.
Imagine you’re building an e-commerce business selling designer clothes made to order by students. Something unique, more slow-fashion than Zara.
Start: you have an idea, but you’re not sure it makes senseYou don’t turn on analytics or A/B tests yet, because you have nothing to measure.:
- First you want to understand the
- situation:
- do people even want to buy clothes made by students?
what matters more to them: price, uniqueness, personalization?
what are they worried about? (quality? returns? lead time?)
You talk to prospective users (interviews/focus groups), observe how they buy clothes online (field studies), and if the topic is more complex, you walk through the whole purchase process with them (contextual inquiry).
→ When to do this? When you don’t have a solution yet and worry you’re designing “in a vacuum”.
The first ideas start to emerge
- You have some direction now: customization, designer stories, limited drops.
- Instead of building the store right away, you test it quickly:
- you show mockups of the product page,
you check reactions to “design your own outfit”,
you ask whether anyone would be willing to pay more for something unique.
You can also try co-design, asking users to lay out their ideal product page themselves.
→ When to do this? When you have a direction/idea but don’t want to invest time and budget in a full prototype yet.
You have a prototype (your first solution)
- A clickable store prototype exists now.
- And this is where usability testing comes in:
- do users understand how to order a custom-fit outfit?
is the personalization process too complicated?
do they know how long they’ll wait and what they’ll get?
You give them concrete tasks and watch where they get lost.
→ When to do this? Always before shipping anything bigger.
- You have a working product, so it’s time for numbers
- The store launches. Now you want to know what’s working and what isn’t.
- Benchmarking: is conversion growing?
Surveys: why does someone abandon their cart?
A/B tests: does “Design your own outfit” work better than “Create something unique”?
→ When to do this? When you need decisions grounded in numbers, not just intuition.
The product is alive!
People are using the store. And suddenly you’re seeing things nobody told you in an interview.
Through tools like Google Analytics 4 or heatmaps, you see where users click, where they drop off, or where they pause. 🙂
Clickstream shows you their whole path, and feedback (emails, forms, tickets) gives you context. It might turn out people said they wanted personalization… but don’t use it, or the opposite, they spend the most time there..
→ When to do this? All the time
To sum up, UX research isn’t a single choice, it’s continuously adapting to what the project needs To close – what’s worth remembering?
If you’ve made it this far, you can probably already see one thing: UX research isn’t a simple checklist of rules. It’s more a way of thinking.Research isn’t black and white.There’s no single “correct” approach and no one perfect method. In practice, what wins out far more often is
creatively combining techniques and matching them to the situation,rather than sticking rigidly to a formula.Over time you start to feel it. You see when something “isn’t sitting right”, when you need more context, and when you need more data. And that’s where
experience does an enormous amount of the work.
- So don’t be afraid to make mistakes.
- UX research isn’t “pick one method and go”. It’s constantly switching between:
- understanding the problem
testing solutions measuring the effects
And if there’s one thing to take away from this article, it’s this: there’s no single good method, only a good decision for that particular moment.And if you feel you need support, whether in diagnosing your situation or choosing the right methods,
