All posts
UX Research Topics
Podcasts
Field Guide
SUBSCRIBE TO OUR NEWSLETTER
FOLLOW US
Whether you’ve run a study before or are a complete beginner, this post is a great refresher on UX research.
User experience research is the practice of identifying the behavior patterns, thoughts, and needs of customers by gathering user feedback from observation, task analysis, and other methods. In plain terms: UX research uncovers how a target audience interacts with your product to gain insights on the impact that it has on them.
From our observation, companies typically employ user experience research to improve their product design, but can struggle to make that research and user testing as impactful as they’d like.
Questions that typically arise include:
The good news is: you can conduct user experience research in a cost-effective way that helps to improve your product. We’ve compiled a number of best practices in this guide.
Whether you’ve run a study or two, or are a complete beginner, this will be a great refresher along with some additional advice from our years of experience in the industry. By the end of this guide, you’ll have a higher confidence in your approach to UX research and be able to do it both faster and better.
Note: Looking for a specific audience to participate in your UX research? User Interviews offers a complete platform for finding and managing participants. Tell us who you want, and we’ll get them on your calendar. Sign up for free.
Note: While this is a general overview of UX research with some helpful tips mixed in, our UX Research Field Guide has in-depth information about each step of the process as a go-to reference. In it you’ll learn the ins and outs of how to conduct more and better research, and how to use it to help your team.
User experience research is an investigation of users and their needs. It’s a methodological way of observing how people interact with your product to gain insights that help you improve it.
How do you define “user experience?”
User experience is the overall interaction and impressions a person has when using a product or design. Often, user experience is evaluated in terms of whether using a product is intuitive or challenging for someone to use.
What is a “good” user experience (UX)?
Is your design logically organized for how someone will use it? In other words, can a person use your product successfully and with relative ease? If a user experience matches the abilities, limitations, needs, and values of a target audience, that is a great user experience.
What is UX design?
UX design is the process of creating helpful, accessible, and appealing designs for users to interact with. Whether you’re designing a website, an app, or a physical product, insights from UX research get applied as UX design.
What is the purpose of user experience design?
Every company wants to ensure a pleasing experience for their customers, otherwise they will leave. UX design helps encourage customers to engage with and continue to return to the company's product.
Should you do user experience research given limited time and resources?
Every study and product is different. That said, the return on an investment in research tends to be worth it because of the time and cost saved in the long run. For example, one Forrester study showed that companies reap $100 for every $1 invested in UX design. There’s also a great video about what is so costly about skipping user experience research.
User experience research is so effective because it helps you make design decisions based on real evidence. Understanding users, their needs, and the design requirements to meet those needs helps teams:
When you base your company and product on the data from user experience research and testing, it’s a no-brainer that doing so would help the odds of your success.
When should you conduct user research for your project? It might surprise you to hear that you can do research at any stage.
We’ve noticed that even some veteran researchers don’t plan for testing at multiple phases. Allocating budget for testing early on, during, and at the end of the project lifecycle can actually save resources and increase the value of your product overall.
Think of your research as a funnel where broad questions get asked at the top and specific details at the bottom. Conveniently, this will generally match with your product development phases.
These are the general research guidelines we recommend:
Findings from the earliest research you do will typically have the biggest impact on your project. In most cases, allocate more time and budget to the early stage. After all, it’s best to get as many useful insights as possible before investing further.
At this stage, you’re finding out who the users are and their general challenges. Broader UX research methods come in handy in the early stages.
Thinking in terms of the UX research funnel, you’d likely place ethnographic research at the top. Learning about people and their lives in a way that is generally representative of the user helps confirm you are on the right track and notice common themes.
To illustrate this: Let’s say your business is a grocery store. With your current resources, you can either improve the in-store experience or create a grocery delivery app for your customers to order with from home.
During early development, you would research how your users behave, clustered in terms of demographics and general habits. You might find that your customers are mostly young people in their 20s and they order groceries delivered more often than they go in to visit the store. Your customers belonging to older demographics prefer to shop in-store, but they make up a much smaller percentage of your users.
Now, you can conclude that a delivery app makes more sense as an investment based on the research. With this information you can save time and resources that may have gone into improving the in-store experience (and even the resources that may have gone in to targeting those older demographics with advertising or content).
Your project likely has milestones during its development. See where along those milestones it might make sense to do further design research—where can you reduce uncertainty and risk with further insight?
Continuing with our grocery delivery app example, you can do further testing of your dominant user base of young people in their 20s and get a closer look at their behaviors and needs to make sure you perfect your delivery app for them.
Now is a good time to answer questions like:
For example, you might discover that users often forget an item in their grocery delivery order. You might use that insight to prioritize a grocery list feature in your app to make it easier for them to remember to add it to their order.
Continuous improvement on a product has a high value for the business and users. So, conserve some budget for research toward the end of the project (and even periodically once your product is live) to get insights from users interacting with the completed product.
With the grocery store’s delivery app example, we understand who the user is in terms of high level behaviors and demographics thanks to research we did in early development. We know what their pain points are, like forgetting to order certain items. Now, you can test the grocery list solution you’ve created to see whether it meets your users’ needs as you expected.
This is when usability testing and validation are most important. A completed beta version of your project gives you the opportunity to work out any kinks and adjust it to match user needs where it lacks. For example, you might decide to do A/B testing of your design to see where placement of the grocery list in the app works best so that it’s intuitive for your users to find.
As you do more testing, you might do user interviews and get feedback that some people aren’t using the grocery list feature and are still forgetting things in their order.
So, you decide to test notifications from the app to remind them to complete their grocery list, or a modal to suggest adding a product they’ve bought in the past. All this is to help improve their experience in ordering groceries from you and increase the amount of groceries you sell!
This is the type of continuous refinement and improvement that unlocks the value of user experience research for your business and customers.
At a basic level, a user research test should be a replicable process of learning and finding the user’s needs. It should fit the design process at any stage.
You’ll first do project planning for all of the logistics. Then, the actual testing. Finally, you’ll use it to refine and improve your product. Rinse and repeat.
Note: User interviews are appropriate at every stage of the product development process. Why? They’re incredibly flexible and give rich data and context around your users for whatever you want to learn.
For the big-picture view of a research workflow, the 5 Steps of UX Research created by Erin Sanders, also called the Research Learning Spiral, are a useful framework for thinking about the main research steps:
Like any scientific test, you’ll first determine the questions you want your research to answer. Then, you’ll identify assumptions about your users to test. Next, what research will you do to answer these questions and test your assumptions?
Once you’ve answered the questions above and factored them into a one-page research plan that you can present to stakeholders, you can start gathering the knowledge you need through the selected research methods:
Organizing these first steps into a simple document as a research plan will help you and the team stay organized. Now, you can begin conducting your research.
As you synthesize, be open to new discoveries and opportunities.
Now, we’ll go over some common user research methods and the scenarios to use them in.
There are a lot of project-specific variables that will affect this. So it’s helpful to divide methods into larger groups.
Depending on the questions you want to answer, the research you conduct will likely consist of a combination of quantitative research and qualitative research.
Quantitative research is measured using numbers, often in large sample sizes. For example the data could be, “How many users filled a form?” or “What percentage of people clicked on this button?”
Use quantitative methods to find what users are doing and the statistical likelihood of it happening in the future.
Qualitative research is more open-ended and unstructured than quantitative research. For example, the data could be a narrative description of user behavior by the researcher or interview answers from the participant. It might answer questions like, “Why didn’t some users fill out the form?” or “What was it on the page that caused people to click or not?
Use qualitative methods to find why users are doing what they do.
As a method, user experience research interviews are accessible to companies of any size, at any phase of the development process.
The insights from interviews are well worth the effort — as long as you’re asking good questions.
User interviews can be conducted one-on-one or with multiple people in a focus group. There a few interview types. Which type you use depends on your project and goals.
Directed Interviews / Discovery:
A researcher asks the participant questions they answer. This is your “standard” interview and the one you would most commonly conduct.
You can gain insights from directed interviews whether you’re interviewing a few users or a hundred. You can easily compare and contrast the answers from different users.
Non-Directed Interviews:
Think of non-directed interviews as opening up a conversation between the researcher and user. Rather than the question-answer format, the researcher will establish some general guidelines for the conversation and let the interviewee do most of the talking.
Non-directed interviews are useful when you’re learning about touchy subjects that are emotional for users, and when you want answers for questions that participants might not respond well to when asked directly.
Ethnographic Interviews:
This type of interview is the most observational for the researcher. In ethnographic interviews the researcher will observe how the participant behaves in their own environment, such as work or at home.
Ethnographic interviews are useful for seeing the process it takes for users to accomplish tasks, understanding the context for how they use a product, and looking at user behavior when what they say may not match up with what they do.
So, how do you find interview participants?
You’re likely already aware that recruiting the right participants for interviews is crucial.
Experienced researchers know to screen participants ahead of time to have them lined up for the study. But the time and resources that go into finding and screening participants can quickly add up between:
These challenges are increased when it comes to finding niche participants. For example, you may have criteria for their profession, geography, or demographics. Limiting your pool of test subjects for any reason can make screening slower and more expensive.
You can study participants in person or online, but online is generally the easier, more flexible method for finding and screening participants with particular attributes. Plus, it’s also easier to record and conduct your study.
Note: At User Interviews, we specialize in finding high quality participants to fill your study as quickly as possible. If we don’t have participants in our database, we’ll go find more who fit your criteria. Tell us who you want, and we’ll get them on your calendar. Sign up for free.
Here are some additional tips we have for screening participants and conducting online interviews.
While interviews are a highly valuable research method, they are best when combined with other UX research tasks. Check out this Awkward Silences podcast episode for more on conducting remote interviews.
Combining different tests at different stages helps you gain better insight.
Here are the other tests you can consider conducting and when to use them:
Surveys and questionnaires are best used as a bookend: very early on in the design stage or at the end of testing. They are less suited to the middle of development when product changes are more dynamic.
Doing a survey up front is a good screening strategy before doing an interview to make sure they have the pain points you are trying to solve. Once development is near completion, you can follow up with the same users to test out your new version and see if you’ve succeeded in providing a solution.
Surveys and questionnaires aren’t limited to the research process—you can use them for the live product. When you launch, survey for what real users found valuable to get feedback. If you save their contact information, this is a method of passive recruitment for future tests as much as it is data collection.
That being said, surveys almost always suck, according to Erika Hall.
Card sorts are best-suited for the middle or end of the design process. They help you figure out how to organize your design in terms of the order and grouping of items before it gets made. It can also help you improve its organization as a prototype.
This testing method helps to reveal how users cluster the things on the cards in their heads. You can use their grouping inclinations to inform your usability in terms of classification and information architecture.
For example, as shown above, if you’re designing a menu in your grocery app to organize beverages: do your users tend to group fizzy beverages all together? Or, do they divide sparkling beverages into alcoholic drinks like sparkling wine and non-alcoholic drinks like sparkling water? Sugary drinks and non-sugary drinks?
Card sorts are best done unmoderated to avoid bias, as users will sometimes attempt to read your signals for how to sort things. Ultimately, it should tell you where to place different pieces of information relative to each other for a good customer experience.
A tree test (more details here) is a simple test that you can conduct unmoderated. Think of tree tests as a reverse to card sorting. They are good for testing during the middle of the process when you are still organizing your design’s structure. With a simple text representation of site or information architecture, you can get quantifiable data on participant behavior.
For example, you might find that when asked to place an order, 30% of participants clicked on “contact” instead of the purchase button and take steps to correct it.
Once you have a version of your design in place, you can begin conducting actual usability tests.
Usability tests can be moderated or unmoderated, depending on the context. For example, you might present a task, such as saving a playlist, and observe how they figure it out.
For moderated tests, you’ll give clearcut instructions for users to follow. However, clear instructions are just as important when conducting unmoderated usability tests. Without a list of things to do next, your test users can get lost very quickly. Present them with challenges such as, “create your grocery list” and observe how they go about doing it, then present them with another task to complete.
It’s also helpful to walk through an unmoderated usability test yourself or send it to colleagues to confirm it all makes sense before performing an official test.
.png)
Guerilla tests are designed to fit in with time constraints. Usually, it involves presenting your prototype or feature to whoever may be available and asking for their thoughts.
For example, you can stand on the street with your product and ask passersby to use it and give you their feedback.
This a less structured method of usability testing that may not hold up on its own, but guerilla testing can still yield great insights when time is of the essence and confidentiality is not important.
A/B tests are also great when you want to get fast information on your prototype’s usability.
Presenting two design versions, you can see the differences in user behavior between them and keep the iteration that is easier to use.
By necessity A/B testing is unmoderated. There are numerous research tools for A/B testing such as Optimizely, VWO, and Adobe Target.
Once you’ve completed your study consisting of one or several of the above methods, you can use that data to communicate your findings and apply them to the project.
Usually, product managers and other stakeholders will want to see the insights and recommendations derived from user research. Creating reports, presentations, and recommendations with this information will help to communicate what you’ve uncovered.
The steps for creating deliverables and applying them to the design process are:
1. Identify general trends: What conclusions can you reach by interpreting the information you’ve gathered broadly?
2. Analyze quantitative data: What do the numbers tell you about how people are using your design? For example, how long does it take the average user to complete a task, or do certain users gravitate more toward different features?
3. Analyze qualitative data: Turn your observations of user behavior into further information about the users. For example, what did they like the most or least about the product? What common patterns and themes were there in user responses?
4. Organize your data: Use a visual tool such as a spreadsheet, a wall of post-it notes, or software like Delve, MAXQDA, and Dedoose to organize and present research data.
5. Make recommendations: Based on the data, you can make recommendations to apply to the next product iteration. That can mean anything from eliminating a pop-up to changing the placement of a “SHOP NOW” button, or even adding entirely new functionality.
Finally, you’ll turn the recommendations into concrete changes on your design, and then test it again.
We have our own storytelling structure for reports and presentations designed for UX researchers and UX designers, along with best practices in our field guide.
There are several specific report types you may include in your report or presentation.
Usability Reports: What were the usability findings? These reports provide information on what the test plan was, the methods used, and key metrics displayed as visuals.
Analytics Reports: What is the UX trendline as the user interface (UI) changes? Is the UX improving over time? For example, did form completions, sign ups, or purchases rise or fall between the current product design and the last?
Competitive Analysis Reports: How does the product or service compare to competing solutions on the market? Different solutions for the same base of users will have different strengths, weaknesses, and other attributes. Identifying where your product fits in helps find areas you are doing well in, see where to improve, and where you can compete.
As another deliverable, many UX research will create personas to represent the different user groups that make up the target audience.
Chances are, you won’t have a single type of user, but different categories of users. Creating personas of the different types of people your product will impact can help tailor its functions to them.
To create personas, take note of common differences in user:
Personas can seem trite if you haven’t created them before. However, they can be very useful internally for product improvements and externally for presenting the array of different users to stakeholders.
At this point, you have a bird’s-eye view of the user experience research landscape.
The more testing you do, the quicker and more streamlined it becomes as your skills with testing improve with practice. Your capacity to create the best possible version of a product will also improve.
We know that finding, screening, and interviewing users for research can seem like an intimidating process. It doesn’t have to be overwhelming though.
We’ve devoted our own UX research and design capacities to creating the best possible platform for you to find research participants that really help your project.
Note: Looking for a specific audience to participate in your UX research? User Interviews offers a complete platform for finding and managing participants. Tell us who you want, and we’ll get them on your calendar. Sign up for free.
Ziev Beresh is a writer and content marketer who specializes in long-form articles for agencies, SaaS, and startups. When he’s not writing, he’s either reading, cooking, or enjoying the outdoors with his adorable 3-pound toy poodle.
