Research aids in decision-making at every stage of product development, helping product managers and other team members reduce risks and minimize the costs of testing ideas. Of course, research cannot provide 100% certainty or guarantee success in every project, but it can save teams and companies significant time and money.

Qualitative research is the most common tool in a product manager’s arsenal for gathering user information and testing hypotheses. However, there’s a temptation to rely solely on qualitative research, sometimes unnecessarily. As a result, it’s important to understand the benefits (and limitations) of qualitative research, which we will discuss below.

In this article, we’ve collected essential knowledge about qualitative research that can help product managers use it as effectively as possible. Over the course of this guide, we’ll talk about primary methods for qualitative research, when to apply them, and how to draw conclusions from their results. In addition, we will also review the stages of conducting user interviews and give practical recommendations for each of them.

This guide is a result of our collaboration with experienced researchers:

What is qualitative research, and what are its limitations?

Put simply, qualitative research can help answer the questions “how?” and “why?”.

Qualitative research can help product managers to formulate and refine hypotheses that can be tested with quantitative methods (strategies based on survey data, behavioral statistics, and more).

The goal of qualitative research is to understand the root of current or intended user behavior. Well-executed research can help product managers uncover how people think, what their motives are, how they structure their decisions, and how (and why) they use particular products.

Qualitative research methods cannot count things, estimate proportions, or identify the prevalence of certain characteristics within large populations of users. To reach those goals, quantitative research methods should be used. For example, it’s incorrect to conclude from a series of interviews that if 5 out of 10 clients mentioned a problem, 50% of clients have encountered the same issue. This is because qualitative methodologies do not imply an extrapolation of their results to a general population (ex. all users). In other words, you cannot say that all of your users will behave in the same way as a set of respondents do.

That said, this doesn’t mean that qualitative research is bad, wrong, or unreliable. However, it signals that this methodology has a different purpose: to help product managers understand the essence of user behavior. Qualitative research can assist PMs in figuring out the reasons why users act in particular ways. This is very important information for strategic decision-making.

In qualitative research, a sample of users doesn’t need to necessarily replicate the structure of the general population. For example, within the context of user interviews, you can select respondents based on different principles according to your research goals. In these instances, you can single out “typical” or “representative” users, or choose to focus on edge cases.

Types of qualitative research

In the broadest sense, qualitative studies include various types of interviews, group discussions, observations, and diary studies. Product teams most often use:

  • In-depth interviews — unstructured or semi-structured conversations with respondents. Here, problem identification and solution-oriented interviews are usually distinguished.
  • Usability tests — structured sessions that allow researchers to observe the completion of tasks within the product.

Diaries, focus groups, and expert interviews are used much less frequently.

In-depth interviews and usability tests are often combined in a single session. This means that it’s quite typical to conduct a usability test that contains several elements of an interview.

All of these methods can solve different challenges that a product manager encounters in their work.

Problem-solving interviews can help product managers understand the context in which users engage with a product. For example, product managers can learn more about users’ lifestyle, what problems they’re focused on solving, what tasks they’ve been assigned, and what they value.

Solution interviews provide product managers with an opportunity to test ideas, and discover whether potential solutions can address a user’s problems (and to what extent).

Similarly, usability tests allow PMs to assess how convenient a product is for solving particular user problems.

On the other hand, diaries bring research teams as close as possible to “real life” user situations, and allow PMs to track user behavior and usage dynamics in a natural context.

Focus groups and other formats of group discussion can help PMs understand how users discuss a product or problem in a group setting, and can provide additional learnings that may emerge from this microcosm model.

Expert interviews stand somewhat apart from other forms. They are not concerned with the personal experience of experts but can help PMs clarify the market structure, audience characteristics, and trends of a particular domain. These interviews are broad in scope because the experts often have a professional overview of markets as a whole.

→ Test your product management and data skills with this free Growth Skills Assessment Test.

→ Learn data-driven product management in Simulator by GoPractice.

→ Learn growth and realize the maximum potential of your product in Product Growth Simulator.

→ Learn to apply generative AI to create products and automate processes in Generative AI for Product Managers – Mini Simulator.

→ Learn AI/ML through practice by completing four projects around the most common AI problems in AI/ML Simulator for Product Managers.

When to conduct research

In short, qualitative research can be applied at all stages of product development.

All product discovery processes, one way or another, involve obtaining additional information when:

  • Launching an entirely new product (Ex. when only ideas, concepts, prototypes, or MVPs exist)
  • Entering a new market (Ex. another country or segment)
  • Expanding the functionality of an existing product
  • Collecting feedback on an existing product/solution
    • Interpreting behavioral metrics or results of experiments
    • Understanding a user’s general frame of mind, 
    • Gathering feedback for further feature improvements
    • Prioritizing existing work

Sometimes, qualitative research may not be necessary in these situations, as it may be impractical.

When not to conduct research

When results won’t influence a decision

When a decision has already been made, or changes to a product cannot be implemented for various reasons (ie. lack of resources), product managers should not prioritize research. 

This point seems obvious, but in practice, such situations occur regularly. For example, sometimes research requests are made when a team wants additional confirmation of their beliefs about a particular issue, and is not open to exploring alternative outcomes or recommendations.

Sometimes, there’s a desire to conduct research to broaden a team’s horizons more generally, but without a clear understanding of how research results will help a business. In this case, it’s necessary to assess the feasibility of research projects and consider the company’s ability to allocate resources for activities that won’t yield immediate results.

When it’s cheaper to implement and test immediately 

Sometimes, new features can be implemented very simply and quickly, and then tested on real audiences with minimal risk. In this case, conducting research would be an unnecessary expenditure of money and time.

When it’s impossible to reproduce realistic conditions 

Some user tasks are very difficult to reproduce in “laboratory” conditions. Qualitative research makes it difficult for product managers to identify users’ subconscious needs and goals, and to observe counterintuitive behavior.  It’s especially difficult for researchers to replicate multiple factors when trying to understand user behavior in specific contexts. For example, qualitative research is unlikely to be helpful when optimizing specific areas of an interface design in order to increase conversions or views.

How to conduct interviews

Now, let’s focus on the most popular research method in product development — interviews. For simplicity’s sake, the situations described below all pertain to B2C products. Conducting interviews in support of B2B products is more nuanced, and will not be the focus of this article.

Any research method consists of a protocol; a distinct set of steps. The content of each step may change depending on your research goals, but the focus of each step should remain the same. As with most qualitative research, when conducting interviews, follow the steps below: 

  • Goal setting
  • Research design
  • Recruiting
  • Conducting interviews (Field stage)
  • Analyzing results
  • Synthesis

Now, let’s look at each of these stages in more detail.

Setting goals

This is the most crucial stage for researchers for one simple reason—you cannot reverse it. If you do not correctly define a clear and specific research goal from the start, you may risk moving in the wrong direction. As a result, others will be unable to replicate and learn from the results of your work.

At first, goal setting as an objective may seem obvious. When goal setting, PMs need to understand and identify important preliminaries right away, including:

Research objective(s)

To identify a research objective, start by assessing the company’s short-term and long-term goals. How are they related? What business problems can you help resolve as a result of your research? Which metrics should you focus on impacting? Once you’ve answered these questions, you’ll be on your way to formulating a cogent research objective (or objectives).

A qualitative study can be guided by several objectives, or key questions to answer– and that’s okay. That being said, be wary of scope creep. Typically, interviewees’ capacity for qualitative research is limited, and few people will agree to talk with researchers for long periods of time (ex. two hours or more). Usually, you can cover two to three questions in detail over the course of an hour. For example, in a one hour interview session, you can:

  • Find out what tasks (jobs-to-be-done) users are assigned to
  • Discover user pains and needs associated with key tasks
  • Learn the procedures that users follow to complete jobs-to-be-done

Research Context

Barbara Minto, in her book The Pyramid Principle, provides an excellent framework for defining context for qualitative studies. For example, different contexts may change the direction of the research. In one case, a company could research users because it’s launching a new product and has no idea about user problems. Another context could be that the company is losing revenue for some reason. In the second case, you’ll research how the experience has changed recently and why.

In order to establish context around a particular problem, Barbara Minto recommends developing an understanding of three areas:

  • Situation: Understand how the company operated in the past, before a particular problem arose.What major decisions were made? What did leadership value?
    • For example, assume that the product team at Company X made a successful product for parents, and historically occupied a successful niche (upper-middle class parents).
  • Problem: Understand how the situation changed.
    • Using the example of Company X, perhaps an economic crisis occurred–which led to lower spending among upper-middle class parents. 
  • Question: Understand what consequences resulted from the problem, and begin to consider approaches to solving it.
    • For Company X, a question like “which new segment should we target?” could be identified here.

In some cases, executives may approach you with a fully formulated research question. For example, the leadership team at Company X might ask “Which new audience segment should we consider targeting?”  

In these situations, it’s still important for PMs to understand the historical context of a problem, and identify where the need for research originated. 

By establishing context, you can understand the high-level goals and limitations informing your qualitative research. 

Results format

After identifying goals and establishing context, PMs should define the expected format of their results. Sometimes, qualitative research can result in a visual diagram in Miro, a presentation, or a simple data table.

When considering potential formats, find out who, specifically, will be responsible for applying the results of your research. If it’s you, imagine the task(s) you might create as a result of your findings. Based on those tasks, identify the primary information you’ll need to leverage, and consider how you can make it accessible. 

When undertaking research on behalf of external clients, your results should be formatted based on the roles and objectives of each participant. For example, a marketer may be more interested in focusing on users’ emotions, while a product manager might focus on their jobs-to-be-done. On the other hand, CEOs and executives may be focused on developing a more general understanding of a particular segment.

In any case, it’s essential to report results in a format that is convenient for the end user. To find out which formats will be convenient for clients or stakeholders, be sure to ask them what specific decisions will be made based on your research–this can help you envision the perfect format for your research results.

Essential information about the target audience

Researching the wrong audience segment is a common mistake in the field of qualitative research. To avoid it, determine the key parameters that define your target audience up front. Who do you need to survey? Who should be excluded? By using specific parameters (based on things like interests, location, and age) you can ensure you’ve found the right people to engage with. 

Later in this piece, we’ll discuss how to use audience parameters to screen research participants.  

Research design

Effective research design translates business goals into a set of discrete research tasks. At this stage, researchers should focus on taking a question, deconstructing it, and identifying tools to help answer it.

To deconstruct a daunting research question, break it down into small components.

For example, if you’re conducting in-depth interviews to identify the Jobs-to-be-Done (JTBD) for a particular audience segment, you’ll need to ask the following questions:

  • What task does the user have to complete?
  • What solution do they currently use?
  • What problems do they encounter often?
  • How significant are these problems?  (Ex. Which problems does the user designate as “important?”)

In this way, a single question can be broken down into four components. 

Your primary tool at this stage is a research script or guide, which should contain a list of questions you want to ask interviewees. Therefore, choosing the right tools boils down to selecting appropriate questions for the interview guide.

At GoPractice, we try to structure these scripts into sections that reflect the goals of each interview. For goals like “identify jobs-to-be-done” and “identify key problems,” the script will have two main sections: one that contains questions about jobs, and another that contains questions about common issues. This helps both interviewers and interviewees maintain focus over the course of a session. 

Recruiting

The recruiting process consists of two tasks: screening respondents and recruiting them directly.

Screening is necessary to filter out non-target respondents for your research. Using a framework to disqualify interview candidates can help you avoid wasting your time (and respondents’ time). 

This process can be completed with the help of a qualification questionnaire (also known as a ‘screener’). A screener is essentially a list of questions whose answers will help determine the eligibility of respondents. When developing a list of screening questions, consider taking these characteristics into account:

  • Frequency: How often do users perform a particular task?
    • For example, a researcher for UberEats may be interested in respondents that order delivery once a week, as opposed to those who order delivery once a year.
  • Priority: How much money do users spend on solving particular problems?
    • For example, a researcher at Udemy may be interested in respondents who spend upwards of $1000 on education per year.
  • Demographics: What socio-demographic characteristics (gender, age, location, etc.) does the audience have?
    • For example, a researcher at Company X may be interested in recruiting respondents with at least two children.

It’s important to ask screening questions before in-depth interviews begin in order to filter out ineligible participants. To screen participants, you may want to use an in-platform chat, a Google form, or a recruiting announcement. When sending out announcements to large groups of people, be sure to specify in detail exactly who you’re looking for. 

Sometimes, finding interviewees is simple: you can request to survey your customer base directly. However, when you need to search for potential users, this process can become more difficult, especially for teams with modest budgets. However, there are different ways to tackle this problem.

The simplest option is to think about where your target audience lives: what social networks and community forums they use, where they go, and which thought leaders they pay attention to. By understanding your target audience, recruiting channels will appear naturally.

Many recruiting methods exist, but it’s difficult to ascertain which ones will be the most effective. To help get you started, here are several paid and free recruitment strategies:

Free*

  • Personal network: Post on social networks, and ask acquaintances to do the same.
  • Snowball method: At the end of an interview, ask who else they can recommend.
  • Current audience: Send out emails or messages to current users.
  • Social networks: search your audience as followers of thematic channels and groups of your competitors such as Reddit, Facebook, Instagram, TikTok.
  • Other social apps: try to find a particular person using Tinder, Twitch, iTalki, Tandem.
  • Communities: try to find communities about your specific topic at Facebook, web forums, private chats or google “your topic + communities”.
  • LinkedIn: search for B2B respondents here using filters.
  • Channels and websites with job listings: sometimes the easiest way to find a target respondent is to find his CV or a vacancy posted by a company.
  • Business chats on Facebook.
  • Competitors’ communities and social networks.

*These recruitment channels are conditionally free: While you may not have to pay to search for potential respondents, you may have to pay respondents directly for their time.

Paid

  • Respondent panels and services: these companies specialize in recruiting, so you just need to set the description of your target audience and pay them a fee (Ex. User intirviews, Respondent.io)
  • Bloggers: you can buy a paid placement about your research in any channel about your topic (Ex. on YouTube, Instagram, Facebook etc.)
  • Communities: the same concept as with bloggers. Communities sometimes allow paid placements and they are really welcoming if you promise to share the results of your research.
  • Self-hosted platform: make a short landing page or a survey form and drive traffic to it (Ex. usually Facebook or Instagram ads).
  • Professional recruiters: try to find recruiters or research agencies that can recruit respondents for money.

Conducting interviews

Studying the art of interviewing won’t make you into a professional interviewer. Willingness to practice, coupled with the ability to dive deep during a conversation, are both necessary. As you begin interviewing respondents regularly, here are a few things to remember:

Don’t blindly trust recommendations

It’s easy to find recommendations online about the “correct” way to conduct interviews, but you shouldn’t trust all of them blindly.

This guide offers a basic structure for conducting qualitative research. To take a deep dive into the subject of interviewing, we recommend familiarizing yourself with basic psychological principles behind decision-making, as explained in the book Thinking, Fast and Slow by Daniel Kahneman. 

As you begin your interviewing journey, remember to test theories in practice, experiment, and listen to common sense.

Start with onboarding

At the beginning of an interview, it’s important to outline what your respondent can expect from the session. Often, people may not remember who you are, what you want from them, and what will happen after the interview has concluded. To provide context, introduce yourself, explain your reason for reaching out, outline the topic of the interview, and explain how long the call will take. Be sure to record your conversation, but don’t forget to ask for the respondent’s permission beforehand.

Start with the basics

After establishing context, ask the person to tell you about themselves. For best results, engage in small talk, establish a friendly connection, and “warm up” the respondent. To establish healthy conversation, avoid burdening the person with complex questions during the first 10 minutes of interaction.

Dig deeper

A good sign that you’re not digging deep enough is if you’re simply reading one interview question after the other–without asking any follow-up questions. The goal of qualitative research is to understand why and how something works in the user’s mind, and the answer is almost never on the surface. To conduct research successfully, you need to understand a user’s needs and goals: to do so, try using the “five whys”, or other methods.

Follow what’s interesting

Remember, you’re not obligated to read interview questions one after another (or ask them all). Dive deeper during moments where you can sense the potential for interesting insights. If you get stuck, it’s okay to switch to another topic. Imagine the respondent as a boat, and you’re the helmsman: you may be floating downstream, but you can steer the boat as you see fit.

Confidence in interviewing comes with time. Once you start to memorize interview questions well and gain experience, you’ll learn to steer conversations however you please. Then, your interviews will feel more organic, and you won’t miss anything important by switching topics.

Sometimes, a respondent may spontaneously touch upon a topic from your list of questions, even though you’ve planned to discuss it in another section. It’s important not to lose track and pick up the topic later, within its section–by interrupting your interviewee, they might lose track of their answers, causing you to miss out on important details.

Don’t lead to an answer

Avoid giving hints towards a possible answer or sharing your hypothesis with respondents. If you want to know whether or not users need filters when searching, don’t ask about filters directly. Instead, ask users to describe how they typically use the search functionality. If filters aren’t mentioned, they’re probably not that important. 

Don’t let the user invent solutions

Interviewees are often eager to help you build a great product. In some cases, they might suggest solutions to particular problems. In these situations, thank the respondent for their opinion, but don’t focus on implementing their recommendations directly. Instead, focus on your respondent’s experience and their problems. The user doesn’t know what solution is best for your product–but after your research is complete, you should!

Remember, people lie

This is a very common problem. When being interviewed, many users don’t intend to lie–they just want to appear more impressive than they actually are.

To become a better interviewer, learn to recognize when people might be stretching the truth. When in doubt, ask respondents to ground their statements in experience. For example, if a respondent tells you they go to the gym often, ask how many times they’ve been in the last month. There’s a good chance they’ll answer “not once.”

During conversations with respondents, strive for specificity. If you’re stuck, remember to ask “when?” and “how much?” when talking about specific activities. If an interviewee is telling a story, ask them to elaborate and recall the last instance of a particular problem.

Like all of us, interviewees are also prone to memory lapses and perceptual distortions. As our memory compresses, we tend to remember events from three years ago as if they occurred last year. When looking for specific details about the past, help interviewees refresh their memories by asking them about the context in which a particular situation occurred. What else was happening at that moment? What other problems were they trying to solve? By establishing context with respondents, you have a better chance of gathering more accurate data.

Regularly revisit the goal

During the interview process, it’s easy to forget about your research goal. Try setting three alarms for any time of the day. When your alarm goes off, answer the question, “What’s the goal of what I’m doing?” Almost always, you’ll find you’re not doing what’s needed. Humans are wired for distraction. As an interviewer, it’s useful to remind yourself periodically to return to your goal.

To stay focused, conduct mini-retrospectives during the research process. As you reflect on your progress, pose questions to yourself and your team, such as “do current results help answer research questions? What needs to be adjusted?” As a result of these retrospectives, you may need to add new criteria to your screener, or add/remove questions from your interview script.

Keep in mind that no interview script can (or should) be perfect from the start. If your interview questions don’t change during your conversations, you’re likely not learning anything new.

Data analysis

At this stage, it’s time to analyze the information collected during fieldwork (interviews) in order to identify important findings, isolate patterns, and discover insights. For qualitative researchers, this stage usually consists of transcription (decoding) and analysis.

Before you can analyze interviews, you’ll need to transcribe them first. At present, services that transcribe speech into text are becoming better and more accessible for researchers.

In our experience, the most accurate transcription tools are powered by Whisper’s AI engine. There are various ways to use Whisper on your device. You also can look at Dovetail, TL;DR, or built-in tools from Zoom or YouTube. 

When analyzing qualitative data, try to match different entities together and compile them into a coherent view of your user’s world.

In fact, qualitative research can reveal a multitude of things about your audience, including their context, goals, jobs-to-be-done, user journeys, triggers, decisions, experiences of choice and use, insights, problems and needs, fears, barriers, emotions, evaluations, forecasts, and stories.

Work with what you need to learn within a scope of research. For example, I write brief notes on the most interesting parts of each interview. As a result, it’s easy to access highlights from a particular conversation, which are convenient to share with the team. Then, if something catches my eye or is interesting to the team, I can delve into the interview recording and extract more details.

When listening to the recording of an interview, you may notice interesting topics that you might have missed during the initial conversation. Unfortunately, this happens quite often. You’ll be surprised how much you didn’t hear during the interview itself because you were thinking about the next question, or were distracted by something else. As you read or listen to interview transcripts, make sure to highlight this previously ‘lost’ information.

In order to identify patterns, it can be helpful to annotate interview transcripts. You can document interview highlights however you like–I like to use separate pages in Notion for jobs-to-be-done and key user pains, but many text editors offer tagging solutions to label important text. In addition, specialized solutions like Dovetail offer various tools for marking up and analyzing text.

The final version of your analysis can differ depending on your research goals. Sometimes, a large diagram in Miro may be needed. In other situations, an expanded table containing questions and answers, or a simple slideshow, is useful. While sharing the results of your research is important, don’t get carried away with the documentation process. Remember the results you’ve promised to communicate and why they’re important. As a general rule, it’s better to be brief and to the point than to be elaborate and useless. 

As you prepare your documentation, remove any excess information, and focus on highlighting key points.

One more tip: don’t ignore contradictory facts. If one interview contradicts your findings, you’ll feel tempted to ignore it. Don’t do that. It’s important to maintain research objectivity, even if it muddles the coherence of your findings. 

Synthesis

So, we’ve conducted an analysis, made beautiful diagrams, and drawn conclusions. Is there anything else needed to complete our qualitative research project?

Here lies a serious misconception for inexperienced researchers: analysis does not always help solve business problems. Instead, analysis provides facts. During the synthesis stage, you need to turn facts into a plan of action. This is a separate task. 

When you’re conducting research for yourself, you may not experience this problem. In these cases, you have to complete research because it’s needed for your work.

If you’re doing research on behalf of someone else, try this exercise. Imagine that you need to implement changes based on that data you’ve obtained from your research. In this hypothetical scenario, gaps and weaknesses may come to light.

Now, let’s return to the “boring” part of qualitative research–the beginning, where we learned about setting goals and establishing a format to communicate results. At this stage, it’s important to ensure that your research-backed recommendations can be easily applied.

Ensuring the team uses research results

This is a crucial part of the research process, so don’t hesitate to allocate additional time and resources to it. Ultimately, the outcome of your research will depend on the quality of the recommendations derived from it. Formulating these recommendations should be your focus at this stage.

Involve your team during the synthesis phase–this includes anyone who helped conduct the research or is interested in the outcome. Ideally, you can also invite people with general experience in an area that’s relevant to your research. Once you’ve settled on a list of invitees, hold a meeting. The  format of this meeting should be fairly standard:

  1. Gather invitees
  2. Present key findings from your research
  3. Open any document and start brainstorming solutions. Allow everyone to contribute individually to foster independent thinking.
  4. Discuss these solutions together.
  5. Conclude by voting on solutions to select the best ideas.

The goal of this meeting is to connect findings from your research with experience from industry experts. Therefore, sharing insights derived from research should be your starting point. After presenting your findings, take a specific task, pain, or need and try to figure out how it can be resolved within the group. It’s often helpful to supplement conclusions you’ve drawn with quotes from respondents in order to provide participants with context.

If you continue to repeat this process, eventually, you’ll have a set of actionable solutions. Up until this point, you’ve only had observations.

The secret to finding synthesis as a researcher is to cultivate a deep understanding of your business. The better you understand the industry you’re operating in and the business you’re serving, the higher the quality of your ideas and recommendations will be. If you lack industry expertise as an early-career researcher, try to involve more experienced people in generating joint solutions. As you do so, strive to understand their logic, and hypothesize how your research and recommendations might impact key metrics.

The importance of establishing a research process

If you conduct qualitative research more than once a year, it’s advisable to consider the effectiveness of your research process. An efficient research system can simplify key tasks for you and your team, enabling you to accumulate a knowledge base, refer back to it, and reuse the results of past research with ease.

Firstly, make regular interactions with users habitual for your team. Sustained contact with users will help lower barriers to testing product hypotheses, reducing the number of untested ideas that go into development. Consider integrating research into your development process as a hypothesis check. This could take the form of a prototype test or at least one user interview per week/month (depending on the team’s size, pace, and available resources). This way, you’ll regularly receive new inputs from your users. It’s not necessary to research and test all hypotheses and ideas; but it’s important for any team to find its own balance regarding what’s sensible to test (and what’s not). Think about optimizing your research process as building a conveyor belt: an effective system will help you cultivate a healthy research culture within your team.

Secondly, describe your research process and share it with others. Consider creating checklists and distributing documents so that important processes aren’t siloed. This way, every team member can engage in effective research work.

Finally, create a convenient environment for storing and reusing your results: a research repository. This could take the form of a table that contains insights or common issues. A database like this can help provide new team members with important context, track chronic product problems, and make it easier to identify trends.

In conclusion, here are some practical tips for beginners when conducting qualitative research:

  • Allow plenty of time for research. There’s an illusion that conducting an interview is a quick and easy task. As this guide demonstrates, usually the opposite is true: interviews involve a lot of behind-the-scenes work. Therefore, for a full research cycle (involving 5-8 respondents), allocate about three weeks of your time.
  • Don’t forget about the synthesis stage. Many PMs might think that qualitative research is only about fieldwork. In reality, preparation, data analysis, and implementation of results are the most meaningful and labor-intensive parts of any research process.
  • Begin with the end in mind. Start any research project not by writing questions, but by understanding how your results will be applied in practice. If this is not done, conversations with users may end up being detached from business goals.
  • Learn basic psychology. Subjectivity, selective perception, and memory distortions occur at all stages of the research process. To avoid this, learn about cognitive biases and try to catch yourself before you fall for them during the process.
  • Consider interviewee incentives. If you pay respondents for their time, how might that affect your research process? Additionally, consider that compensation accelerates and simplifies the process of recruiting respondents for interviews. Also, remember that respondents who agree to talk for free and those willing to be interviewed for a fee can represent vastly different groups. Think about how this relates to your research task, and whether there’s a risk of getting biased results.
  • Conduct a cost-benefit analysis before recruiting. Recruitment is a resource-intensive process. Don’t be overly optimistic here–it might be cheaper for your company to pay recruiting fees instead of paying for your time to be spent recruiting. This logic applies to the entire research process–sometimes “free” solutions may turn out to be more expensive than they appear.

Learn more

→ Quantitative vs qualitative data: what is the difference and when should you use one instead of the other

Illustration by Anna Golde for GoPractice