One of the first surveys I created came from a business idea I was deeply passionate about. I had envisioned a religious social media platform that would bring together faith-based communities online. I invested time in crafting what I believed were thoughtful and comprehensive questions. I was convinced that I had considered everything. I was completely mistaken.
Just before sending out the survey, I added an open-ended question at the conclusion of the survey. It asked participants if they had any additional insights they wanted to share. I assumed there wouldn’t be much feedback. However, I was wrong again. That final question revealed that my supposedly perfect survey had some serious gaps.
Even twenty years later, I still remember one particular comment. I had included a question asking how people used Facebook, and I provided what I thought was a complete list of common uses. However, one respondent pointed out that the list didn’t allow them to mention their primary reason for using Facebook, which was to play games.
At the time, games like Farmville were extremely popular on Facebook. A significant number of users were logging in just to play those games. Since I didn’t use Facebook that way, I had completely missed this trend. It wasn’t just a gap in my answer choices. This was a clear example of what happens when behavior-focused questions are not asked. When you also fail to give people space to describe their actual routines, you risk missing something important.
That early survey experience taught me a valuable lesson. I had asked what I thought were the right questions, sent it to the right people, and felt confident in my data. Yet, I hadn’t truly focused on behavior or allowed people to tell their own stories. That misstep shaped how I view surveys today.
Surveys remain one of the most popular tools used by entrepreneurs to gather insights. However, many still fall into the same trap I did. They ask the wrong kinds of questions, reach the wrong people, and lead to conclusions that sound promising but lack real value.
The best surveys take a different path. They explore what people have actually done, what decisions they’ve made, and how they describe their experiences. This article outlines five essential principles for designing behavior-focused surveys that can help you uncover insights worth building on.
Principle 1: Reach Real Users, Not Convenient Ones
This principle comes first for a reason. It highlights a fundamental truth of research: good data starts with the right people. The best questions in the world won’t matter if they’re answered by the wrong people. If you send your survey to people who don’t fit your customer profile, your data will be bad from the start.
Many entrepreneurs, especially in the early stages, turn to those immediately around them for feedback. They ask friends, classmates, or social media followers. These people are easy to access, but they rarely reflect the real needs, habits, or decision-making patterns of your intended users. The feedback might seem useful at first, but it can lead you in the wrong direction.
Why It Matters
Hearing positive feedback from the wrong audience can give you a false sense of confidence and momentum. For example, if a product is intended for single parents but the survey respondents are mostly college students, the feedback may seem helpful, but it will push the product in the wrong direction. Without a strong match between your respondents and your real users, you risk building something irrelevant. It could end up being a solution your actual target market neither wants nor needs.
How to Reach the Right People
Start by clearly defining who your ideal user is. Avoid vague descriptions and instead create a well-rounded profile that includes demographics, behaviors, challenges, and goals. This user profile should guide where you conduct outreach and how you frame your questions. Once you know who you are targeting, find out where these individuals naturally spend their time. Look into industry-specific forums, niche subreddits, Slack groups, Facebook communities, or even in-person gatherings such as conferences and local meetups. Join these spaces and spend time participating before asking for feedback.
It’s also important to respect your participants’ time. Incentivizing responses with something of value, like a gift card, early product access, exclusive content, or even a thank-you note with summarized results, can help increase response rates and build positive rapport. Finally, add short screening questions at the start of your survey to ensure participants meet your target profile. These should focus on behaviors, experiences, or roles relevant to your product. If someone doesn’t fit, thank them and politely guide them away from completing the full survey.
Case Study: Canva’s Early Feedback Misfire
When Canva was in its early development stages, the team behind the platform was eager to validate its concept. They shared mockups and ideas with friends, colleagues, and other tech-savvy individuals in their networks. The feedback was overwhelmingly positive. People said they loved the idea of a simple, user-friendly design platform.
Encouraged by this feedback, Canva’s team moved forward quickly. But once the product launched, adoption among their actual target users was slower than expected. These users were non-designers with little to no design experience. The problem was clear in hindsight. The team had collected opinions from people who were not the intended users. These early testers were confident with technology, familiar with design tools, and not representative of the everyday professionals Canva was trying to serve.
This experience led Canva to change its research strategy. They began focusing on teachers, office managers, small business owners, and other non-design professionals. They asked about past behavior, such as how these users created presentations or flyers before Canva. This helped the team identify usability gaps and simplify the platform further. Once they focused on the right audience, user growth accelerated. Canva’s shift from convenient feedback to behavior-focused insights from real users was a turning point in their early growth.
Source: 10 Strategic Moves That Grew Canva to a $40 Billion Valuation
Checklist: Is This Person Your Real User?
- Are they currently facing the specific problem your product or service is meant to solve?
- Have they already taken steps to address or overcome that problem? What have they tried?
- Are they in a position to make a decision about using or purchasing a solution like yours? Would they see enough value to take action?
Additional Material
- How to Create an Ideal Customer Profile In 7 Steps (With a Free Template)
- Video: How to find and target your ideal customer
- Video: How to get people to take your survey
Principle 2: Test Behavior, Not Opinion
People often respond to surveys with what they think the survey creator wants to hear, especially if they feel unsure or want to be polite. They also tend to misjudge their future actions. Behavioral science and market research both show that people are poor at predicting what they will do when placed in hypothetical situations.
For example, research by George Loewenstein and Drazen Prelec (1993) demonstrated that people often misjudge their own future preferences and actions. Similarly, Daniel Kahneman and Amos Tversky’s foundational work on prospect theory highlights how decision-making under uncertainty deviates from expected rational behavior. Another relevant study by Koriat, Lichtenstein, and Fischhoff (1980) explored the overconfidence phenomenon, revealing how individuals often feel certain about decisions that later prove inaccurate.
These findings show that hypothetical responses are unreliable predictors of real-world behavior. As a result, when a survey asks what someone might do, the answer is often based on intention or imagination, not fact. To get reliable insights, frame your questions around real decisions and past experiences. Ask about actual behavior rather than what people believe they would do in theory.
Why It Matters
Opinions are easy to give. People often say what sounds good or what they think the survey creator wants to hear. Real behavior, however, involves effort. When someone takes an action such as making a purchase, clicking a button, or signing up for a service, they are signaling genuine interest. These actions come at a cost, whether it’s time, money, or energy. That cost makes the resulting insight more reliable than anything someone might say in a hypothetical situation.
Good vs. Bad Questions
The difference between good and bad survey questions often comes down to what you are asking your respondent to recall or imagine. Opinion-based questions typically invite guesses. On the other hand, behavior-based questions ask about actions that have already occurred. These types of questions are grounded in reality and produce more reliable and useful data.
Here is a simple comparison showing how a shift in wording can help you move from abstract opinions to grounded behaviors:
| Opinion-Based | Behavior-Based |
|---|---|
| “Would you use this?” | “When was the last time you used something like this?” |
| “Do you care about healthy eating?” | “What did you eat for breakfast this week?” |
| “Would you pay for this app?” | “What apps have you paid for in the past year?” |
Behavior-based questions make it easier to understand what people actually do, not just what they think they might do. They are tied to specific events or decisions, which is essential for identifying patterns and making informed product or service choices.
Tips for Testing Behavior in Surveys
Ask about specific things people have done. Frame questions around actual decisions or events so you’re gathering facts rather than intentions. For instance, instead of asking if someone would use a product like yours, ask when they last used something similar and what influenced their decision.
Avoid questions that start with “would you” or any phrasing that invites speculation. These questions encourage people to guess or say what they believe sounds good rather than what they’ve actually done.
Let respondents explain their answers with short text boxes when possible. Open-text responses give you a window into context, habits, and motivations that multiple choice alone can miss. Even one sentence of clarification can reveal more insight than a checkbox ever could.
Case Study: Duolingo’s Survey Refinement for Real Engagement Data
Duolingo, the popular language-learning app, once launched a survey to better understand how to keep users motivated to return daily. In their initial efforts, they asked broad opinion-based questions like “Do you find learning languages fun?” and “Would you recommend Duolingo to a friend?” The answers were positive, but app engagement remained inconsistent.
Realizing they were not getting actionable insights, the team restructured their approach. They began asking behavior-based questions like, “How many times did you open the app this week?” and “What is the first thing you do after completing a lesson?” These new questions gave them specific data on patterns, drop-off points, and user routines. The feedback also led to the creation of more targeted reminders and personalized features like “Streak Saver” to re-engage users who had recently missed a day.
Source: How Duolingo Gamified Monthly Active Users: Lessons in Habit Formation
Additional Material
- Stop Asking “Would You Use This?”
- 3 Reasons You Need to Stop Asking People What They Think About Your Startup Idea
- Video: Buffer’s Validation Story
- Video: The Art of Asking Questions
Principle 3: Gather Narratives, Not Just Numbers
Quantitative data can help you spot trends and measure growth. But when you’re trying to understand why people behave the way they do, numbers alone are not enough. They can tell you what is happening, but they rarely explain the underlying motivations.
This is where qualitative insights come into play. Context, emotion, and lived experience are essential if you want to create something that truly meets user needs. The best way to surface these insights is to ask open-ended questions that lead people to tell stories. These stories often reveal barriers, frustrations, workarounds, and unexpected behaviors that you wouldn’t uncover with a yes or no question or a satisfaction score.
Why It Matters
Stories help uncover the motivation behind a user’s choice. They show what happened before the decision, what influenced the action, and how the person felt during the experience. This type of detail gives you insight into patterns, unmet needs, emotional triggers, and barriers that data points alone cannot provide. Listening to stories gives you the depth needed to make sense of behavior and to design solutions that are more meaningful and effective.
How to Prompt Storytelling in Surveys
Use questions that invite specific stories from real experiences. These should focus on actual decisions or moments of frustration, not general impressions or opinions. Here are a few examples to get started:
- Tell me about the last time you tried to solve this problem.
- What part of that experience was the most frustrating or unexpected?
- How did you decide which product or solution to go with?
- What were you doing just before you made that choice?
- What happened after?
By focusing your questions on one clear event, you help respondents provide detailed answers that reveal motivations, behaviors, and emotional responses. Avoid broad or vague prompts such as “What do you think about…” or “What would you do if…” because they encourage speculation instead of storytelling.
Case Study: HelloFresh Uses Open-Ended Surveys to Discover Pain Points
HelloFresh, the meal kit delivery service, wanted to improve customer retention and better understand what caused subscribers to cancel. Rather than relying solely on multiple-choice questions like “Why did you cancel your subscription?” they added open-ended narrative prompts such as, “Can you tell us about your most recent experience with HelloFresh?”
These prompts led to detailed stories from users who described specific issues, such as missing ingredients, unclear cooking instructions, or delivery timing challenges. While structured survey data gave them general trends, the stories gave them clarity about which experiences truly frustrated customers.
By analyzing these written responses, HelloFresh was able to prioritize changes to packaging, delivery timing, and even recipe card design. This approach helped them address the real obstacles customers faced, leading to stronger satisfaction and better retention.
Source: How HelloFresh Turned Product Feedback into New Revenue Streams
Additional Material
- How to Turn Narrative Survey Responses into Data You Can Use
- Your quick guide to open-ended questions in surveys
- Video: The Art of Persuasive Storytelling
Principle 4: Ground Your Questions in When and What Actually Happened
Vague survey questions often lead to incomplete or misleading answers because they rely on general memory. People struggle to recall habits or decisions when the prompt is too broad or lacks context.
To gain better responses, anchor your questions to specific timeframes or meaningful events. For example, instead of asking, “How often do you eat out?” ask, “How many times did you eat at a restaurant in the past seven days?” This approach improves recall and produces more accurate, detailed responses that are grounded in real behavior.
Why It Matters
People often struggle to recall general habits or vague routines. The more abstract the question, the more likely the response will be incomplete or influenced by guesswork.
However, when you ask about a specific period, such as the last seven days, or focus on a particular event or experience, it becomes much easier for respondents to remember and describe their actions accurately. This level of specificity improves the quality of your data and provides insights that are grounded in real behavior, not assumptions.
Tourangeau, Rips, and Rasinski (2000), in their research, The Psychology of Survey Response, found that people provide more accurate answers when surveys focus on a specific time period or event. It’s easier to remember real behavior when questions focus on something clear and recent. For example, asking “How many times did you exercise last week?” works better than “Do you exercise often?” because it gives people a timeframe they can picture. This makes responses more accurate and useful.
Survey Techniques
- Use recent and specific timeframes. Ask about a defined window such as the last week or month. For example, instead of asking, “How often do you go out to eat?” ask, “How many times did you eat at a restaurant in the past seven days?” Questions like this improve memory recall and lead to more accurate answers.
- Refer to identifiable real-world experiences. Ground your questions in events that respondents can easily recall. You might ask, “What did you do the last time you planned a family meal?” or “What app did you last use to track your expenses?” Tying questions to specific events helps people give clearer and more meaningful responses.
- Avoid vague language. Steer clear of words like “usually,” “generally,” or “often.” These phrases are open to interpretation and can lead to inconsistent or vague responses. For example, instead of asking, “Do you usually exercise?” consider asking, “How many times did you exercise in the past seven days?” The more precise your wording, the more reliable your data will be.
- Ask about real actions instead of routines. Regular behaviors that happen automatically, such as checking your email or scrolling through social media, can be hard for people to describe clearly. When asked to summarize how often they do these things, they often guess or give vague answers instead of reporting specific actions. Questions like, “How often do you check your email?” often result in vague or exaggerated answers. A better alternative is to focus on specific past behavior. For example, asking, “What was the purpose of the last email you sent?” helps respondents recall a real instance, which makes the answer more reliable and useful.
- Be intentional about the timing and context of your questions. Be intentional about the timing and context of your questions. This principle builds on the first two points by encouraging you to consider not only when something happened or what the person did, but also where it happened, how it occurred, and what circumstances surrounded the behavior. For example, instead of asking, “Do you check your phone often?” you might ask, “What app did you use the last time you unlocked your phone today, and where were you when that happened?” Including both timing and situational context helps people recall their actual behavior more accurately, resulting in clearer and more actionable feedback.
Case Study: Netflix’s Viewing Data Improves Content Decisions
Netflix is known for using behavioral data, such as what users watch, when they watch, and how they watch, to drive major business decisions. Instead of relying on user surveys that ask “What kind of movies do you like?” or “Would you watch this show?”, Netflix focuses on what users have already done.
For example, when developing the series House of Cards, Netflix used behavioral data to identify three key insights:
- Viewers who watched the original British version of House of Cards also watched movies with Kevin Spacey or directed by David Fincher.
- Those viewers were more likely to finish shows quickly and binge multiple episodes.
- Political dramas performed well across several demographics.
These insights didn’t come from opinion-based surveys. They came from millions of real viewing behaviors. Netflix used this data to confidently greenlight House of Cards with Spacey and Fincher attached. The result was a major success that helped launch Netflix’s reputation as a content powerhouse.
By grounding their decisions in recent, specific actions, such as what people actually watched instead of what they said they liked, Netflix avoided the trap of misleading opinion data and created a new model for behavior-driven product and content development.
Source: Making Smarter Business Decisions with Big Data: A Netflix Case Study
Additional Material
Principle 5: Replace Yes or No Questions with Decision-Based Choices
Yes or no questions are quick and simple, but they often miss the deeper context behind a choice. People may answer yes because they want to be polite or because it feels like the “right” answer, not because they would actually take action.
Instead, use questions that mirror real-world decisions. Ask what someone did the last time they faced a similar situation, or present options that involve tradeoffs, like time, effort, or money. These types of questions are more likely to reveal what people would actually do in real life.
Why It Matters
Simple yes or no questions overlook the deeper reasons behind decisions. Real-world choices often involve constraints and tradeoffs. For example, someone might choose a slower shipping method to save money, or avoid a product because of a minor feature flaw. Surveys should reflect this kind of thinking by asking people to weigh different options.
Research in behavioral decision-making supports this approach. Payne, Bettman, and Johnson (1993) explain in their book The Adaptive Decision Maker that people adapt their decision strategies depending on the complexity of the task, available information, and perceived costs and benefits. This means your survey questions should reflect realistic tradeoffs rather than assume simple binary answers.
These strategies help uncover what people are more likely to do in practice, not just what they claim they prefer.
Case Study: Booking.com Uses Decision-Based Testing to Drive Conversions
Booking.com is known for constantly A/B testing small changes on its website to understand user decision-making. Instead of relying on simple opinion polls, the company uses real behavioral experiments to learn what people actually do when booking travel. For instance, they test different versions of urgency messages, pricing displays, and refund options to see how these influence booking decisions. Their goal is to simulate real-life choices—like whether to book now or wait—and use the results to optimize user experience and drive revenue.
In a 2016 presentation, Booking.com revealed they run over 1,000 experiments at any given time, and each one is based on actual user behavior, not speculation. By asking users to make real decisions in real contexts, they gain much clearer insights than what a yes/no survey could provide.
Source: Lessons from Booking.com experimentation culture
Additional Material
- Beyond the Checkbox: Redesigning Surveys for Reliable Behavioral Insights
- Behavioral Economics Techniques for Honest Survey Responses
- Video: The Pros and Cons of a Yes/No Question Type
Conclusion: Better Surveys Lead to Better Businesses
Back when I tested my faith-based social media idea, I thought I had built the perfect survey. I had spent hours refining it, convinced that I had asked all the right questions. Yet, it wasn’t until one honest comment pointed out a blind spot that I realized how much I had missed.
It’s a powerful moment when someone opens your eyes to something you didn’t see, especially when you’re building something that matters to you. As entrepreneurs, we don’t just want data. We want to understand the people we’re trying to serve. Their lives, their habits, their frustrations. When we ask the right questions, the kind that reveal behavior and real choices, we earn the insight that makes everything we’re doing worthwhile.
So when you sit down to write your next survey, don’t aim for perfection. Aim for connection. Ask questions that bring you closer to the truth of what people do, not what they say they might do. Surveys that explore real decisions help you create something useful. They remove assumptions and highlight what truly matters. If you want your idea to survive and grow, your research needs to reflect your users’ actual experience.
The strongest businesses don’t begin with guesses. They begin with understanding.