What is Customer Discovery | An Introductory Guide, Part 1

What is Customer Discovery | An Introductory Guide

In the past 20 years there have been a series of new methodological and philosophical breakdowns of how to start a business. The term “customer discovery”, though it was coined in the early 2000s and ‘discovery’ tactics of this kind have existed for centuries, has more recently become a common turn of phrase in the tech and entrepreneurship world.

But what is customer discovery? How do you do it? And what do you need to know about conducting qualitative interviews before you get started?

This article is the first of 3 exploring different aspects of customer discovery, from what it is, to how it works, and even quick guides on how to do it. 

If you’d like to get weekly emails walking you through different aspects of customer discovery and similar topics, make sure to sign up to our email course at the bottom of the page.

[Click here join our email short course on How To Conduct Great Qualitative Interviews]

Or explore our blog and this series for more details:

In this article we’re going to talk about the history and importance of customer discovery as a concept, and then walk through the basics of what, why and most critically how to do customer discovery in the modern (and now mostly remote) world.

For those of you who know this topic very well, you may feel that I have blurred the lines between customer discovery, and a variety of qualitative validation methods, and I’d like to assure you that it was intentional. As where these techniques can be more or less rigorous, they share the same fundamental strategy.

[P.S. it goes without saying, but in case you weren’t aware, our tool is designed to help people like you run a customer discovery project. If you have no idea what I’m talking about, check out Yleos once you’re done reading the article!]

What Is Customer Discovery

While many of the published philosophies of the tech world are focused on raising money, company team and culture, or product design, the most controversial (and arguably impactful) concepts are those that focus on the importance of understanding your customer. 

I say controversial, because as idealists, most of us love ideas, and in business we are often taught that it’s our ideas that count, that get us investment or allow us to build cool things.

But therein lies the dilemma. You see studies on entrepreneurship, have shown repeatedly that the average success rate over the first 5 years of new tech businesses is less than 10%. Studies done by our parent company and others have started to show it may even be sub-2% success!

We also know that because of the increasingly competitive environment created by large and consolidated firms, there are fewer and fewer entrepreneurs every day. In fact, the global average of new businesses started per-year and per-capita has plummeted by over 30% in 30 years; and new the per capita share of businesses in US business activity alone has dropped by more than 40%!

In short, the tech hype of entrepreneurship captures the hearts of many, but rarely teaches them how to avoid its high risks and statistical likelihood of failure.

Among its many advantages, customer discovery can be viewed as a risk mitigator. It is a tactic that uses empathy to reduce the chances of failure.

But given that all companies are different, it is often difficult for startups or new product teams to apply concepts like customer discovery, or related validation strategies directly without previous experience. Customer discovery can in fact prove itself a risk. If it’s done very badly it can produce justification for unnecessary risk instead of mitigating it.

So how do you do customer discovery? Below let’s talk through the critical theory and tactics required for effectively de-risking the development of a new product or service.

How To Do Customer Discovery

Luckily for us, customer discovery is not exactly a “new” way of approaching business. 

There is written evidence as far back as traders on the silk road of merchants collecting qualitative inputs (talking to customers) to understand what they value most, and how to serve them better than their competitors. This is in itself a form of customer discovery, and a validation tactic that you can casually employ.

But the trick here is not simply the asking of questions. It’s the targeting and design of the questions you ask.

In the next article on specifically how to do customer discovery, we’ll walk through the steps and structures for getting the job done. First, there are a few critical concepts you should read below before starting that process. 

While customer discovery can be a meticulously crafted process, which we’ll get into below, it lives or dies by two simple precepts: 

1. who you ask


2. how you ask

Let’s take a quick look at both…

Who You Ask

This seems like a simple problem, but is perhaps the most often overlooked by first timers. 

Remembering that it’s called CUSTOMER discovery, you should be starting a conversion with people who are in fact a narrowed segment of potential or current customers.

If your focus is on an issue experienced by one person but paid for by another, in which case your user and your customer are two different people, you’ll need to make sure to talk to both. 

[If you’re not sure how to identify or segment customers, we recommend starting here…]

More importantly, you need to talk to potential customers who you don’t already know

Pre-existing relationships are the quickest way to get input, but often the quickest path to misdirection. Your parents, siblings, friends and confidants are all heavily biased by their pre-existing relationship with YOU, and will give you answers influenced by that relationship whether it be overtly supportive, negative, inquisitive or otherwise. 

As much as it’s relevant to build things the people around you can embrace and understand, customer discovery is about proving that you know who your customers are beyond your inner circle, and empathising with their needs.

You can get around quite a lot of inner-circle bias if you master the art of question design (more on this below), but there is a little-spoken about concept hidden in the need for you to go speak to new people, which is this:

If you don’t know how to find potential customers to speak to now, what’s to say you will be able to find them later, when you have a product or service to sell?

You see, customer discovery is part of an ongoing validation process for any product or service. It provides data to inform your decision making, from what a product or service should be, to who it’s for, how it should work, and how it will be marketed.

In this way you can look at customer discovery as your first test:

  • Can you identify people who are experiencing the problem you want to solve?
  • Do they need a better solution to that problem than they have now? 
  • Can you empathise with their needs enough to design an appropriate solution?

The first time you do this you are likely to fail the test. In fact, you may fail a number of times before the stars align and your conversations are validating something you can take action towards. But that’s a good thing.

In fact even now when our teams do customer discovery we are usually hoping our first round of interviews will show us we were wrong about something. 

By proving ourselves wrong it shows us we are effectively managing the bias of our assumptions instead of embracing them. Not to say we’re not right sometimes, but even then we often find details in people’s experiences that will change our approach.

In short, “failure” in the discovery process is a good thing! Proving yourself wrong is how you avoid spending months, years, and a lot of money building something that won’t work. But if you let go of your ideas, and keep digging, you’ll often find something brilliant that will.


There is a critical point here that needs to be made before you go any further.

If it’s your first time exploring the topic of customer discovery, or validation in this way, you may be making assumptions about the methods you use to “talk to customers”. 

In fact, we find most people without clear direction will default to using surveys as a way to collect customer data.

While surveys can be immensely valuable and have their place in the collection of customer data, this is NOT what we’re talking about when we discuss customer discovery.

The only genuinely useful way to start a customer discovery process is to have real conversations, one-on-one, with real people.

We call these conversations “interviews”, or in some cases “discovery interviews.” 

Interviews have more context when done in person, but are easier to conduct remotely.

Discovery interviews are best executed with a “script” or “discussion guide” that you’ve designed for the interviews based on your knowledge and assumptions around the topic of discussion. 

And your interviewees are best referenced as “participants”, as it is a more humanising term than “user” or “customer”.

Your script is also a way to make sure you collect similar and comparative data from your different participants, allowing you to look at patterns or isolated events as part of the overall focus of your discovery. 

You see, while a survey will tell you what happened by asking participants to reflect on past events, interviews allow you to explore why things happened by digging layers deeper into someone’s experiences, in their own words.

Of course you can try to do that in surveys, but you will always be limited by the construct that is the survey, and most people tend to write much less than they talk.

The value of interviews (or ‘primary qualitative data’ for you academic types) is that it gives you the room to identify opportunities that you may otherwise have never known existed.

By way of example, in 2014 when my colleagues and I were starting CareerFoundry, an online course platform for changing your profession, it wasn’t until we did interviews that we realised that people weren’t struggling with the decision to buy a vocational online course because of its price. 

In fact, they were struggling with the decision to buy a vocational course because they felt that the course was making a commitment to change their life by way of their career. Making a career change is a much bigger and scarier decision than paying a few hundred dollars for a course.

What was critically important about this discovery was that even though we planned to sell courses, a customer wasn’t buying a course from us, they were buying a change in career; potentially in lifestyle. 

But the customers we spoke to did not necessarily recognise or say that these bigger issues were what prevented them from buying courses. 

Participants would often say it’s expensive, or simply “I’m not ready”. But when you asked them to talk about what led to their decision to seek out a course, or how they felt about the decision to change their career – what catalysed it and how they planned to start that process – it became clear that the purchase of a course was a pivotal moment in that journey. 

Buying the course was the commitment to make a much bigger change.

Once we knew this, we designed our course waiting list, onboarding, and purchase funnel NOT to “sell the course”, but to help potential customers navigate the decision about their career. We didn’t change the concept entirely towards “selling a new career”. Instead, we used empathy in our interviews with participants to identify what people needed to reach an empowered decision of their own, and then helped them through it.

Once a customer had decided they were definitely ready, the course purchase was a no-brainer.

No survey could have given us this information because it wasn’t until we were conducting interviews and seeing people’s emotions around this topic that we understood that the course purchase and life-change were so deeply linked. 

This is also to say that discovery interviews are never a one-and-done. Like the products it aims to create, it is an iterative learning process, wherein your script is an ever-evolving MVP, just like the product it will help you build.

How You Ask

You’ll notice that I’ve worded this as “how you ask” and not “what you ask”. 

This is because the actual questions will range greatly depending on your topic of focus, meaning WHAT you ask will and should change. That is the part that requires you to plan out your discovery process. 

By contrast, the principles of HOW to ask good questions can be carried across any conversation you have, customer discovery or otherwise.

Remember that for better or worse, most people like to speak about themselves and their opinions; but when asked a direct question they will often give reactionary answers that counter how they would behave when unobserved.

You could oversimplify this section of the article by saying it is mostly about how to avoid asking direct questions in favour of those that lead to genuine discovery. 

If you’re not already familiar with the concept of open vs. closed or leading questions, I recommend you do a little reading on it.

To get you started, here is a quick break down of the three question types and some reference links for deeper diving:

Open vs. Closed Questions

Open vs Closed Questions In Customer Discovery

Let’s get one thing straight right up front: Open questions are good. 

They can be too open, or not open enough, but will generally make up the majority of an interview script.

They should become a point of study and analysis for anyone going through the customer discovery process as they give space for your interview participants to talk openly about something, ideally without feeling judged.

This doesn’t mean that closed questions are “bad”, but they have a distinctly different purpose, and much less value in the discovery process.

Let’s look at it in the context of “what” and “why”. 

A closed question, which is characterised by a question which produces a “yes” or “no” (binary) answer, can tell you “what” happened. Surveys are often constructed with different types of closed questions for easy quantitative analysis (ie. 80% said this, 20% said that = what happened).

An open question, characterised by a question that produces a non-binary answer, can tell you why something is, happened, or experienced in the way it was. Open questions can be designed to encourage reflection (why did it happen?), or projection (why would that happen?), but principally allow the participant to explain how they personally experience something.

You can always test to see if a question is open or closed by asking yourself, or people around you to see how they respond. 

If you/they give you binary or reactive answers, you are likely asking closed or leading (see below) questions. If you/they give explorative answers or talk about their experience in a literally “open” way, you have likely asked an open question.

The most important distinction here is that open questions allow you to empathise with someone’s experiences. This extends far beyond customer discovery and into real life as well, where numerous social studies have shown that relationships between friends and partners who ask open questions are both happier and longer lasting.

More reading on open vs closed questions can be found here!

Leading Questions

Most often referenced in court-room TV and movies, a leading question is a dangerous and misleading tool that anyone doing discovery or any qualitative research needs to avoid. 

In its literal meaning, a leading question is one that leads the respondent to an answer, to either give you the answer you want or present you with an argument to it.

Apart from the ethical issues this presents in court, in customer discovery this is the quickest way to validate your ideas while completely missing out on the truth. Or in the worst case scenario, starting an argument with your participant based on their beliefs. 

An example of a blatantly leading question type is the “Don’t you think…”?

  • Don’t you think this article is great?


  • Don’t you think chairs should be more comfortable?

They can also be directive by making a statement, followed by seeking affirmation:

  • You were in Amsterdam last week, weren’t you?


  • You didn’t see the stop sign, did you?

These questions tell the participant what you want them to say, rather than asking them what they actually did or felt.

The largest issue is that these types of questions have become common in our daily interactions. We often seek a binary response (often affirmative) to our existing perception of something instead of asking a question that allows an individual to explain how they view things.

As you can see, leading questions are generally “closed” in nature (see above section on closed questions), but they can also be more complex and cruel in the right context. We’ll leave those examples to the courtroom. 

In customer discovery, leading questions can be used to transition between conversation topics, but should be used with extreme care as the answers will often be what you want them to be instead of the truth.

You can read more about leading question types here, which gives poignant examples of how leading questions are used to make good but misleading media.

The Missing Layer In Customer Discovery: Problem Discovery

So far in this article we’ve looked at the lego blocks required for good customer discovery. In the next article in this series we’ll break down how to actually go about doing customer discovery for a product or service.

But to finish up this article I’d like to talk about a difficult yet critical concept that is so often missed in the discovery process.

In fact, I would argue that most ideological resistance on this topic from startups and product teams, or claims that “it didn’t work” or “didn’t help”, are either because of an addiction to one’s own ideas, or more likely, this missing layer:

Problem Discovery…

Problem Discovery Is A Critical Part Of Customer Discovery

Now if you are someone who does a lot of qualitative research already, you might consider problem discovery a separate stage of research that precedes customer discovery, but as an entrepreneur or product manager it is, and should be, an integral part of this process.

Problem discovery is effectively the assumption that the problem you want to solve, may or may not actually exist. But whether or not it exists (and here is the important bit!), it may not be what customers perceive as their problem.

I cannot emphasise that last statement enough…

Problem discovery is about distinguishing between the problem people have, and the problem people perceive.

But why is that important? If they have a problem, I can see it’s a problem, everyone’s going through it, no one has a solution, what’s the difference?

Well the difference is often quite simple: whether or not they have the problem, what you really need to know is: Are they actively seeking a solution?

You see it’s not even about wanting a solution, or needing one, it’s whether they are actively looking to solve a problem they have or not.

As as a product or service this is critical as if your intended customer is not actively seeking a solution, they will rarely if ever be willing to pay for one.

As smart people with ideas, we’ll often see problems that are ubiquitous and consistent, they are often problems that we ourselves experience in life or as a professional. But often with one problem comes many, and what I perceive as “painful” enough to find a solution for, may in fact not be the problem you think is most important.

In the next article on How To Do Customer Discovery (coming soon!), we’ll look at how to ensure your interview script explores the problem as well as the customer’s values, needs and expectations.

But before you go, here are some examples of customer discovery in action:

Customer Discovery Examples

Let’s go through some examples we’ve seen over the years. We’ve anonymised these case studies to protect the companies and teams they represent, but these are real examples from companies we’ve worked with, advised, or been part of through this process.

Keep an eye out not just for the “discovery” in these examples, but also for how the problem-to-be-solved changes; particularly the separation between root problem and perceived problem.

Better Breakfast Discovery Project

Working mums struggle to eat a good breakfast. Between getting ready for work, and kids ready for school, it’s often a missed opportunity. This discovery project was on a particular segment of working mums, based on the type of work they do and lifestyle they have.

We assumed that missing breakfast was a time issue, a problem of convenience more than anything to do with the food. The assumption here is that if you create a highly convenient breakfast option, perhaps in the form of nutrition bars or pre-made smoothies, it will solve their problem. 

But with so many options already on the market, why are mums still skipping breakfast?

We set up a script, and went out to interview working mums, but focused on asking questions around their morning journey, looking for a root cause rather than just how they manage time. We very quickly discovered that while yes, convenience is still at the core of the issue, mums would prefer not to eat at all if they cannot eat something that they enjoy eating.

So we dug more into how working mums define “enjoy” in relation to food (given the constraints of the morning), and discovered that there are three important factors in their choice to eat, apart from convenience. These were: taste, variety, and healthiness. 

Working mums of this particular segment would rather not eat than eat something that tastes boring. They would rather not eat than eat the same thing every morning. They would rather not eat than eat something unhealthy.

So while convenience is a problem with many potential solutions already in the market, the segment of mums we interviewed need healthy, and enjoyable-to-eat variety for their on-the-go morning meals.

While this was a fantastic discovery, we needed to dig deeper. Not only did we take a long hard look at the competition out there; we expanded the discovery script to include asking about where and how these mums would expect to find such comprehensive meal options. How would the ideal quick-breakfast work? What kind of variety would they like? What are the healthy factors they look for? And so forth…

Our initial script also gave us important things to explore around their morning routines and how food for the kids was made. We then looked for opportunities to seamlessly merge food prep for their kids with the preparation of their own breakfast.

After 10-15 interviews, we analysed the data and found these aforementioned desires as patterns throughout. We were onto something! So in an effort to create an MVP (minimal viable product), we cooked up a series of samples to test, and with many tweaks over the following months designed a breakfast product series for working mums.

Would you have had the same assumptions as we did? Would you have focused on the same things? For us the critical discovery was the “why”, followed by the “where” and “how” they would like to solve it. The outcome was a solution that, with time and testing, meant working mums ate breakfast, and a business was born. 

Alternatively, we could have spent thousands of hours prototyping the hippest, sexiest, most convenient breakfast bar to hit the market only to end up with weak sales and the occasional unfulfilled customer. But instead, after 10-15 interviews, we got to the root problem, and helped launch a brand that immediately captured its core market.

Good Grades Discovery Project

A pair of university student founders wanted to explore the topic of helping kids get better grades. To do discovery in this they first had to map out the stakeholders. They identified that you have parents who are your customers, kids who are your users, and potentially third and fourth stakeholders such as teachers and school administrators.

The team believed that the root cause of the customer problem was the need to help kids manage their homework better. That is, tools or services for helping kids get homework done.

They interviewed the parents first which partially validated their assumptions. Parents see bad grades and teachers tell them homework is an important part of fixing it. Parents then feel there is a need for heightened accountability, or outside help for their child. Parents of this segment most often referred to outside help in the form of tutoring.

But in speaking to kids the founders saw a very different story. It’s not that kids didn’t want to do homework, or that it was difficult; the kids didn’t care about homework at all. Digging in further, the founders discovered that the social pressures of school were so distracting that it entirely disincentivised the work to get good grades, both in the classroom and at home.

Now the founders had a dilemma. The root ‘cause’ of the problem for the user was a lack of social incentives associated with how they communicate with peers. The perceived problem of the customer is homework based. If you give the parents what they want the kids won’t use it, and if you give the kids what they want the parents won’t find or pay for it.

What do you do?

By digging further, the founders hoped to unearth an opportunity to solve the parents’ perceived problem, while empowering or enabling a solution for the root problem.

How might we incentivise home/class work as part of a child’s social interests? The founders landed on a variety of gamification methods, wherein rewards could then be tangibly applied to other aspects of the children’s lives.

But what about proof of work done? Parents in the interviews expressed a need to “see progress”, and feel like they have both input and control, thus the assumption a tutor would be most helpful. 

But could the circle be closed by the parent being part of the system? If children are incentivised to get homework done, but can’t unlock incentives unless a parent is shown and confirms it’s done, are both sides able to solve their respective problems?

There was more learning to do, so the founders prepared for a new round of interviews, but this time armed with a series of incentives and simple prototypes to test. 

One prototype showed particular promise and became the basis of the future app, where parent-validated proof of work then provided social incentives to kids. The beauty of the prototype was that all it took was a WhatsApp group where the founders manually played the role of the platform and rewards system. 

10s of thousands of developer dollars could have been wasted developing the wrong solution, and instead we found a model so simple and clean we could test it in a WhatsApp group.

What else would you be looking for in this complex dynamic? What other no-code prototypes could you test?

What’s Next for Customer Discovery?

I hope you’ve learned something from this article. In case you missed something, here is a quick nav back to different sections of the article:

In the next article we’ll talk more about How To Do Customer Discovery (coming soon!), with steps, question examples and more.

But if you’re new to qualitative research (ie, conducting interviews), I highly recommend signing up to our email short course on How To Conduct Great Qualitative Interviews. You can sign up directly below. It’s an excellent introduction to the topic, pointing to a wealth of resources for a deeper dive.

And don’t forget, if you need a tool to help you along the way check out our app, Yleos. We built Yleos specifically to help ourselves do customer discovery faster, better, and more easily, so it can help you too!

Thanks for reading, and see you next time!

Emil and Team

Want to learn more about how to conduct great qualitative interviews?

Whether you’re new to qual or looking for a new perspective on techniques and tactics, we’ve got the stuff!

“Built from a combination of unique content and interviews with user researchers from around the world; the Yleos Newsletter is a unique perspective on the world of qualitative research.”

Jerry C., Research Professional