When You Should Not Listen to Your Customers

Henry Ford left us with some open questions.

Published in

Product Coalition

11 min readOct 10, 2023

“If somebody had been asked to do research about the year 1800 into how transportation could be improved, what would he have done? He might have looked for some improved diet that would give horses greater stamina. He might have sought some way of breeding faster horses. He might have wondered whether coaching inns could be better spaced along the highways or whether more pliable springs could be installed in stage coaches.” — D. E. Berlyne

As product managers, we have all enthusiastically dived into customer feedback, hoping to find insights that will take our product game to the next level.

Yet, time and time again, we hit a wall.

More often than not, the feedback is insufficient or contradictory. It fails to deliver valuable insights, game-changing ideas, or even just a decent overview of what users need.

Probably everyone has heard the alleged Henry Ford quote: “If I’d ask people what they’d want, they’d say faster horses”. While this statement illustrates the limitations of direct customer feedback, it does not explain how to work around those limitations.

To assist with the practicalities of product management, I propose segmenting user needs in a way that differentiates between the expressed needs that people volunteer, the unexpressed needs they do not communicate, and the unrecognised and potential needs they cannot accurately describe themselves.

This differentiation helps not only to dig deeper for valuable insights; it also explains why users words and actions do not always correlate, why direct feedback does not reveal market-creating insights, and also why, sometimes, there is no explicit demand to back up good product ideas.

In this article, I will delve into the real-life examples and common challenges with each of those layers, as well as how to overcome those challenges.

1. Expressed needs

The top of the iceberg, the most visible layer, includes needs that users can and will articulate.

These are the problems that customers openly discuss or complain about, the functionality that they request, and the improvements they suggest.

Expressed needs are rightfully the basis of most iterative product developments.

A common difficulty of working with expressed needs is that sometimes people confuse what they want with what they like or need, so they can also express preferences that are not true or optimal.

“While we may think we know what we want, we’re often wrong.” — A.Ansari

For instance, consumers often ask for more choices when shopping, but it is well known that having too many choices can be overwhelming and lead to shopping cart abandonment altogether.

I’ll talk more about how to handle the interference of subconscious needs later in the article.

The second difficulty is that people often frame their needs as solutions. And due to the Einstellung effect, the chances are that the solution they request will be similar to existing solutions they have experienced, rather than the optimal solution for their situation. If horses are all they know, then faster horses are what they will be asking for.

To overcome the Einstellung effect, product managers can start asking customers why they want the things they want. And not just once, but multiple times in a row, as explained in the 5 Whys methodology. This examination helps to reveal the problem behind the solution.

Taking the time to explore the customer problem may seem like a lot of work, but it can be beneficial for everyone involved if done right. For instance, when a startup called Kajako incorporated the 5 Whys into their feedback loops, they ended up solving more problems with less work.

2. Unexpressed needs

Then there are user needs that users are aware of but do not express. Not every user knows what is relevant or important to share, how to give feedback, or thinks that giving feedback is worth their time.

In fact, most users who struggle with a product might churn silently without voicing their concerns, leaving behind huge gaps in customer knowledge.

“You didn’t ask, so I didn’t tell.”

So the first challenge with unexpressed needs is just the discovery of them. It is impossible to solve problems you do not know about.

To find out what people have left unsaid, one has to be proactive about customer research. Regularly connect with past, existing, and potential customers to explore their lives beyond their product usage.

The best generative research is user-centric, not product-centric. Users’ lives do not revolve around our products, so all our interview questions should not revolve around our products either. Ask about context and workflows instead of features and bells.

Some unvoiced product-specific problems can be detected from analytics. For example, a high abandonment rate during onboarding is a good indication that something is not working.

However, it is important to understand that product analytics is not a silver bullet. Product data can tell you what people are doing, but not why they are doing it or how they are feeling about it.

Product analytics also does not show what people are doing outside of your product and what other solutions they are using to get their needs fully met. For instance, if a person is used to watching cartoons on Disney+, she will not be requesting or searching for them on Netflix, leaving no clues behind that this is also a need unless proactive research is done.

Both the quantitative and qualitative research methods have their strengths and flaws; that is why the combination of them usually works best.

3. Unrecognised needs

“The problem with listening to your customers is that they can’t tell you about a need that they don’t even know they have.” — C. Griffiths

Unrecognised needs are those that people themselves do not know they have but can be detected with observation.

Most people experience struggles that are so common that they do not identify these issues as problems to begin with.

“Pretty much everyone in the nineties dragged their CDs with them. Pretty much everyone had a beat-up leather case that lived in their car because it was too bulky for their bag. But pretty much nobody thought about it as a problem with a solution. Everyone just assumed it was part of life — if you wanted to listen to your music, you’d need to bring your CDs.” — Build

As history has demonstrated, just because something has not been classified as a problem does not mean it could not be an opportunity.

But to make things more complicated, there are other types of unrecognised needs to be aware of. There are also desires that are so deeply rooted in the subconscious that people are not aware of their degree of influence, such as the need for social validation or the preference for familiar experiences.

These subconscious preferences and biases can greatly affect behaviour and decision-making, sometimes leading individuals to make choices that do not align with their expressed goals, values, or needs.

Working with unrecognised needs has its challenges. When people are forced to explain things they do not understand themselves, they may do so very confidently but inaccurately, leading teams to make non-optimal product decisions as a result.

“The first online-dating services tried to find matches for clients based almost exclusively on what clients said they wanted. But pretty soon they realized that the kind of partner people said they were looking for didn’t match up with the kind of partner they were actually interested in.
Amarnath Thombre, Match.com’s president, discovered this by analyzing the discrepancy between the characteristics people said they wanted in a romantic partner (age, religion, hair color and the like) and the characteristics of the people whom they contacted on the site. When you watched their actual browsing habits — who they looked at and contacted — they went way outside of what they said they wanted.” — Time

It is not that people intentionally lie; it is just that they do not know or cannot accurately predict their true needs or desires.

That makes working with unrecognised needs as much an art as a science: you cannot rely only on what people say; you have to look at what they do.

This can be done with data analysis, customer shadowing, and other observational methods that do not rely on self-reporting. Observational methods involve watching and analysing people’s behaviours, actions, and choices as they unfold in reality, not in theory.

Observing is how you can spot both the “non-problem” opportunities and the contradictions between what people say and do. It also helps to uncover the problems and needs people know about but didn’t think to mention.

So whenever possible, observe first and ask questions later.

Another difficulty of working with unrecognised needs is that, because they are not clearly verbalised and are sometimes deduced from observations instead, the demand for a solution is not going to be as obvious as it would be with expressed needs.

But no one demanded a car in the 18th century either, while everyone could have observed that horses needed breaks to rest. “Not a problem” and “a boring observation”, some might say. “This is how it has always been.”

However, in hindsight, many 10x innovations—such as a car, the internet, or a computer—could have been a result of “a boring observation” and reacting to it with a creative solution.

Breeding homing pigeons that could cover a given space with ever increasing rapidity did not give us the laws of telegraphy, nor did breeding faster horses bring us the steam locomotive. — Edward J. v. K. Menge

Sometimes it is wise to challenge the status quo, and even a seemingly mundane observation could serve as a reason.

4. Potential needs

“The best way to predict the future is to invent it.“ — A. Kay

The deepest part of the iceberg represents people’s potential needs. Potential needs are those that do not exist yet, usually because the context for the need does not exist yet.

Before cars, no one needed seat belts; before the Walkman, no one needed to listen to music on the go; and before the internet, no one expected to have the option to buy things online.

New solutions allow us to do things that were not possible before, creating new behaviours, demands, needs, and preferences—preferences that do not exist until new conditions create them.

That is not only true for radical innovations, but every time something new gets built, only scale and risk differ.

The main difficulty of building new things is that people are notoriously bad at predicting the future, and there might be very little to learn from the past.

Today, we know that the iPhone became a huge success, but before its launch, it was not clear at all. In fact, Steve Jobs had a hard time convincing the rest of the organisation because the data did not support his vision.

“The marketing team fought Steve Jobs the hardest /…/ a lot of us rebelled. In 2005 the most popular “smart” phone by far was the BlackBerry — fondly known as the Crackberry. People were addicted. /…/ the BlackBerry die-hards would always tell you that the very best thing about their very favorite gadget was obvious. It was the keyboard.
/…/ So when Steve told the team his vision for Apple’s first phone — one giant touchscreen, no hardware keyboard — there was an almost audible gasp. People whispered in the hallways, “Are we really going to make a keyboardless phone?” — Build

It is safe to have data to make decisions for us. But the problem with predictive analytics is that it works only when nothing changes between the past and the future. When you build things, you change things, and just like you cannot predict where you are going based on what you see in the rearview mirror, past data cannot foresee the future either.

To compensate for the unusefulness of past data, new data can be created. This is where trial-and-error experimenting comes in. Experimenting means taking a chance on an idea without prior certainty that it is going to work.

Building and testing prototypes and running pilots can all improve the amount of information available and reduce the number of unknowns.

But even with extensive prototyping and testing, success can look like failure at first. Apple I failed, Apple II was lacklustre, it was not until the third iteration of the Macintosh that Apple finally found “it”: “the standard for convenient, user-friendly computing to which the rest of the industry ultimately had to conform.” — Innovators Dilemma.

Tony Fadell, who led the teams that created the iPod, iPhone, and Nest Thermostat, emphasises in his book “Build” that it is going to take at least three generations to get a new product right, given that the vision and timing on the market are right too.

According to his experience, V1 is built on vision, customer insights, and data—exactly in that order—and V2 can be built on the same things but in the opposite order.

So when building for the future, you need not only to take some risks and iterate; you also need patience and probably a lot of luck.

Exploring the known and the unknown needs

To make the long story short again:

The needs and preferences that customers express are usually easily accessible; one only needs to read customer support tickets, social media posts, and online reviews.

But people do not always say out loud everything they think. So a lot of information stays hidden. To uncover the needs that have not been volunteered, one has to be proactive about research by doing generative interviews and exploring analytics.

However, interview-based methods can only uncover needs that users can recognise and articulate. But sometimes people do not know what they want or like, so a lot of insights can be gathered from observations, science, and (big) data.

But even observational methods cannot uncover needs and preferences that have not emerged yet. When building something new, the most relevant information comes not from the past but from the trial-and-error of testing new ideas in real life.

When the goal is to create products that not just meet but exceed expectations, listening to people is a good start.

However, serving your customers right is not just about listening and reacting; it is about observing, predicting, and sometimes even challenging the status quo.

If past innovators had just been listening, we would still be riding horses and sending letters with pigeons.

To quote N&N:

Watch what people actually do.
Do not believe what people say they do.
Definitely don’t believe what people predict they may do in the future.

Be bold and get creative.

PS! One last suggestion. When researching needs, don’t focus solely on your existing customers; think about your past and future customers too. I write about the diminishing returns of customer obsession here.

References: