Extrapolating from a Data Set of One or Fewer
How product leaders can build customer empathy without over pivoting on their own needs
We used to joke about product teams that “extrapolate from a data set of one or fewer.” The idea was that one of the risks every PM faces is feeling like their experiences are the only ones that matter.
I would like to take this opportunity to remind you that for most products, you are likely not the target audience. This statement may seem like common sense, but it is all too easy to forget.
I remember a presentation that a Facebook team gave about a product they wanted to build to help people meet up for brunch on Saturday mornings. This was maybe the third or even fourth pitch we heard of a similar type. This sticks with me, because it illustrates how limiting it can be to make assumptions based on our own experiences. Most people do not spontaneously decide to have brunch with their friends. Most people do not live in a city with lots of brunch places. Most people can't afford to drop $50 for brunch on a whim. Most people have kids, errands to run, or other obligations, and can’t easily do last-minute meetups. It became a bit of a running joke: “Every year or so, some new bright-eyed and bushy-tailed Product Leader says, ‘Let's go build a product to help people meet up and have brunch.’”
Discovering our blindspots
As PMs, we can be biased. Our habits, tastes, goals, and individual stories all contribute to the ways we see things. While this can be useful, it can also inhibit our ability to build effectively. I remember when I started working on Facebook Marketplace with a small group of people at the company. We loved the product, but so many people would ask us why anyone would want to buy used stuff with it. I remember sharing the story of how I bought a used refrigerator on Marketplace, and one of the other executives said to me, “We don't pay you enough to buy a new refrigerator?” Okay, sure. Maybe if you were a tech worker making six figures and living in a big city, you wouldn't need to buy used stuff. But the rest of the world was not like us, and that was something I understood, because I used the product myself.
However, as it turned out, I wasn't immune to blind spots either. Firstly, I never expected the Autos and Rentals category to take off, but somehow it did. We sold our old minivan on Marketplace so quickly over one weekend that my husband actually asked, “How can I get people to stop contacting me? I'm getting too many messages to sort through them all.” We once helped my in-laws rent out part of their place, and again, there was huge demand for a modest space to live.
Secondly, in many countries, people didn't use Facebook Marketplace like Craigslist. Rather than buying used goods from one another, they treated it like a literal marketplace, where they could pick up food and discover new businesses. That possibility had never occurred to me, even though I was on Marketplace myself. This illustrates how, no matter how familiar we are with our products, we are limited by our own backgrounds, locations, and experiences.
Go where your customers guide you
In the early days of PayPal, eBay sellers started adopting PayPal to use for transactions. It served two purposes: 1) it was a free way to accept credit cards for transactions, and 2) they made $5 for each buyer they got to sign up for the service. However, the leaders of PayPal debated whether to block transactions from eBay, not quite understanding what sellers were trying to do. Eventually, it became the best customer acquisition tool and PayPal’s largest business for many years.
Often, your customers will show you what you are missing. Take the time to listen to them and discover what they are trying to tell you.
There are three ways to see what users are doing with your product:
Ask them. Use focus groups, surveys, and follow-me-homes to understand how users interact with your product in the real world.
Look at the data. Examine what users are doing using information like searches, click streams, and activities.
Look at trends outside the company. Research sectors that are taking off. Investigate how competing products are addressing the same customers.
I remember desperately trying to figure out how to get my mom to the doctor's office in time when her appointment was moved and I could no longer take her. I ended up calling an Uber for her. It was one of those moments when I desperately wished they had a product specifically for medical visits. I am sure I was not alone in that moment—so many people live alone and need help getting to and from medical appointments. (Eventually, Uber built a feature for taking people to the doctor.)
It may sound simple, but looking at how people are actually using your product is one of the best ways to figure out how to make it better.
Understand the limitations of dogfooding while using it to address your blind spots
I've previously written about dogfooding: the practice of using your products to understand their limitations. While dogfooding is not a fix-all solution, it's still worth doing. Many of the products we build are not for people just like us, but for people with very different needs and wants. Understanding those needs and wants is more important than solving our own problems.
This has been a journey for me at Ancestry. While I can build a family tree with the knowledge that I have, my family and my in-laws were immigrants, and thus have limited historical records available on our platform. That said, it's magical to talk with our customers who have traveled back in time to learn about their families' histories through the incredible content that we have on our platform. Leading a company with such a diverse set of users means that I have to make a lot of decisions intellectually, not viscerally. I want to make sure we're serving our customers well, and that means building empathy for them.
In the end, I started building my brother-in-law's family tree. It was so incredible to see it come to life, person by person, through births, marriages, death records, Newspapers.com, and photos shared by other Ancestry members.
If you are in the core audience, make sure you're not an audience of one
One of the things that makes it difficult to differentiate between what you want and what customers want is that, when you're a customer, of a kind, but you have very specific needs. You need to learn how to generalize those needs to larger groups of addressable audiences, rather than just people like you. Here's how:
Identify the problem you're trying to solve. How big is it, really? How often do customers run into this issue?
Explain why you are doing things differently. How big are the pain points? Are you creating a vitamin or a painkiller?
Understand how generalizable the solution is. Is this addressing a very specific group of people, or can it be expanded to new areas?
Research others who have tried to solve this problem. Has anyone else solved it in a unique way? What differentiates your solution?
We all have blind spots—different preferences, backgrounds, and priorities—as both users and PMs. But if we aren't careful, we risk allowing those blind spots to impact our products. The best way to prevent this is to make sure we are building empathy for our customers and getting a broad range of perspectives. By taking a step back, observing our users, and using our own products, we can ensure that customers' unique needs are being met.
Perspectives is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.