Two Lessons For Using Feedback

How to navigate the polarizing world of critique in product design — and make better decisions doing it

J.T. Trollman
Design at Meta
Published in
8 min readSep 15, 2015


Remember Sisyphus, that forlorn fellow from Greek mythology who was always rolling a boulder uphill? Give that gent a MacBook and some prototyping software, and ask him to build a product that will please 100% of the people who use it. Therein lies the truly impossible task.

Go this way! No, go that way!

Creative work is polarizing — from architecture, to writing, to digital product design. Building something pleasing to everyone is a task doomed to failure, because the diversity of opinions out there is both stupefying and bifurcated: People gush over your new painting. They hate that chair you designed. They love your blog article. They detest that app you just launched. Strong opinions are often the only ones voiced, and the loudest voices sometimes come from the smallest segments of your audience. Facing that clamor of mixed messages becomes a task even our mythological Greek friend would balk at.

We all owe it to ourselves to take a step back. Breathe. Yes, listening to outside critique is important; but we can’t lose the plot along the way. You’ll always run into that strong project manager, investor, user, or startup founder who will insist on a certain path. The trick is to learn how to listen.

How do we take criticism?

I’ve stumbled my way through a couple of different careers: although I’m a product designer at Facebook today, for five years I was doggedly charging down the path of journalism.

As a writer, my experience reacting to negative feedback often became an exercise in accepting that any response can be a good response. As long as I was confident in my facts and my story, my heart didn’t need to jump in my throat at the sight of an angry email: readership rate told the tale. I’d guess that tirade-prone people were ten times as likely to speak their mind, and move on to the next article.

Product design is a whole other beast, though. Every irritated tweet, comment, or (God forbid) negative article about what you’ve built is enough for you or your investors to catch night terrors — or at the very least, seriously rethink your strategy. Especially when you don’t have the right metrics to review, that fear bubbles up: Does everyone hate this? Is this the beginning of the end? Will I never get to meet Obama now? Having worked at both a five-person startup and a 10,000-employee tech giant, my gut reaction to critique is always the same: Sound the alarm, we need to change everything!

Let’s take that breath for a minute, though. There’s a way to eke truth out of all those one-star App Store reviews. In fact, there are two ways.

1. Know Your Core Value

The real lesson is not to ignore the problems people say they have with your product. What’s important is how you listen to those problems.

This is a treatise for temperance: through all of the slings and arrows of outrageous opinions, one key to level-headed decisions is knowing why your product exists in the first place.

When in doubt, think about Batman.

Spandex is always a good look for creative-types.

Beyond the obvious appeal of wearing a cape and fighting clowns, what really makes Batman appealing to creative-types is that all his actions stem from a core truth he holds. Batman strives to never kill or seek revenge: he seeks to avenge.

Everything Batman does is through that lens — and everything your product does should be through its lens of purpose, too. At Facebook, there have been countless times where I’ve been stymied about a project’s direction until I remembered to look at it through our lens of purpose: Facebook is about making the world more open and connected. That core statement has helped guide how we approach problems more times than we can count: when designer Laura Javier created the first version of the photo-sharing app Moments, for example, the goal she had was a uniquely “Facebook” take on a photos app: “make it easy to give photos to friends, and get photos back.

That statement perfectly encapsulates the “why” of a Facebook photos app. There are scores of other apps out there for viewing snapshots, all by different companies; Facebook could have easily made a me-too app that showed you all your albums. But making the world more open and connected is our lens, and so connecting people became a critical part of any Facebook photos experience.

That means just like you won’t see a gun in Batman’s tool belt of detective gear, you shouldn’t see a product by your team that betrays your company’s values. What’s your product’s reason for existing? What larger user problem are you trying to solve? If it’s not already apparent, relentlessly pursue the question until you find an answer. Use simple frameworks like Sakichi Toyoda’s Five Whys, or more involved methods; whatever it takes to cut to the heart of your rationale.

Learn your core value, and you have a guiding light through the noise of countless opinions. It may be that those loud, haranguing voices are right! But until you view them through your purpose-tinted glasses, it’s difficult to see fact from farce.

2. When In Doubt, Count

The second lesson for using feedback starts with knowing how people form opinions — and how they voice them.

It turns out my guess earlier — the one about how people with extreme opinions express them more readily — is a psychologically proven phenomenon. A study at Stanford found that not only do people with the most extreme opinions express them faster and louder, but they can also easily fool themselves into thinking they are the majority opinion. These hardcore haranguers feel like they are the champions for the masses, even when they aren’t: “Everyone hates your new sprocket thing! Change it! Bring back the old sprocket!”

The face of accepting feedback.

The thing is, those “champions for the masses” opinions are often wrong. Hundreds of loud voices can proclaim people want one thing, but the way people actually use that product can tell an entirely different story.

So count people. Count early, and count often. Data can tell you what opinions to hear; and conversely, opinions can tell you what data to read. Counter-intuitive? Let me explain.

Not every data-driven decision is straightforward: In this case, the opinions will likely tell me which data to listen to. Other times, it’s the inverse.

Data is useful both before and after you’ve made your new sprocket. Post-creation is the most obvious: Whether you’re a writer or a coder, checking out how many people are reading and sharing your article — or actually using your product in a particular way — is a reliable way of conveying the most common experiences. (In the made-up graph above, not everyone loves my haircut: just a couple of the people I care most about.)

Knowing the right data to read is a challenge in itself, of course. Understanding your product’s core value (our first lesson) can help guide you to the right metrics: if in the example above my haircut’s core value is making me, my mother, and my girlfriend happy, you can guess which bar graph I’ll listen to. But if all opinions were equal, I’d switch up my style: in this hypothetical survey, most people hate my sweet new bowl cut.

A more concrete example is Facebook’s support products team, which found that “user satisfaction after a support experience” became a core metric we wanted to improve. In most cases, we work hard to build products that raise the bar on that metric: the more people who are satisfied, the better we think we’re doing. But the majority doesn’t always rule: although few people ever voice concerns about suicide or self-harm, they’re such intense experiences that we spend months working on products to help connect folks with friends (and professionals) when they arise.

Counting before building can be informative, too: a year ago, I worked on a team that crunched data about how people were using Facebook’s Android app in countries like India. What we saw showed that when Wi-Fi wasn’t available, people switched their phone to “airplane mode” rather than eat up the 30MB/month of data they could afford. Data like that helped us decide to focus more on how our app behaved offline. On top of that, it also led us to develop and publish an entire network connection class system: by knowing a phone’s network speed, we can think about improvements like sending smaller images to your News Feed on a 2G network versus a 5G one, speeding up your performance dramatically. For the billions of people in developing nations — who make up many of the 44% of us still only connected to 2G data speeds — that type of sensitivity makes a huge difference.

“Using data effectively” doesn’t mean replacing intuition with Excel. And “Knowing your product’s core value” doesn’t mean being single-minded about your pursuits, doggedly ignoring divergent opinions. In the end, you do listen to the people using your product (or reading your article, or what-have-you), for the simple reason that at least some opinions will guide your path toward improvement.

But by being a bit smarter about how we listen to the people around us — be they users or coworkers, funders or founders — we can be that much smarter about the problems we choose to tackle. It may be the product you built was addressing the wrong problem entirely: for our pal Sisyphus, we might not give him a wheelbarrow to make his endless boulder-rolling easier. Chances are he wants something to eliminate the boulder in the first place.

It’s that level of decision-making that our core values and counting can inform. And those two ways of looking at feedback will guide your decisions better than any single opinion.

Cover photo credit: Kristina Alexanderson from Flickr



J.T. Trollman
Design at Meta

Product design manager and curious individual. I love AI, maps, problem-solving, transit, privacy & safety issues, and photography. SF native.