Join 31,816 readers for design news, cool links, and updates from Jake & JZ every... once in a while.
This post was originally meant to be a chapter in the Sprint book. One of the hardest things about writing is cutting parts you like—I love this story, but it didn’t fit in the book. Now that Gimlet has become so successful (they were acquired by Spotify for $230 million), this window into their early days is kinda fun to see. Hope you enjoy it!
If you’ve ever listened to This American Life, the blockbuster radio show from Chicago Public Radio, you’ve probably heard Alex Blumberg’s voice. After years as a producer on This American Life, and after creating another hit show called Planet Money, he left the world of public radio and started a podcast company called Gimlet Media.
The timing seemed perfect. With the growing ubiquity of smartphones, and digital audio in cars, podcasts looked ready to explode. And Alex flat out knew how to build audiences. After all, both This American Life and Planet Money had millions of loyal listeners tuning in each week. With Gimlet, Alex planned to build a slate of new podcasts and find new ways to make them profitable.
There was a problem, however. Podcasts are fundamentally limited by their format. Like records and cassettes, podcasts are audio-only. There are no images or video, no connections to friends or social networks, and — perhaps most critical to Gimlet’s success — no built-in payment system.
One of the promising opportunities for Gimlet was selling premium content: extra interviews, extended shows, and the like. But with podcasts, it was like Gimlet was a band playing at a concert hall with no ticket window. If people came to the show, that was great, but there was no easy way for them to pay.
There was a seemingly obvious solution: Gimlet could make their own app. If they offered fans a dedicated Gimlet app — instead of relying on platforms like iTunes to distribute their shows — they could invent a new business model for podcasts. They could offer one-click purchases within the app, and even cross-promote related books, merchandise, or other shows.
A Gimlet app was an obvious solution, but it wasn’t obviously a good idea. Gimlet didn’t have a technology team — their people had titles like “producer” and “editor,” not “software engineer.” And Alex worried that the app project would be overwhelming for his fledgling company.
The question nagged at Alex: should Gimlet build an app? To really seize the opportunity, and make his company a massive success, did he need to tackle the podcast problem? And if he did… what would it look like?
“You sure this is the place?”
I was standing outside a locked door on a busy Brooklyn street with my partners John Zeratsky and Braden Kowitz. We were in town for a sprint with our friends at Gimlet Media. We run sprints with startups all the time, and we thought we could help Gimlet answer their questions about whether to build an app.
It was November, and well below freezing — the first day of a cold snap with biting winds. We’re California types, and the icy gusts had us huddled in the doorway with our wool hats pulled low.
The building was supposedly Gimlet’s headquarters, but by all appearances, it was actually an abandoned theater warehouse. Through grimy windows, we saw stacks of wooden backdrops crowding an unlit corridor. It looked like nobody had been inside for at least twenty years.
A few minutes later, we spilled out of the elevator and onto the 5th floor. The warm air smelled like a cross between an art classroom and an antique shop. Alex Blumberg was waiting to meet us. With bulky headphones over a wool hat, a brick-sized black gadget in one hand and a foot-long directional microphone in the other, he looked more like a Ghostbuster than the CEO of a startup. He’d be recording our entire sprint as material for the StartUp podcast. (Spoiler: you can listen to the finished episode here.)
“So, what do you think of our office?” Alex pointed the mic in John’s direction.
“Um.” John looked around. Protruding from the ceiling were fans, round lamps, factory lamps, box lamps, rusty pipes, and what looked like a dryer hose. Rows of mismatched desks and dining tables were lined with worn out wooden, leather, and fabric chairs. No two of anything matched.
Alex pointed the mic at me.
“Um,” I said.
He pointed the mic at Braden.
“It’s different,” Braden said. “I like it…” His eyes darted around. “I think.”
Gimlet’s co-founder, Matt Lieber carried over a cardboard box. “Let’s head back to the sprint room,” he said. “I have to warn you, there’s no heat in there.”
The three of us exchanged a glance. We stopped taking off our coats.
The sprint room was funky. In keeping with the general decor, one end of the room was piled with oddments — a broken chair, three board games, and a disconnected industrial sink.
The room had a battered wooden table long enough for a three-generation Thanksgiving dinner. Matt placed his box in the middle; inside were a dozen apples, trail mix, and bags of bulk almonds. He explained that he works for about an hour a month at a food cooperative in Brooklyn. “These apples are unbelievable. You can’t get them at a regular grocery store.”
There was no fancy video setup, just a ten-year-old monitor sitting on the end of the table. And Matt wasn’t kidding. It was cold — probably about 45 degrees. But the sprint room was perfect. It was big enough for all of us. One entire wall was covered with IdeaPaint — the space-age stuff that turns any surface into a whiteboard. And the apples and trail mix were delicious. We had everything we needed for a great sprint.
“OK,” said Alex. “Starting now, our schedules are clear for the week. So what happens next?”
“Are we really going to compete with iTunes?”
By Monday afternoon, the whiteboards were plastered with diagrams and notes. About a hundred questions were written on individual Post-its. There was a list with no fewer than 32 different ideas for things Gimlet could build — everything from a sleep timer to exclusive episodes. There was a map: a simple diagram on the whiteboard showing a customer’s progress from learning about a podcast all the way through becoming a regular listener. Alex and Matt had spread out everything they knew, they’d interviewed their employees, and they’d even brought in friends with expertise. Now it was time to make some decisions.
It was that map that was causing consternation. It was impossible to look at the diagram without seeing the 800-pound gorilla that dominated the podcast landscape: Apple’s iTunes software. iTunes was a great service for Gimlet — providing the bulk of their subscriptions — but it also perpetuated the same old distribution model.
“We’ve got to pick a spot for our prototype and test,” Matt reminded him. “But if you don’t want to reinvent distribution, we don’t have to.”
Alex scanned the diagram and nodded. “Yup,” he said. “Let’s focus here.”
He pointed to the far right of the diagram, far downstream from the shadow of iTunes. “When I look at the opportunity, it’s not about getting people in. People are getting in already. Our listener numbers are strong. If we can just do a better job at the end of the cycle, we’ll have a giant impact on the business.”
Alex’s decision wasn’t just some spur-of-the-moment thing. Earlier that afternoon, he’d brought in a secret weapon expert: Charles Duhigg, a friend of Matt’s, an investor in Gimlet — and, to the rest of the world, author of the bestselling book, The Power of Habit. A couple of hours ago, Alex and Matt had cross-examined Charles and gotten him to explain how the mechanics of habit formation made the best podcasts a part of their listeners lives.
Charles described great podcasts as a kind of “reward.” Many listeners, he said, save podcasts for a special treat — listening when they clean dishes, fold clothes, or run an annoying errand. With their podcasts, Gimlet has created an honest-to-goodness dopamine producer. Charles had suggested they use their power for good.
Now, remembering Charles’ advice, Alex pointed to a question written in thick marker on a sticky note: How might we create a stronger connection with listeners?
“It’s all about the existing listeners,” Alex said.
Matt sighed. “Well, that’s a relief.” After all the anxiety, it turned out there was an opportunity that did not require them to compete with Apple. Alex and Matt had discussed this question — should we become a technology company and make an app? — for months. Every time they talked about it, they felt stuck. Now, by spending a few hours mapping out the problem before jumping to solutions, they began to see a way forward. By focusing on existing users, they could do something that felt authentic to their mission. Not only that, but they wouldn’t have to try to take on iTunes and the rest of the podcast ecosystem.
“So that’s the plan,” Matt said. “We’ll prototype an app for existing Gimlet fans.”
The Gimlet team could recruit from their existing listeners for Friday’s test. They had a plan, they had a deadline, and it was time to get to work.
Matt and Alex followed the rest of the sprint process. On Tuesday, they sketched solutions, and on Wednesday, they chose their favorites. On Thursday, Alex recorded audio tracks for the prototype. The whole thing, soundtrack and all, was stitched into a series of Keynote slides so it looked like an iPhone app.
On Friday, the listeners came in. One by one, they listened to the end of a StartUp episode — which had been custom recorded by Alex just for the sprint. In the audio clip, Alex announced the new Gimlet Media app, and asked them to download it. The prototype even included a fake version of the App Store, so that the whole experience seemed real.
Check out the prototype in this video!
By the end of the day, a few things were clear. People hated an app that talked back. They found the rest of the app interesting, though nobody fell in love with it. But the biggest finding of all? Everyone loved getting access to special episodes and other content in the app — and they were willing to pay for it.
So did Gimlet really need an app to deliver that content? Alex and Matt started to wonder. The sprint allowed them to fast-forward to a possible future. They saw customers react to a realistic façade of a Gimlet app, and, just as important, Alex and Matt saw that realistic app for themselves.
In the end, they decided that an app was not the magic solution. The magic was in their connection with listeners, and in the content itself. Perhaps they would build an app one day, but they didn’t have to wait for that day in order to launch paid content. (In fact, a few months later, Gimlet introduced paid content with a simple message from Alex in the podcast — after all, it worked in the prototype.)
And maybe, Alex and Matt realized, they wouldn’t have to make one big decision to alter the future of the company. Maybe they could keep sprinting and prototyping their way to a plan. “We can fake it,” said Alex, “‘til we make it.”
Before the sprint, Gimlet was intimidated by the idea of becoming a tech company. On Monday, they started making progress and their stress level plummeted. They even found reason to be excited by this gnarly challenge. They remembered something they’d known all along: the importance of the connection they had with listeners. Their hypothetical app might be a way to strengthen that bond.
It’s all great in hindsight, but on Monday morning, everyone was a little terrified. As a matter of fact, we get antsy before every sprint. We think things like, “Is the process going to work this time?” and, “We don’t know what we’re doing!”
Those pre-sprint jitters are perfectly natural. And it’s exactly why every sprint starts with a full day of talking, planning, and mapping out the problem. Crazy, right? When you only have five days to solve a problem, spending a whole day not solving it might seem insane.
Actually, it’d be insane not to start this way. Later in the week, the sprint team generates solutions, makes decisions, and builds a prototype at an incredibly fast pace. It only works if you begin with the right foundation.
On Monday of their sprint, Gimlet carefully prepared for the week. The first thing they did was to gather information. They revisited the basics, like where listeners come from. They brought up old ideas (like each team member’s favorite listening app) alongside new information (like Charles Duhigg’s model of habit mechanics). This kind of information comes up all the time at Gimlet. It’s a small, close-knit team, and they’re constantly talking. But, as in most organizations, their time is fragmented, and those nuggets of information are easy to forget. That Monday, they got to talk about all of it — all at once, and at the right time to do something about it.
After they unpacked everything they knew, every brain in the room was loaded with the same information — something that almost never happens in normal work life. They were prepared to map the problem. They drew a customer story diagram showing how listeners discovered podcasts, why they built meaningful listening habits, and what brought them back again and again. As they drew that diagram and reflected on all they’d seen, they realized that what was most important — and what felt true to their mission, the reason they’d started this company in the first place — was building a closer relationship with listeners. They made the tough decision to ignore almost all of that journey and laser focus right at the end, where they had the opportunity to strengthen the bond with their most loyal audience.
Imagine a bizarro world where Gimlet jumped right into solutions. They likely would have had ideas that spanned the entire diagram. One person might have championed an app, while another wondered about doing a partnership with Netflix. Each person would have been limited by what he or she knew or remembered, rather than the shared insights of the group. And without the clearly established guideline of bringing loyal customers closer, it would have been impossible to know which ideas to try first.
You might notice that the bizarro world sounds a lot like business as usual for most teams. Jumping to solutions sounds super efficient, and as result, it’s awfully tempting. The sprint process — with its built-in full day of planning — is a great antidote to that kind of thinking. It gives you a recipe for working smarter with your team, finding better solutions, and having more fun.