Sprint: How to Solve Big Problems and Test New Ideas in Just Five Days

by Jake Knapp

  Braden smiled and held his hands palm up. “I don’t know what the regions mean.” There was silence. We avoided looking at James. After all, Braden’s brave admission might be seen as heresy.

  “That’s okay,” said James. The floodgate opened. John and Jake didn’t know the difference between coffee regions, and neither did Daniel Burka. We drank coffee together constantly, but none of us had ever admitted to our lack of sophistication.

  Then Serah Giarusso, Blue Bottle’s customer service lead, snapped her fingers. “What do we do in the cafés?” she asked. After all, she went on, The Braden Situation must happen to baristas all the time: a customer comes in for coffee beans, but isn’t sure which kind to buy.

  James is a slow and thoughtful speaker. He paused for a moment before he answered. “The brew method is very important,” he said. “So we train the baristas to ask the customer a simple question: ‘How do you make coffee at home?’ ” James explained that, depending on whether the customer used a Chemex, or a French press, or a Mr. Coffee, or whatever, the baristas could recommend a bean to match.

  “ ‘How do you make coffee at home . . . ?’ ” Braden repeated. Everyone jotted notes. James had started the sprint by explaining his vision: that the online store should match the hospitality of the cafés. It felt as if we were onto something.

  The team spent the following day sketching ideas for the store. On Wednesday morning we had fifteen different solutions. That’s too many to test with customers, so the team voted on their favorites as a way to narrow it down. Then James, the decision-maker, made the final pick of three sketches to test.

  The first sketch showed a literal approach to making the website match the cafés: It looked like the inside of a Blue Bottle café, complete with wooden shelves. The second sketch included lots of text, to mirror the conversations baristas often have with customers. Finally, James chose a third sketch that organized coffee by brew method, bringing the “How do you make coffee at home?” question right onto the computer screen.

  James had chosen three competing ideas. So which one should we prototype and test? The idea of a website that looked like the café was the most appealing. Blue Bottle’s aesthetic is celebrated, and a matching website would look different from anything else in the market. We had to try that idea, and it wasn’t compatible with the other solutions. But those other solutions were also really intriguing. We couldn’t quite decide.

  So we decided to prototype all three. After all, we didn’t need a functioning website. To appear real in our test, each fake online store only required a few key screens. Working together with the Blue Bottle team, we used Keynote presentation software to make a series of slides that looked like three real websites. With a little ingenuity, and without any computer programming at all, we stitched those screens into a prototype that our test customers could use.

  On Friday, the team watched the customer interviews. One at a time, coffee drinkers shopped on several websites, with Blue Bottle’s three prototypes slipped in among the competitors. (To avoid tipping off the customers, we gave each prototype a fake name.)

  Patterns emerged. The store with wooden shelves, which everyone had such high hopes for? We thought the prototype was beautiful, but customers said it was “cheesy” and “not trustworthy.” But the other two prototypes fared far better. The “How do you make coffee at home?” design worked seamlessly. And the “lots of text” design shocked us: People actually read all those words, and the extra information brought Blue Bottle’s voice and expertise to life. As one customer said, “These guys know coffee.”

  James and the Blue Bottle team built confidence with their sprint. They were much closer to defining how their online store would work. What’s more, they’d done it in a way that felt true to their principles of hospitality. They believed the online store could be an authentic Blue Bottle experience.

  A few months later, Blue Bottle launched their new website, and their online sales growth doubled. The next year, they acquired a coffee subscription company. With a bigger team and new technology, they expanded the web store and began experimenting with new offerings. They knew it would take years to get the online store right—but in the sprint, they started on their path.

  The bigger the challenge, the better the sprint

  If you’re starting a project that will take months or years—like Blue Bottle and their new online store—a sprint makes an excellent kickoff. But sprints aren’t only for long-term projects. Here are three challenging situations where sprints can help:

  High Stakes

  Like Blue Bottle Coffee, you’re facing a big problem and the solution will require a lot of time and money. It’s as if you’re the captain of a ship. A sprint is your chance to check the navigation charts and steer in the right direction before going full steam ahead.

  Not Enough Time

  You’re up against a deadline, like Savioke rushing to get their robot ready for the hotel pilot. You need good solutions, fast. As the name suggests, a sprint is built for speed.

  Just Plain Stuck

  Some important projects are hard to start. Others lose momentum along the way. In these situations, a sprint can be a booster rocket: a fresh approach to problem solving that helps you escape gravity’s clutches.

  * * *

  When we talk to startups about sprints, we encourage them to go after their most important problem. Running a sprint requires a lot of energy and focus. Don’t go for the small win, or the nice-to-have project, because people won’t bring their best efforts. They probably won’t even clear their schedules in the first place.

  So how big is too big? Sure, sprints work great for websites and other software challenges. But what about really large, complicated problems?

  Not long ago, Jake visited his friend David Lowe, a vice president of a company called Graco that manufactures pumps and sprayers. Graco is not a small startup. They’re a multinational company who have been in business for more than ninety years.

  The company was developing a new kind of industrial pump—a machine used in assembly lines. David, the VP, wondered if a sprint might help lower the risk of the project. After all, it would take eighteen months and millions of dollars to design and manufacture the new pump. How could he be sure they were on the right track?

  Jake doesn’t know anything about industrial assembly lines, but out of curiosity, he joined a meeting with the engineering team. “I’ll be honest,” Jake said. “An industrial pump sounds too complicated to prototype and test in a week.”

  But the team wouldn’t give up so easily. If limited to just five days, they could prototype a brochure for the pump’s new features and try it in sales visits. That kind of test could answer questions about marketability.

  But what about the pump itself? The engineers had ideas for that, too. To test ease-of-use, they could 3D print new nozzles and attach them to existing pumps. To test installation, they could bring cables and hoses to nearby manufacturing plants and get reactions from assembly line workers. These tests wouldn’t be perfect. But they would answer big questions, before the pump even existed.

  Jake was wrong. The industrial pump wasn’t too complicated for a sprint. The team of engineers accepted the five-day constraint and used their domain expertise to think creatively. They sliced the challenge into important questions, and shortcuts started to appear.

  The lesson? No problem is too large for a sprint. Yes, this statement sounds absurd, but there are two big reasons why it’s true. First, the sprint forces your team to focus on the most pressing questions. Second, the sprint allows you to learn from just the surface of a finished product. Blue Bottle could use a slide show to prototype the surface of a website—before they built the software and inventory processes to make it really work. Graco could use a brochure to prototype the surface of a sales conversation—before they engineered and built the product they were selling.

  Solve the surface first

  The surface is important. It’s wher
e your product or service meets customers. Human beings are complex and fickle, so it’s impossible to predict how they’ll react to a brand-new solution. When our new ideas fail, it’s usually because we were overconfident about how well customers would understand and how much they would care.

  Get that surface right, and you can work backward to figure out the underlying systems or technology. Focusing on the surface allows you to move fast and answer big questions before you commit to execution, which is why any challenge, no matter how large, can benefit from a sprint.

  2

  Team

  Ocean’s Eleven, starring George Clooney and Brad Pitt,I is one of the all-time great caper movies. In the film, Danny Ocean, an ex-con played by Clooney, organizes a band of career criminals for a once-in-a-lifetime heist. Their target: a Las Vegas casino on the night of a big prizefight, when $150 million will be in the vault. The odds are against them, the clock is ticking, and it takes an intricate strategy and every special skill the team possesses to pull it off. There’s a pickpocket, an explosives guy, even an acrobat. It’s excellent cinema.

  A sprint resembles that perfectly orchestrated heist. You and your team put your talents, time, and energy to their best use, taking on an overwhelming challenge and using your wits (and a little trickery) to overcome every obstacle that crosses your path. To pull it off, you need the right team. You shouldn’t need a pickpocket, but you will need a leader and a set of diverse skills.

  To build the perfect sprint team, first you’re going to need a Danny Ocean: someone with authority to make decisions. That person is the Decider, a role so important we went ahead and capitalized it. The Decider is the official decision-maker for the project. At many startups we work with, it’s a founder or CEO. At bigger companies, it might be a VP, a product manager, or another team leader. These Deciders generally understand the problem in depth, and they often have strong opinions and criteria to help find the right solution.

  Take Blue Bottle Coffee’s sprint. Having CEO James Freeman in the room was critical. He was there to talk about Blue Bottle’s core values and share his vision for an online store that matched their standards of hospitality. He chose the sketches that best aligned with that vision. And he knew how the baristas were trained, a detail that unlocked a surprising solution.

  But it isn’t just expertise and vision that makes decision-makers so crucial. There’s another important reason you need them involved in your sprint, and we learned about it the hard way. See, one of the early sprints we tried was a big flop. To protect the innocent, let’s call the company SquidCo.II We’ll tell you who wasn’t innocent: Jake, John, and Braden. We screwed up.

  We’d carefully invited everyone from SquidCo’s team who worked on the project. Everyone, that is, except for one person: Sam, SquidCo’s chief product officer. Sam was going to be traveling, but the week worked for everybody else. So we helped SquidCo run a sprint. They made a prototype and tested it. The prototype did well with their customers, and the team was ready to start building.

  But when Sam returned, the project ended. What happened? The solution had tested well—but Sam didn’t think we had picked the right problem to solve in the first place. There were other, more important priorities for the team.

  The SquidCo sprint failure was our fault. We’d tried to guess what Sam would say, and we’d failed. The Decider should have been in the room.

  Get a Decider (or two)

  The Decider must be involved in the sprint. If you, dear reader, are the Decider, clear your schedule and get in the room. If you’re not, you must convince the Decider to join. You might feel nervous; after all, it’s a big time commitment for a new process. If your Decider is reluctant, try one or more of these arguments:

  Rapid Progress

  Emphasize the amount of progress you’ll make in your sprint: In just one week, you’ll have a realistic prototype. Some Deciders are not excited about customer tests (at least, until they see one firsthand), but almost everyone loves fast results.

  It’s an Experiment

  Consider your first sprint an experiment. When it’s over, the Decider can help evaluate how effective it was. We’ve found that many people who are hesitant to change the way they work are open to a onetime experiment.

  Explain the Tradeoffs

  Show the Decider a list of big meetings and work items you and your team will miss during the sprint week. Tell her which items you will skip and which you will postpone, and why.

  It’s About Focus

  Be honest about your motivations. If the quality of your work is suffering because your team’s regular work schedule is too scattered, say so. Tell the Decider that instead of doing an okay job on everything, you’ll do an excellent job on one thing.

  If the Decider agrees to the sprint but can’t spare a full week, invite her to join you at a few key points. On Monday, she can share her perspective on the problem. On Wednesday, she can help choose the right idea to test. And on Friday, she should stop by to see how customers react to the prototype.

  If she’s only going to make cameo appearances, your Decider needs to have an official delegate in the room. In many of our sprints with startups, the CEO appoints one or two people from the sprint team to act as Deciders when she’s not there. In one sprint, the CEO sent the design director an email that read, “I hereby grant you all decision-making authority for this project.” Absurd? Yes. Effective? Absolutely. This official power transfer added tremendous clarity, the kind of clarity we wish we’d had with SquidCo.

  And if your Decider doesn’t believe the sprint to be worthwhile? If she won’t even stop by for a cameo? Hold up! That’s a giant red flag. You might have the wrong project. Take your time, talk with the Decider, and figure out which big challenge would be better.III

  Once you’ve got a Decider (or two) committed to the sprint, it’s time to assemble your sprint team. These are the people who will be in the room with you, all day, every day during the sprint. On Monday, they’ll work with you to understand the problem and choose which part to focus on. Throughout the week, they’ll be the ones sketching solutions, critiquing ideas, building the prototype, and watching the customer interviews.

  Ocean’s Seven

  We’ve found the ideal size for a sprint to be seven people or fewer. With eight people, or nine, or more, the sprint moves more slowly, and you’ll have to work harder to keep everyone focused and productive. With seven or fewer, everything is easier. (Yes, yes—we know there were eleven people in Ocean’s Eleven. It was just a movie!)

  So who should you include? Of course you’ll want some of the folks who build the product or run the service—the engineers, designers, product managers, and so on. After all, they know how your company’s products and services work and they might already have ideas about the problem at hand.

  But you shouldn’t limit your sprint team to just those who normally work together. Sprints are most successful with a mix of people: the core people who work on execution along with a few extra experts with specialized knowledge.

  In Savioke’s sprint, we got great ideas from the people you’d expect, like the roboticists and the head of design. But one of the most important contributors turned out to be Izumi Yaskawa. Izumi wasn’t part of the team that built the robot, but as Savioke’s head of business development, she knew more than anyone about how hotels operated and what they wanted from the robot.

  For Blue Bottle Coffee, important insights came from the customer service manager and the CFO, people who normally wouldn’t have been involved in building the website. In other sprints, we’ve had winning solutions come from cardiologists, mathematicians, and farming consultants. The common traits they all shared? They had deep expertise and they were excited about the challenge. Those are people you want in your sprint.

  Recruit a team of seven (or fewer)

  Choosing whom to include isn’t always easy, so we’ve created a cheat sheet. You don’t have to include each and every role listed here. And
for some roles, you might choose two or three. Just remember that a mix is good.

  Decider

  Who makes decisions for your team? Perhaps it’s the CEO, or maybe it’s just the “CEO” of this particular project. If she can’t join for the whole time, make sure she makes a couple of appearances and delegates a Decider (or two) who can be in the room at all times.

  Examples: CEO, founder, product manager, head of design

  Finance expert

  Who can explain where the money comes from (and where it goes)?

  Examples: CEO, CFO, business development manager

  Marketing expert

  Who crafts your company’s messages?

  Examples: CMO, marketer, PR, community manager

  Customer expert

  Who regularly talks to your customers one-on-one?

  Examples: researcher, sales, customer support

  Tech/logistics expert

  Who best understands what your company can build and deliver?

  Examples: CTO, engineer

  Design expert

  Who designs the products your company makes?

  Examples: designer, product manager

  The word “team” is pretty cheap, but in a sprint, a team is really a team. You’ll be working side by side for five days. By Friday, you’ll be a problem-solving machine, and you’ll share a deep understanding of the challenge and the possible solutions. This collaborative atmosphere makes the sprint a great time to include people who don’t necessarily agree with you.

  Bring the troublemaker

  Before every sprint, we ask: Who might cause trouble if he or she isn’t included? We don’t mean people who argue just for the sake of arguing. We mean that smart person who has strong, contrary opinions, and whom you might be slightly uncomfortable with including in your sprint.