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

by Jake Knapp

  Stewart told us he had a hunch about “Bot Team.” Potential customers had difficulty imagining what it would be like to use Slack at work. With the simulation provided by “Bot Team,” Stewart predicted those customers would get it right away.

  Stewart is a repeat entrepreneur who is known for having really good hunches. He’d followed a hunch to create Slack after the game Glitch failed to catch on. A decade earlier, he’d followed a hunch to create the photo sharing service Flickr. Needless to say, when Stewart said he had a hunch about the sketch called “Bot Team,” we listened. Still, Merci, the product manager, worried that the fake team might confuse customers. Not only that, she estimated that the engineering required to implement it properly could take four to six months.

  Merci had credibility, too. She was an experienced entrepreneur, having started a software company of her own before joining Slack. And as head of the project, she was also a Decider for the sprint. Her supervotes had gone to another sketch: “The Tenacious Tour,” a solution that explained Slack’s interface step-by-step.

  This supervote conflict posed a problem, because we couldn’t figure out how to put “Bot Team” and “The Tenacious Tour” together in the same prototype. It would be way too much explanation for one website. With two great ideas and no way to combine them, there was only one sensible course of action. It was time for a Rumble.

  • • •

  On Wednesday morning, your team will make a Sticky Decision to narrow down to the most promising sketches. But what if, like Slack, you end up with two (or even three) winning sketches that can’t coexist? Since Deciders get three supervotes, this kind of conflict happens all the time. It might sound like a problem, but actually it’s a bonanza.

  When you have two good, conflicting ideas, you don’t have to choose between them at all. Instead, you can prototype both, and in Friday’s test, you’ll be able to see how each one fares with your customers. Your prototypes will battle head-to-head, like professional wrestlers whacking each other with folding chairs. We call this kind of test a Rumble.

  A Rumble allows your team to explore multiple options at once. For Slack, it meant building two prototypes: one for “The Tenacious Tour” and one for “Bot Team.” Merci and Stewart didn’t have to argue, or compromise on a watered-down solution. With a sprint, they could get data in just five days—before they made a commitment. (Later in this book, we’ll tell you whose hunch was correct.)

  Of course, it doesn’t always make sense to do a Rumble. Sometimes, there’s just one winning sketch. Sometimes, there are many winners, but they all fit together. Savioke’s winning solutions for their robot personality—sound effects, survey, and happy dance—could all coexist in one prototype. Which was lucky, because we only had one robot.

  If you think you can combine your winning sketches into one product, don’t bother with a Rumble. Instead, put them together into your best shot at solving the problem. This all-in-one approach has advantages, too. Your prototype will be longer and more detailed.

  Rumble or all-in-one

  If you have more than one winning solution, involve the whole team in a short discussion about whether to do a Rumble or combine the winners into a single prototype. Typically, this decision about format is easy. If it’s not, you can always ask the Decider to make the call.

  Now, if you decide to do a Rumble, you’ll have one more small problem. If you show your customers two prototypes of the same product, you risk sounding like an optometrist: “Which version do you prefer? A, or B? A? Or B?”I

  Luckily, the resolution to this murky situation is easy, and even fun: You get to create some fake brands. Once your prototypes have their own distinct names and look, customers will be able to tell them apart.

  In Slack’s sprint, we decided to use the Slack brand for one prototype, but we needed a fake name for the other one. We knew customers wouldn’t take a prototype seriously if it had a name like “Acme” or “Clown Pants.” It had to sound like a realistic competitor to Slack. After thinking up a few options, the team chose “Gather” for the second prototype. That name was perfect: It wasn’t a real product, but it sounded like it could be.

  Blue Bottle Coffee faced a similar challenge when they tested different ideas for their online store. They needed fake brand names that sounded like real coffee companies, and they came up with “Linden Alley Coffee,” “Telescope Coffee,” and “Potting Shed Coffee.”

  Inventing fake brands is fun, but it’s also a potential time waster. To keep the process short, we use an all-purpose brainstorm substitute that we call Note-and-Vote. Here’s how it works:

  Note-and-Vote

  Throughout the sprint, you’ll have times when you need to gather information or ideas from the group and then make a decision. The Note-and-Vote is a shortcut. It only takes about ten minutes, and it works great for everything from fake brand names to deciding where to get lunch.

  1. Give each team member a piece of paper and a pen.

  2. Everyone takes three minutes and quietly writes down ideas.

  3. Everyone takes two minutes to self-edit his or her list down to the best two or three ideas.

  4. Write each person’s top ideas on the whiteboard. In a sprint with seven people, you’ll have roughly fifteen to twenty ideas in all.

  5. Everyone takes two minutes and quietly chooses his or her favorite idea from the whiteboard.

  6. Going around the room, each person calls out his or her favorite. For each “vote,” draw a dot next to the chosen idea on the whiteboard.

  7. The Decider makes the final decision. As always, she can choose to follow the votes or not.

  * * *

  By lunchtime on Wednesday, you will have decided which sketches have the best chance of answering your sprint questions and helping you reach your long-term goal. You’ll also decide whether to combine those winning ideas into one prototype or build two or three and test them in a Rumble. Next, it’s time to turn all these decisions into a plan of action so you can finish your prototype in time for Friday’s test.

  * * *

  I. Not that there’s anything wrong with optometrists. We love optometrists.

  12

  Storyboard

  By Wednesday afternoon, you’ll be able to feel Friday’s test with customers looming ahead. Because of the short timeline, it’s tempting to jump into prototyping as soon as you’ve selected your winning ideas. But if you start prototyping without a plan, you’ll get bogged down by small, unanswered questions. Pieces won’t fit together, and your prototype could fall apart.

  On Wednesday afternoon, you’ll answer those small questions and make a plan. Specifically, you’ll take the winning sketches and string them together into a storyboard. This will be similar to the three-panel storyboards you sketched on Tuesday, but it will be longer: about ten to fifteen panels, all tightly connected into one cohesive story.

  This kind of long-form storyboarding is a common practice in movie production. Pixar, the film studio behind movies like Toy Story and The Incredibles, spends months getting their storyboards right before committing to animation. For Pixar, the up-front effort makes sense: It’s much easier to change storyboards than to re-render animation or re-record voice tracks with super-famous actors.

  Sprints have a shorter timeline and smaller scale than a Pixar production. But storyboarding is still worthwhile. You’ll use your storyboard to imagine your finished prototype, so you can spot problems and points of confusion before the prototype is built. By taking care of those decisions up front, you’ll be free to focus on Thursday.

  Slack’s storyboard showed how their prototype would work, following the customers as they read about the two products (Slack and Gather) in a news article, then clicked through to the websites, and, hopefully, signed up for the service.

  At a glance, this storyboard might look like the world’s most boring (and poorly drawn) comic strip. But for the Slack team, it was a masterpiece. The storyboard contained all of our best ide
as, stuck together in a story we could all understand, and we hoped it would make sense to customers, too. When we looked at the whiteboard, we saw this:

  When you’re finished, your storyboard will make just as much sense to your team as Slack’s storyboard did to them. Next, we’ll dive right in and talk about how to make one, and as we go, we’ll show you how Slack put theirs together.

  First of all, somebody needs to be the storyboard “artist.” We put the word “artist” in quotes because the job doesn’t require artistic talent. In this case, the “artist” is just someone willing to write on the whiteboard a lot. (It might be another good time to give the Facilitator a break.)

  Draw a grid

  First, you need a big grid with around fifteen frames. Draw a bunch of boxes on an empty whiteboard, each about the size of two sheets of paper. If you have a hard time drawing long straight lines (and who doesn’t), use masking tape instead of a marker.

  You’ll start drawing your storyboard in the top left box of the grid. This frame will be the first moment that customers experience on Friday. So . . . what should it be? What’s the best opening scene for your prototype?

  If you get it right, the opening scene will boost the quality of your test. The right context can help customers forget they’re trying a prototype and react to your product in a natural way—just as if they had come across it on their own. If you’re prototyping an app, start in the App Store. If you’re prototyping a new cereal box, start on a grocery shelf. And if you’re prototyping business communication software?

  In real life, Slack was getting lots of great press. Many of their new customers discovered the service by reading an article about it. So Merci suggested we use a fake New York Times article for our opening scene. The article could be about “new trends in office software”—giving us the perfect opportunity to introduce our two prototypes, Slack and Gather. Here’s how we drew it on the storyboard:

  The fake news article is a useful opening scene. We used the same method in our sprint with Blue Bottle, when we opened with a (fake) New York Times article about three (fake) up-and-coming coffee companies.

  But there are lots of ways to open your storyboard. Flatiron Health wondered if existing users of their software would change their workflow for a new clinical-trial tool. A news article wouldn’t have made much sense. Instead, Flatiron’s opening scene was an email inbox—the place research coordinators would receive notifications from the new system. For Savioke, the opening scene was checking in to a hotel and forgetting a toothbrush. The trick is to take one or two steps upstream from the beginning of the actual solution you want to test.

  Choose an opening scene

  How do customers find out your company exists? Where are they and what are they doing just before they use your product? Our favorite opening scenes are simple:

  • Web search with your website nestled among the results

  • Magazine with an advertisement for your service

  • Store shelf with your product sitting beside its competitors

  • App Store with your app in it

  • News article that mentions your service, and possibly some competitors

  • Facebook or Twitter feed with your product shared among the other posts

  There are other possible opening scenes. Your prototype might begin with an everyday routine: a doctor’s folder of paper reports, an engineer’s email inbox, or a teacher’s classroom newsletter. If you’re testing a new kind of store, you might start the moment people enter the front door.

  It’s almost always a good idea to present your solution alongside the competition. As a matter of fact, you can ask customers to test out your competitors’ products on Friday right alongside your own prototype.

  • • •

  Once you choose an opening scene, you only have nine hundred more decisions to make before you’re done with your storyboard. Just kidding . . . kind of.

  Storyboarding is a simple process, with a ton of tiny decisions along the way. Those tiny decisions can be tiring, but remember—you’re doing your future self a favor. Every decision you make now is something you won’t have to think about when you build your prototypes.

  Fill out the storyboard

  Once you’ve selected an opening scene, the storyboard “artist” should draw it in the first frame (the “artist” will be standing at the whiteboard while everyone else gathers around). From there, you’ll build out your story, one frame at a time, just like a comic book. As you go, you’ll discuss each step as a team.

  Whenever possible, use the sticky notes from your winning sketches and stick them onto the whiteboard. When you come to a gap—a step in the story not already illustrated by one of the solution sketches—don’t fill it in unless it’s critical to testing your idea. It’s okay if some parts of your prototype don’t work. You can have buttons that don’t function and menu items that are unavailable. Surprisingly, these “dead ends” are generally easy for customers to ignore in Friday’s test.

  If you decide the gap does need to be filled, try to use something from your “maybe-later” sketches, or from your existing product. Avoid inventing a new solution on the spot. Coming up with ideas on Wednesday afternoon isn’t a good use of time or effort. You will have to do some drawing, of course: filling in gaps when necessary and expanding on the winning sketches so that your prototype will be a believable story. Remember that the drawing doesn’t have to be fancy. If the scene happens on screen, draw buttons and words and a little arrow clicking. If the scene happens in real life, draw stick figures and speech bubbles.

  Making your storyboard will likely take up the entire afternoon. To make sure you finish by 5 p.m., follow these guidelines:

  Work with what you have.

  Resist inventing new ideas and just work with the good ideas you already came up with.

  Don’t write together.

  Your storyboard should include rough headlines and important phrases, but don’t try to perfect your writing as a group. Group copywriting is a recipe for bland, meandering junk, not to mention lots of wasted time. Instead, use the writing from your solution sketches, or just leave it until Thursday.

  Include just enough detail.

  Put enough detail in your storyboard so that nobody has to ask questions like “What happens next?” or “What goes here?” when they are prototyping on Thursday. But don’t get too specific. You don’t need to perfect every frame or figure out every nuance. It’s okay to say: “Whoever builds this tomorrow can decide that detail.” And then move on.

  The Decider decides.

  Storyboarding is difficult because you already spent a lot of your limited decision-making energy in the morning. To make it easier, continue to rely on the Decider. In the Slack sprint, Braden was the “artist” drawing the storyboard, but Merci made the decisions. It was extra work for her, but it kept us fast and opinionated.

  You won’t be able to fit in every good idea and still have a storyboard that makes sense. And you can’t spend all day arguing about what to include. The Decider can ask for advice or defer to experts for some parts—but don’t dissolve back into a democracy.

  When in doubt, take risks.

  Sometimes you can’t fit everything in. Remember that the sprint is great for testing risky solutions that might have a huge payoff. So you’ll have to reverse the way you would normally prioritize. If a small fix is so good and low-risk that you’re already planning to build it next week, then seeing it in a prototype won’t teach you much. Skip those easy wins in favor of big, bold bets.

  Blue Bottle Coffee’s storyboard shows all of the clicks required to select and order coffee beans.

  Detail from Savioke’s storyboard, showing the robot delivery at the guest’s door.

  Keep the story fifteen minutes or less.

  Make sure the whole prototype can be tested in about fifteen minutes. That might seem short, especially since your customer interviews will be sixty minutes long. But you’ll have to a
llow time for your customers to think aloud and answer your questions—not to mention starting up the interview at the beginning and winding it down at the end. Fifteen minutes will take longer than fifteen minutes. And there’s another, practical reason for this limit. Sticking to fifteen minutes will ensure that you focus on the most important solutions—and don’t bite off more than you can prototype. (A rule of thumb: Each storyboard frame equals about one minute in your test.)

  Once you’ve incorporated all of the winning sketches, the storyboard will be complete. And you’ve finished with the hardest part of the sprint. The decisions are made, the plan for your prototype is ready, and Wednesday is a wrap.

  FACILITATOR NOTES

  Don’t Drain the Battery

  Decisions take willpower, and you only have so much to spend each day. You can think of willpower like a battery that starts the morning charged but loses a sip with every decision (a phenomenon called “decision fatigue”). As Facilitator, you’ve got to make sure that charge lasts till 5 p.m.

  Wednesday is one decision after another, and it’s all too easy to drain the battery. By following the Sticky Decision process and steering the team from inventing new ideas, you should be able to make it to 5 p.m. without running out of juice.

  But you’ll have to be mindful. Watch out for discussions that aren’t destined for a quick resolution. When you spot one, push it onto the Decider: