by Jake Knapp
During Each Interview
• Take notes as you watch. Hand out sticky notes and markers. Write down direct quotes, observations, and interpretations. Indicate positive or negative. (p. 219)
After Each Interview
• Stick up notes. Stick your interview notes in the correct row and column on the whiteboard grid. Briefly discuss the interview, but wait to draw conclusions. (p. 220)
• Take a quick break.
At the End of the Day
• Look for patterns. At the end of the day, read the board in silence and write down patterns. Make a list of all the patterns people noticed. Label each as positive, negative, or neutral. (p. 222)
• Wrap up. Review your long-term goal and your sprint questions. Compare with the patterns you saw in the interviews. Decide how to follow-up after the sprint. Write it down. (p. 222)
Frequently Asked Questions
Q: Can I facilitate a sprint without any experience?
A: Yes.
With this book, you have everything you need. In fact, you’re much better prepared than we were when we started out!
Q: Do sprints require long hours?
A: No.
The sprint requires roughly thirty-five hours of work from each participant. We want the team to be rested so they stay sharp and do their best work. You’ll be home for dinner.
Q: Will sprint participants miss a lot of other work?
A: Sort of.
It’s impossible to spend thirty-five hours in a sprint and do your normal job. But since the sprint only runs from 10 a.m. to 5 p.m. each day, participants have the option of spending time in the morning to keep up with other work.
Q: Do sprints work at big companies?
A: Yes.
At big companies, it can be difficult to get time from the Decider and other experts. Focus on scheduling cameo appearances for Monday, and be sure to have the Decider delegate another decision maker who can participate every day.
Q: Do sprints work for hardware products?
A: Yes.
The biggest challenge for a hardware sprint will be prototyping. Here are three techniques for creating a hardware prototype in one day: Modify or build on top of an existing product, even if it’s incomplete. Use 3D printing or other rapid fabrication techniques to prototype your product from scratch. Or create a Brochure Façade, which allows customers to react to your product without seeing the actual product. For more, see page 185.
Q: Do sprints work for [insert super-hard-to-prototype product or service here]?
A: Almost certainly.
Adopt the prototype mindset and pretty much anything is possible. For more on the prototype mindset, see page 168.
Q: Can sprints work at nonprofits?
A: Yes.
Just like startups, nonprofits have big challenges and limited resources. The definition of “target customer” might be different, but questions about nonprofit concerns like fund-raising, public relations, and community services can all be answered by prototyping and testing with real people.
Q: Can sprints work in classrooms?
A: Yes.
The biggest challenge to running a sprint in a classroom is scheduling. If you can find a solid week, go for it! But if your class meets only once or twice a week for a few hours each session, you’ll have to be creative.
At Columbia and Stanford, professors have adapted the sprint process by having students do one “day” each class session (either in class, or as homework with their team). Dividing the process will create a lack of continuity, and a lot of “boot up” time each session. Help students out by encouraging them to take lots of photos. If possible, let them keep their maps, sprint questions, and other notes on Post-it easel pads or something similar.
Q: Can we run a sprint with team members in different places?
A: Maybe.
Running a sprint with people who aren’t physically in the room is tricky. If you want to include them for Monday’s Ask the Experts exercise, or as observers of Friday’s tests, that’s relatively easy to accomplish with a video conference. But if you want to include them for other steps, you’ll need ingenuity and a strong relationship. The bottom line: Nothing that happens on paper or on a whiteboard will work very well for your remote teammates. (Hopefully the technology to solve this problem is just around the corner, but it’s not quite here yet.)
Q: Can I run a sprint by myself?
A: Sort of.
Don’t expect a solo sprint to be as good as a sprint with your team. But we’ve talked to people who have done it successfully, and the techniques for a sprint can be useful on your own. For example, set a timer and force yourself to come up with multiple solutions to a problem. Prototype your ideas to answer specific questions before diving into implementation. See below for more tips on how to use parts of the sprint.
Q: Can we stop after the “Decide” step?
A: No.
This question comes up a lot, and we know it’s tempting. Once you identify promising ideas, you could easily jump ahead and start building them for real. The problem is this: Ideas that seem perfect on Wednesday often prove to be flawed after Friday’s test. By continuing with the sprint, and prototyping and testing, you and your team will learn whether those ideas are really as good as they appear.
Q: Can we run a one-, two-, or three-day sprint?
A: We don’t recommend it.
If you compress the schedule, you either won’t finish prototyping and testing (see above), or you’ll work crazy hours and burn out. Neither is a recipe for good results.
Q: How about a four-day sprint?
A: Maybe.
If the team has experience running five-day sprints, you may be able to compress Monday, Tuesday, and Wednesday’s activities into just two days. However, you should not compress prototyping and testing—allow one full day for each.
Q: If we’ve just finished a sprint, can our follow-up sprint be shorter?
A: Yes.
Follow-up sprints are exceptions to the five-day rule. Since you’ll already have a map and a prototype, as well as results from your first test to help you create new solutions and make decisions, you can often accelerate a follow-up sprint. Two things don’t change: You’ll still need a realistic prototype, and you’ll still need to test with five customers.
Q: Can we use parts of the sprint process without running a whole sprint?
A: Yes.
For a big challenge, use a full sprint. But there are lots of sprint techniques that are useful in other settings. If you need to make a small decision in a meeting, try a Note-and-Vote (p. 146). If you find yourselves frustrated by problems, try writing How Might We notes (p. 73). If you’re talking about solutions in the abstract, do a Four-Step Sketch to make them concrete (p. 109). Every meeting benefits from a Time Timer (p. 47) and a Facilitator writing notes on the whiteboard (p. 36).
And you can conduct customer interviews (p. 204) at absolutely any time—with a prototype, with your real product, with competitors’ products, or even with no product at all. We guarantee you’ll learn something.
Q: Is Friday’s test a focus group?
A: No.
A “focus group” is ten or so customers discussing an idea all together, all at once. Focus groups are plagued by the worst of group dynamics: shy people not talking, loudmouths talking too much, sales pitches, and a group-formed opinion that doesn’t reflect anyone’s honest feelings.
By comparison, Friday’s test features one-on-one interviews and observation of customer reactions. In these interviews, you can believe what you see.
Q: Can we run Friday’s test remotely, over the phone or by video conference?
A: Yes, but extra care is required.
Our partner Michael Margolis runs remote tests all the time, where he uses video-conference software to share computer screens and interview customers. But this kind of interview is more difficult. You’ll need to work extra hard to engage your customer, put her at ease,
and encourage her to think aloud. And the technology presents another challenge. You don’t want to waste valuable time getting your video-conference software up and running, so practice ahead of time and send your customers a detailed how-to guide for connecting.
Q: Can we test our prototype with fewer than five people?
A: No.
After four customer interviews, it’s usually difficult to see patterns. After five, it’s easy. (Check out Jakob Nielsen’s research chart on p. 193 for an illustration of this phenomenon.) If you schedule five people and only four show up, you might be okay. But don’t schedule four or fewer.
Q: Can we test with friends and family?
A: No.
You can only trust the results when you interview customers who match your target profile. Even if your friends and family happen to fit the profile, there’s another big problem: They’re biased, or at the very least they know too much. In your test, you’re looking for honest reactions from real-world customers—something you can never get from someone who knows you.
Q: What about random people at Starbucks?
A: Probably not.
If you’re building a product for random people who hang out at Starbucks, this might work. But even so, you’ll probably need an additional round of screening to find the exact right five customers—the Starbucks regular, the single parent, the business traveler, whatever.
Q: Should we ever interview customers before a sprint?
A: Yes!
We know how difficult it is to make time for customer interviews before a sprint. We usually can’t do it. But if you can, this kind of “pre-research” provides a big head start. It’s especially helpful when you’re starting from scratch and don’t already know a lot about your customers or how they use your product. For example, Blue Bottle was new to selling coffee online, so we interviewed coffee lovers before the sprint to better understand how they shop for coffee.
Q: Are there more resources?
A: Yes.
Check out thesprintbook.com for more information on sprints.
Q: What if my question isn’t answered here?
A: We love answering questions about sprints.
The best way to reach us is on Twitter. Jake is @jakek, John is @jazer, Braden is @kowitz, and our team is @GVDesignTeam.
Thank-You Notes
Jake Knapp
First and foremost, thank you to my lovely wife, Holly, for your trusted advice. As my first reader, she helped shape this book—and saved the rest of you from several boring stories. Thanks also to Luke, who taught me the meaning of time, and Flynn, who kept me on task by asking every few hours if the book was done yet.
Thank you to my family: to Mom, for doing projects with me; to Dad, for being in the stands at every basketball game; to Becky and Roger, for tireless support; and to Steve, Rich, Nancy, Karol, Britton, and Mignonne, for putting up with your little brother. Probably the only person in my family who didn’t help me even the slightest bit is my nephew, Jack Russillo.
On tiny Orcas Island, I had more excellent teachers than I can list—but a big thank-you to Lyn Perry, Colleen O’Brien, Joyce Pearson, Eric Simmons, Steff Steinhorst, and (of course) Tish Knapp. Thank you to my many mentors at work, especially to Jeff Hall at Oakley; to Sheila Carter, Christen Coomer, Robb Anderson, Melinda Nascimbeni, and Dan Rosenfeld at Microsoft; and to Charles Warren, Jeff Veen, and Elaine Montgomery at Google. A very special thank-you to Irene Au, for supporting my sprint experiments from the very beginning, and to Michael Margolis, for years of patience and good humor, and for pointing out the positive side of my procrastination.
Thank you to Caroline O’Connor for guidance on the first blog posts about the sprint process and to Belinda Lanks for spreading those posts to a broader audience through Fast Company. Thank you to the many readers who ran your own sprints, shared your experiences, and asked for more stories and greater detail. Your interest was the spark for this book.
Many thanks to those who gave early advice on the book project: Joe Kraus, Jodi Olson, M.G. Siegler, Gaurav Singal, Kevin Rose, Scott Berkun, and Josh Porter. Thank you to Tim Brown for your insight and encouragement, and to Charles Duhigg for being so generous with your time.
A very, very big thank-you to our agents, Christy Fletcher and Sylvie Greenberg. Christy and Sylvie are the Obi-Wan Kenobis of Sprint. If this book is readable or useful, it is thanks to their expert guidance and ability to put themselves into the reader’s shoes.
Ben Loehnen, our editor at Simon & Schuster, can read entire manuscripts in three subway stops and spot a cliché through a wall of solid steel. But his real superpower was making us sound more intelligent than we actually are. Ben, if you ever decide to quit publishing and edit dialogue for everyday life, I’m interested.
Jon Karp at Simon & Schuster provided early and steadfast enthusiasm. Publishing a book is complicated work, and Sprint made it into the world with the help of Richard Rohrer, Cary Goldstein, Leah Johanson, Jackie Seow, Stephen Bedford, Ruth Lee-Mui, Brit Hvide, and many others at S&S. And a hand-lettered, UV-varnished, triple-embossed (is that a thing?) thank-you to Jessica Hische, who brought the cover to life.
Like a friend who points out spinach in your teeth, our test readers provided forthright criticism to help Sprint look its best. Thank you to Julie Clow, Paul Arcoleo, Mark Benzel, Jake Latcham, Aaron Bright, Kevin Sepehri, Andrea Wong, Jose Pastor, Justin Cook, Jenny Gove, Kai Haley, Nir Eyal, Steph Habif, Jason Ralls, Michael Leggett, Melissa Powel, Xander Pollock, Per Danielsson, Daniel Andefors, and Anna Andefors.
Some thank-yous don’t fit into obvious categories. Alex Ingram reviewed the Monday chapters over and over and over again (it was really a mess at first) and helped explain the sprint from a startup’s perspective. Sunkwan Kim and Elliot Jay Stocks advised us on print design. Becky Warren suggested the Marie Tharp story. Chip and Dan Heath wrote Made to Stick, the book that served as our primary inspiration for Sprint. (Keep on stickin’, Heath Brothers!)
The entire team at GV was exceptionally understanding and supportive throughout the book-writing process. Thanks especially to Mandy Kakavas, Ken Norton, Phoebe Peronto, Rick Klau, Kaili Emmrich, and Tom Hulme for your timely feedback and suggestions. A very special thank-you to Laura Melahn, who provided brutal honesty and energetic encouragement. And another very special thank-you to Jenn Kercher, who navigated approximately 1,001 legal questions—and caught some well-hidden typos. A diamond-studded thank-you to David Krane for your ideas, advice, and enthusiasm. And an extra special, super deluxe thank-you to Bill Maris, whose encouragement and support made Sprint possible.
It is very unlikely that this book would have been finished before the year 2027 without the help of Kristen Brillantes. She organized our time, put up with our chaos, and even listened to a robot voice read the manuscript aloud on a seven-hour road trip. Kristen, you are quite a friend.
From the very first outline to the last detail of the cover, Michael Margolis and Daniel Burka collaborated with John, Braden, and me on this book. They read and reread, provided detailed suggestions and constructive arguments, and convinced me to remove the most unfunny jokes. Michael, Daniel, Braden, and John—it is an honor to work with you.
John Zeratsky
Thank you to my wife and best friend, Michelle. Your love and encouragement make me a better person. I’m so lucky to be stuck with you.
Thanks to my parents, who supported and entertained a huge range of hobbies when I was a child. (This included sailing-yacht design and music production, among other things.) You helped me develop a love of learning that I’m very grateful for.
My grandpa in rural Wisconsin became an unlikely computer geek in the 1970s, and his passion for digital technology fueled my interest. He was more than a loving grandfather. He was a friend, a mentor, and a patient source of technical support.
Thanks to all my friends and colleagues at the Badger Herald. You gave me a thorough introduction to design and journalism. Thanks to you, I got to
design a newspaper, manage print production, write a column about jazz, and even lead our board of directors.
Thank you to Suzy Pingree and Nick Olejniczak. At the University of Wisconsin–Madison, Suzy and Nick made me feel at home—no small feat on a forty-thousand-student campus. Nick introduced me to blogging and taught me web development. Suzy allowed me to enroll in several of her graduate seminars. They introduced me to my first freelance design clients and provided support beyond words.
In 2005 the FeedBurner team gave me an unbelievable opportunity to jump in over my head. Thanks to Matt Shobe, Dick Costolo, Eric Lunt, and Steve Olechowski for my shot at the big time. I still can’t believe you let me do all that stuff.
In 2011 another group of partners gave me an opportunity that I’m not sure I deserved. Thank you, Braden, for recruiting me to join you at GV. Thanks to Bill Maris, David Krane, Joe Kraus, and the rest of GV for embracing design in venture capital before anyone else. I’m honored and humbled to be part of this team.
I have been privileged to work with dozens of GV portfolio companies. Their curious, talented teams gave me the most transformative learning experiences of my career. Thanks in particular to Pocket, Foundation Medicine, Blue Bottle Coffee, Savioke, and Cluster.
To Kristen, Daniel, Jake, Michael, and Braden (yes, you again, Braden): Thank you. I can’t tell you what a pleasure it is to work together. Thanks in particular for encouraging me to embrace writing again—almost ten years after I forgot I was a writer.
Finally, thanks to our agents, Christy Fletcher and Sylvie Greenberg, and our editor, Ben Loehnen. You led us from “what if” to “holy crap, this is happening” with strength and poise. And you gave us access to a literary world that we had only seen in movies and TV shows.
Braden Kowitz
Thank you to my parents, who showed me joy in creativity, taught me how to repair almost anything, and let me wander through the wilderness. They gave me a TI-99/4A computer and an accompanying cassette recorder to store my first program, all before anyone thought that was a good idea. I am so grateful to have inherited the joy of exploration from them.