by Jake Knapp
4
Start at the End
Everybody knows the story of Apollo 13, but just in case, it goes like this: Astronauts head to moon, explosion on spacecraft, nail-biting return to earth. In Ron Howard’s 1995 movie version, there’s a scene where the team at Mission Control gathers around a blackboard to form a plan.
Gene Kranz, the flight director, wears a white vest, a flattop haircut, and a grim expression. He grabs a piece of chalk and draws a simple diagram on the blackboard. It’s a map showing the damaged spacecraft’s path from outer space, around the moon, and (hopefully) back to the earth’s surface—a trip that will take more than two days. The goal is clear: To get the astronauts home safely, Mission Control has to keep them alive and on the right course for every minute of that journey.
Mission Control’s blackboard looked sort of like this.
Throughout the film, Kranz returns to that goal on the blackboard. In the chaos of Mission Control, the simple diagram helps keep the team focused on the right problems. First, they correct the ship’s course to ensure it won’t veer into deep space. Next, they replace a failing air filter so the astronauts can breathe. And only then do they turn their attention to a safe landing.
• • •
When a big problem comes along, like the challenge you selected for your sprint, it’s natural to want to solve it right away. The clock is ticking, the team is amped up, and solutions start popping into everyone’s mind. But if you don’t first slow down, share what you know, and prioritize, you could end up wasting time and effort on the wrong part of the problem.
If Mission Control had worried about the air filter first, they would have missed their window to fix the trajectory, and the Apollo 13 spaceship might have careened off toward Pluto.I Instead, NASA got organized and sorted their priorities before they started on solutions. That’s smart. And that’s the same way your team will start your sprint. In fact (with the luxury of unlimited oxygen) you’ll devote the entire first day of your sprint to planning.
Monday begins with an exercise we call Start at the End: a look ahead—to the end of the sprint week and beyond. Like Gene Kranz and his diagram of the return to planet earth, you and your team will lay out the basics: your long-term goal and the difficult questions that must be answered.
Starting at the end is like being handed the keys to a time machine. If you could jump ahead to the end of your sprint, what questions would be answered? If you went six months or a year further into the future, what would have improved about your business as a result of this project? Even when the future seems obvious, it’s worth taking the time on Monday to make it specific, and write it down. You’ll start with the project’s long-term goal.
Set a long-term goal
To start the conversation, ask your team this question:
“Why are we doing this project? Where do we want to be six months, a year, or even five years from now?”
The discussion could take anywhere from thirty seconds to thirty minutes. If your team doesn’t quite agree about the goal or there’s any lack of clarity, don’t be embarrassed. But do have a discussion and figure it out. Slowing down might be frustrating for a moment, but the satisfaction and confidence of a clear goal will last all week.
Sometimes, setting the long-term goal is easy. Blue Bottle Coffee knew where they were headed in the long term: Bring great coffee to new customers online. Of course, they could have simplified their goal to “sell more coffee online,” but they wanted to keep the quality of the experience high, and they wanted to challenge themselves to reach new customers, not just their existing fans. They wrote a long-term goal that reflected that ambition.
In some sprints, setting the long-term goal requires a short discussion. Savioke wanted to accomplish a lot with the Relay delivery robot. Was the goal about improving the efficiency of the front desk staff? Was it about getting as many robots in as many hotels as possible? Savioke wanted to focus on customers, and use the same goal as the hotels: better guest experience.
Your goal should reflect your team’s principles and aspirations. Don’t worry about overreaching. The sprint process will help you find a good place to start and make real progress toward even the biggest goal. Once you’ve settled on a long-term goal, write it at the top of the whiteboard. It’ll stay there throughout the sprint as a beacon to keep everyone moving in the same direction.
• • •
Okay, time for an attitude adjustment. While writing your long-term goal, you were optimistic. You imagined a perfect future. Now it’s time to get pessimistic. Imagine you’ve gone forward in time one year, and your project was a disaster. What caused it to fail? How did your goal go wrong?
Lurking beneath every goal are dangerous assumptions. The longer those assumptions remain unexamined, the greater the risk. In your sprint, you have a golden opportunity to ferret out assumptions, turn them into questions, and find some answers.
Savioke assumed their Relay robot would create a better guest experience. But they were smart enough to imagine a future where they were wrong, and the robot was awkward or confusing. They had three big questions: Can we make a smooth delivery? (the answer was yes). Will guests find the robot awkward? (the answer was no, except for the sluggish touch screen). And the long shot: Will guests come to the hotel just for the robot? (surprisingly, some people said they would).
Just like the goal, these questions guide the solutions and decisions throughout the sprint. They provide a quasi-checklist that you can refer to throughout the week and evaluate after Friday’s test.
List sprint questions
You’ll list out your sprint questions on a second whiteboard (if you have one). We have a few prompts for getting teams to think about assumptions and questions:
• What questions do we want to answer in this sprint?
• To meet our long-term goal, what has to be true?
• Imagine we travel into the future and our project failed. What might have caused that?
An important part of this exercise is rephrasing assumptions and obstacles into questions. Blue Bottle Coffee assumed they could find a way to convey their expertise through their website, but before the sprint, they weren’t sure how. It’s not difficult to find an assumption such as Blue Bottle’s and turn it into a question:
Q: To reach new customers, what has to be true?
A: They have to trust our expertise.
Q: How can we phrase that as a question?
A: Will customers trust our expertise?
This rephrasing conversation might feel a little weird. Normal people don’t have conversations like this one (unless they’re Jeopardy! contestants). But turning these potential problems into questions makes them easier to track—and easier to answer with sketches, prototypes, and tests. It also creates a subtle shift from uncertainty (which is uncomfortable) to curiosity (which is exciting).
You might end up with only one or two sprint questions. That’s fine. You might come up with a dozen or more. Also, just fine. If you do end up with a long list, don’t worry about deciding which questions are most important. You’ll do that at the end of the day on Monday, when you pick a target for the sprint.
By starting at the end with these questions, you’ll face your fears. Big questions and unknowns can be discomforting, but you’ll feel relieved to see them all listed in one place. You’ll know where you’re headed and what you’re up against.
* * *
I. Pluto, if you’re reading, we still believe you’re a planet.
5
Map
J.R.R. Tolkien’s The Lord of the Rings is an epic adventure, spanning three volumes and hundreds of pages. There are invented languages, histories, backstories, and subplots galore. It’s an awesome story, but it’s also complicated.
Frankly, it’s easy to get confused while reading The Lord of the Rings. But Tolkien’s got your back. At the beginning of the book is a map. As the characters travel through locations such as Mount Doom, t
he Mines of Moria, and the Misty Mountains,I the reader can flip to the map and remind herself where the action’s happening and how it all fits together.
The map you’ll create on Monday isn’t so different: a simple diagram representing lots of complexity. Instead of elves and wizards moving through Middle Earth, your map will show customers moving through your service or product. Not quite as thrilling, but every bit as useful.
The map is a big deal throughout the week. At the end of the day on Monday, you’ll use the map to narrow your broad challenge into a specific target for the sprint. Later in the week, the map will provide structure for your solution sketches and prototype. It helps you keep track of how everything fits together, and it eases the burden on each person’s short-term memory.
But there’s one quality these maps do not have in common with the map from The Lord of the Rings: They’re simple. No matter how complicated the business challenge, it can be mapped with a few words and a few arrows. To show you what we mean, we’d like to introduce you to Flatiron Health—a company with a very complex challenge and a very simple map.
• • •
Outside, a flurry of snow and a lead-gray cloud bank muted the Manhattan skyline, but inside, the conference room was cozy. Four of us (Jake, John, Braden, and Michael Margolis, our research partner) had traveled to New York City for a sprint with Flatiron Health, one of GV’s largest investments. We were hosting the sprint at Google’s office in Manhattan, a former Port Authority building that covers an entire city block. The office floor plan is confusing—Jake got lost three times on the first day—but we’d found our way to an empty room on the ninth floor, pushed the table against one wall, and gathered rolling chairs into a circle around a whiteboard.
We already knew Flatiron’s backstory. The company was founded by a couple of friends, Nat Turner and Zach Weinberg. In the 2000s, Nat and Zach had built an advertising technology company called Invite Media and sold it to Google.
A few years later, the two started thinking about their next startup, and the topic of health care kept coming up. Both had seen friends and family struggling with cancer, and had witnessed, firsthand, the complexities of treatment. Nat and Zach got inspired. Large-scale data analysis, they believed, could sift through piles of medical records and test results and help doctors choose the right treatment at the right time. They left Google and started Flatiron Health.
The startup had tremendous momentum. Flatiron had raised more than $130 million in funding and acquired the industry’s leading electronic medical records company. They’d hired a world-class team of engineers and oncologists and signed on hundreds of cancer clinics as customers. The pieces were in place to begin a project they believed could have a profound effect on cancer outcomes: improving clinical trial enrollment.
Clinical trials provide access to the latest treatments. For some patients, that means drugs which might save their lives. But trials aren’t just about new drugs—they’re also about better data. The data from every trial is collected and organized, helping researchers learn about the efficacy of new and existing therapies.
But in the United States, only 4 percent of all cancer patients are in clinical trials. The other 96 percent of cancer treatment data is unavailable to doctors and researchers who might use it to better understand the disease and better treat future patients.
Flatiron wanted to make trials available to anyone who was eligible. They hoped to build a software tool to help cancer clinics match patients to trials—a painstaking job to do manually, and perhaps the biggest hurdle to trial enrollment. Patients with common forms of cancer might qualify for trials reexamining the efficacy of standard treatment. Patients with rare forms of the disease might qualify for a new, highly targeted therapy. There were so many unique patients and so many trials that it was too much for any human to track.
The company decided to start with a sprint and had assembled a great team. The Decider was Dr. Amy Abernethy, Flatiron’s chief medical officer. Nat, the CEO, was there for a few hours to give us background. Half a dozen of Flatiron’s leaders joined them. There were oncologists and computer engineers, and Alex Ingram, a product manager.II
In the morning, we completed our Start at the End exercises. Choosing the goal (“More patients enrolled in clinical trials”) was easy. We turned our attention to identifying the big sprint questions.
“We have to be fast,” said Amy. She has an unusual accent: equal parts Australia (where she earned her PhD in medicine) and North Carolina (where she spent years running cancer research at Duke University). “If you’ve just been diagnosed with cancer, you can’t sit around while every clinical trial is considered. You’ve got to start treatment now.”
Jake uncapped his whiteboard marker and thought for a moment, trying to turn the problem into a question. Then he wrote on the whiteboard for everyone to see, Can we find matches fast enough?
“Each clinic already has its own ingrained process,” said Alex, the product manager. “These are teams of people who have been working together in the same way for years. We’ve got to offer something drastically better than the status quo, or they’re not going to change their workflow.”
Jake added, Will clinics change their workflow?
With the sprint questions listed, we started on the map. Michael Margolis and Alex Ingram had interviewed staff at cancer clinics, and with help from Amy, they told us how trial enrollment worked.
To match patients with trials, doctors and research coordinators look at long lists of trial requirements: treatment history, blood count, DNA mutations in the cancer cells, and much more. As cancer care has become more sophisticated and targeted, those requirements have gotten more specific. “For a given trial, you might be talking about a handful of eligible patients across the country,” said Amy. “It’s like looking for needles in a haystack.”
Flatiron Health’s long-term goal and sprint questions.
It was an intricate and messy system. But, after an hour of discussion and a lot of revision, we were able to create a simple map:
Flatiron Health’s clinical trial enrollment map.
On the left was a list of the people involved in trial enrollment: the patient and the doctor (who were central to the treatment decision) and the clinic’s research coordinator (who was easy to overlook but might be the best informed about trial availability). From there, the map showed the patient scheduling an appointment, the doctor and staff searching for matching trials, the appointment, the complete enrollment, and finally, the beginning of treatment.
Behind those few simple steps were all kinds of difficulties with the enrollment process: overworked staff, missing data, and communication gaps. As Amy had explained to us, many of the doctors who were supposed to suggest trials didn’t even know which trials were open at their clinic. In the afternoon, we would have time to go through all of the problems and opportunities. But for now, with this map, we had enough to start.
• • •
Flatiron Health had a complicated problem and a straightforward map. Your map should be simple, too. You won’t have to capture every detail and nuance. Instead, you’ll just include the major steps required for customers to move from beginning to completion, in this case from cancer diagnosis to trial enrollment.
Let’s look at a couple more examples. (For bonus points, see if you can spot the common elements in every map.) On Monday of their sprint, Savioke had to organize information about robotics, navigation, hotel operations, and guest habits. This is their map:
Savioke’s robot delivery map.
On the first day of their sprint, Blue Bottle Coffee sorted through information about coffee selection, customer support, café operations, and distribution channels. Here is their map:
Blue Bottle Coffee’s online sales map.
The common elements? Each map is customer-centric, with a list of key actors on the left. Each map is a story, with a beginning, a middle, and an end. And, no matter the business, each map is simple. T
he diagrams are composed of nothing more than words, arrows, and a few boxes. So now that you know what a map looks like, you’re ready to make your own.
Make a map
You’ll draw the first draft of your map on Monday morning, as soon as you’ve written down your long-term goal and sprint questions. Use the same whiteboard you wrote your goal on and dive in. When we’re drawing our maps, we follow these steps (keep in mind, there’s a checklist at the back of the book, so you don’t have to memorize this):
1. List the actors (on the left)
The “actors” are all the important characters in your story. Most often, they’re different kinds of customers. Sometimes, people other than customers—say, your sales team or a government regulator—are important actors and should be listed as well. And sometimes, of course, there’s a robot.
2. Write the ending (on the right)
It’s usually a lot easier to figure out the end than the middle of the story. Flatiron’s story ended with treatment. Savioke’s story ended with a delivery. And Blue Bottle’s story ended with buying coffee.
3. Words and arrows in between
The map should be functional, not a work of art. Words and arrows and the occasional box should be enough. No drawing expertise required.
4. Keep it simple
Your map should have from five to around fifteen steps. If there are more than twenty, it’s probably too complicated. By keeping the map simple, the team can agree on the structure of the problem without getting tied up in competing solutions.
5. Ask for help
As you draw, you should keep asking the team, “Does this map look right?”