This is how efficient work works in Silicon Valley
On a cloudy morning in May 2014, John Zeratsky walked into a beige building in Sunnyvale, California. John wanted to speak to someone at Savioke Labs, one of Google Venture's newest investments. He made his way through a maze of corridors, took a short flight of stairs to a simple wooden door that read "2B" and entered.
high-techCompany look a little disappointing to those who expect red-rimmed computer eyes, StarTrek-like holodecks or top-secret designs. Most of Silicon Valley is essentially a pile of desks, computers, and coffee mugs. But behind door 2B there were piles of circuit boards, plywood cutouts and plastic fittings fresh from a 3D printer, as well as soldering irons, drills and drafts. Yes, top secret designs indeed. "This place looks like a startup should be," thought John.
How robots make our work easier
And then he discovered the machine. It was a three-foot-high cylinder about the size and shape of a kitchen trash can. Its shiny white body had an elegantly tailored, curved shape that broadened up and down. At the top was a small computer screen that looked almost like a face. And the machine could move: it slid across the floor under its own power. "It's the relay robot," said Steve Cousins, Savioke's founder and CEO. Steve wore jeans and a dark t-shirt and was excited like a middle school physics teacher. He looked proudly at his little machine. "It was built here, and it was made of precast." The relay robot, Steve explained, was designed for hotel guest services. He could navigate automatically, ride the elevator by himself, and bring things like toothbrushes, towels and snacks to hotel guests. Steve and John watched the little robot carefully wheel around a desk chair and stop near an electrical outlet.
Savioke (pronounced "Savvy Oak") had a team of world-class engineers and designers, most of whom were former employees of Willow Garage, a prestigious private robotics research laboratory in Silicon Valley. They all had one thing in common: the vision of making everyday life easier for people with the help of robots as hard-working helpers - in restaurants, hospitals, old people's homes and so on. Steve had chosen to start with hotels because they offered a relatively simple and consistent environment with one constant problem: morning and evening work peaks when the front desk was overwhelmed with check-in, check-out, and room service assignments. That was the perfect location for a robot.
AI as a maid
The following month, the first fully operational relay robot went into operation at a nearby hotel, doing real room service for real guests. If a guest had forgotten their toothbrush or razor, the robot would come along. But there was a problem. Steve and his team feared that the guests might not like the servant robot. They might find it a nuisance or even be afraid of it. The robot was a marvel of engineering, but Savioke wasn't sure how the machine should behave towards humans.
The risk of having towels brought by a machine was too impersonal, Steve explained. Savioke's chief designer, Adrian Canoso, came up with a lot of suggestions on how to make the exterior of the robot friendly, but the team still had a lot of decisions to make before the robot was really fit for the public. How should he communicate with the guests? How much personality was too much of a good thing? "And then there was the elevator," said Steve. John nodded. "Personally, I feel uncomfortable just having to ride in an elevator with other people." "Exactly," said Steve, slapping the relay. "And what happens when a robot is added?"
How to make structured decision-making processes more efficient
Savioke had only been in business a few months. The company had focused on development and technology. They had negotiated a pilot project with Starwood, a hotel chain with several hundred properties. But there were still important questions in the room - questions crucial to success, and there were only a few weeks until the start of the pilot project. It was the perfect time for a sprint. Sprint is Google Ventures' unique five-day process with which key questions can be answered by creating prototypes and then testing them on real customers. Sprint is a kind of compilation of the "greatest hits" from business strategy, innovation, behavioral science, design and more - packed into a step-by-step process that any team can use.
The Savioke team considered dozens of ideas for their robot and then used a structured decision-making process to select the most convincing solutions, without group-think processes. A realistic prototype was created in a single day, and in the end, the team won target customers and set up a makeshift research laboratory in a nearby hotel. We would like to tell you that we, the authors, were the genius heroes of this story. It would be wonderful if we could sneak into every possible company and share our brilliant ideas that will make it a resounding success. Unfortunately we are not geniuses. The sprint at Savioke worked because you worked there with real experts: those involved were already part of the team. We just gave them the right process for their job.
This is how efficient problem solving works in 3 steps
And here is the sequence of the Savioke sprint. No robots are built in your company? Never mind. We use the exact same sprint structure for software, services, marketing and many other areas.
- First we put together a number of checklists, including a shopping list for necessary materials or a brief overview of the respective daily routine of the week. You don't have to memorize everything at once, but before we begin we need to carefully plan the process so that it will be a complete success.
- Then the team shoveled a full week off. From Monday to Friday they canceled all meetings, activated the "out of office" response function of their e-mail accounts and focused exclusively on one question: How should the robot behave in the presence of people?
- Next, they set a deadline. Savioke agreed with the hotel to conduct a live test on the Friday of sprint week. The countdown was on and the pressure increased significantly. We only had four days to design a working solution and create a prototype.
- On Monday, the Savioke team analyzed and reviewed everything they knew about the problem. Steve emphasized the importance of guest satisfaction, which is meticulously measured and tracked in hotels. If the relay robot increased satisfaction ratings during the pilot, the hotel would order more robots. If the value stagnated or fell and there was no order, the young start-up would find itself in a tricky position.
- Together we created a kind of overview plan, similar to a map, in order to identify the greatest risks. Imagine this plan like a storyboard in a film: guest meets robot, robot hands guest the toothbrush, guest is enthusiastic about the robot. There were critical moments on this storyboard when the robot and the guest meet for the first time: in the lobby, in the elevator, in the hotel corridor and so on. What point should we focus on? If you only have five days to sprint, you have to focus on well-defined goals. Steve chose the service fulfillment moment. If that works, the guest is delighted. If things go wrong, reception staff may have to spend all day answering questions from confused guests.
Identify recurring problems
One major concern kept cropping up: The team feared that the robot might appear too intelligent on the outside. "We're all spoiled by C-3PO and WALL-E," Steve said. “We expect robots to have feelings and plans, hopes and dreams. But our robots are not that sophisticated. If a guest speaks to him, he won't answer. And if we disappoint the guests, we have failed. «On Tuesday the team switched from problem definition to possible solutions. Instead of a rumbling brainstorming session, each participant worked out their own solution. And not just the designers. Tessa Lau, chief technology officer and chief engineer in robotics, worked on it, Izumi Yaskawa, responsible for business development, and Steve, the start-up's CEO, also pondered possible solutions. On Wednesday morning, all the walls of the conference room were plastered with sketches and notes. Some ideas were new, while others were old ideas that had been discarded or never thought through. All in all, we had 23 competing solutions.
How could we condense them to a few? In most organizations, the decision would require weeks of meetings and endless email traffic. But we only had one day available. The Friday test hovered over our heads like the sword of Damocles, and everyone felt the pressure. By means of voting procedures and a structured discussion, we came to a calm, quick decision without arguments and arguments.
No result without risk
The test would include a number of Savioke designer Adrian Canoso's most daring ideas: a face for the robot and a soundtrack of beeps and ring tones. In addition, the team decided on one of the fascinating but controversial ideas of the solution sketches: If the robot was satisfied, it would perform a little dance of joy. “I still fear we're giving it too much personality,” Steve said. "But we have to take the risk." "Should it explode, we can still drive it back," Tessa said. When she saw the look on our faces, she said, “It was just kidding. Don't worry, the robot can't explode at all. "
When Thursday broke, we had just eight hours to create the prototype for the live test on Friday. That was actually too close. With two tricks we managed to get our prototype ready on time: Much of the hard work had already been done. On Wednesday we agreed on the ideas that we wanted to test and documented each potential solution in detail. There was only execution. The robot didn't have to function autonomously, as it would in the end in a hotel. All he had to do was do one job: bring a toothbrush to a guest's room. Tessa and her engineering colleague Allison Tse programmed and adjusted the robot's movements using an old laptop and a PlayStation controller. Adrian put on large padded headphones and orchestrated the sound effects. The "face" was designed on an iPad and mounted on the robot. At five in the afternoon he was finished.
Never forget the customer
For Friday's test, Savioke arranged interviews with guests at the on-site Starwood hotel in Cupertino, California. At seven in the morning we set up a makeshift research laboratory in one of the hotel rooms by attaching webcams to the wall with electrical tape. And at 9.14 a.m. the first female guest began his interview. The young woman studied the decoration of the hotel room: light wood, neutral tones, a latest generation TV. Modern and pleasant, but not unusual. So what was the interview about? Next to her stood Michael Margolis, research partner at Google Ventures. For the time being, Michael wanted to keep the reason for the test secret as a surprise. He had planned the entire interview in advance to get answers to certain questions from the Savioke team. At the moment he was trying to find out the travel habits of his interviewee and to encourage her to react honestly when the robot appeared. Michael adjusted his glasses and asked a series of questions about their hotel habits. Where did she put her suitcase? When did she open it? And what would she do if she forgot her toothbrush?
"I don't know, I would probably call the front desk." Michael scribbled notes on a clipboard. "Okay." He pointed to the phone on the desk. "Please call reception." No sooner said than done. "No problem," replied the receptionist. "I'll have one brought to you at once." When the young woman hung up the phone, Michael continued his questions. Did she always travel with the same suitcase? When was the last time she left something at home? Drrrring. The phone cut her off. She picked up the receiver and received an automated message: "Your toothbrush is here." Without thinking further, the woman walked to the door and opened it. At the Savioke corporate office, the sprint team had gathered around the video screens in excitement, watching their reaction tense. "Oh my god," she said. "That's a robot!" Slowly its shiny outer shell opened and a toothbrush appeared inside. The robot made a couple of ringtones and beeps as the young woman confirmed room service on the touchscreen. When she rated this experience with five stars, the little machine performed a short dance of joy by swinging back and forth. "That's really cool," said the woman. "If the hotel uses this robot, I always come here."
What really mattered wasn't her words; it was the enthusiastic smile we could see on the video stream. And it was what she wasn't doing - no awkward pauses and no frustration whatsoever while handling the robot. When we watched the first interview on live video, we were very nervous. In the second and third interviews we laughed and even cheered. All guests reacted in the same way. They were thrilled when they saw the robot, and no one had any problems receiving their toothbrush, confirming receipt on the screen and sending the robot back on its way. Several guests wanted to get a second toothbrush just to see him in action again. They even took selfies with him. But nobody tried to involve him in conversation.
Address and eliminate deficiencies
At the end of the day, our whiteboard was covered with green hooks. The risky robot personality - the blinking eyes, the sound effects and, yes, even the "happy dance" - were a total success. Prior to the sprint, Savioke was concerned they were going to promise too much about the robot's capabilities. Now they realized that equipping the small machine with limited "human" features was possibly the secret to increasing customer satisfaction. Of course, not all the details were perfect. The screen was slow to respond. The timing of some of the sound effects was not yet right, and an idea of integrating computer games into the touchscreen did not go down at all with the guests.
These shortcomings meant we had to re-prioritize some of the engineering work, but there was still time. Three weeks later, the robot started his full-time job at the hotel. The relay turned out to be a real hit. Stories about the charming robot have appeared in the New York Times and the Washington Post. In the first month alone, Savioke received more than a billion press mentions. Most importantly, the robot got to the guests. By the end of the summer, Savioke had placed so many orders that the company was barely able to keep up with production. Savioke had taken a risk by giving the robot its own personality. But the team was confident because the sprint gave them the opportunity to quickly test the risky idea.
Answer crucial questions quickly
As a partner of GV, our mission is to help our startups answer these critical questions. We are not consultants who are paid by the hour. We are investors who are successful when the companies we support are successful. In order to help them find a quick solution to their problems and quickly stand on their own two feet, we have optimized our sprint process so that it produces the best results in the shortest possible time. Best of all, however, the process draws on people, knowledge, and tools that every team already has.
By sprinting our start-ups, we shorten endless debates and compress months into a single week. Instead of first launching a minimal product to understand whether an idea is really worthwhile, our companies receive clear data from a realistic prototype in advance. The sprint equips our start-ups with a kind of superpower: They can "fast forward" into the future and anticipate how customers will react to their finished product before they make expensive investments. When a risky idea proves itself in a sprint, the result is overwhelming. However, the greatest return on investment is offered by failures, even if they are painful. Identifying defects that are critical to success after just five days of work is the maximum in efficiency. This is a tough learning process, but it prevents costly failure on the market at an early stage.
Always check the direction you are going
At GV we have sprints with companies like Foundation Medicine (developer of highly specialized methods for cancer diagnosis), Nest (manufacturer of intelligent household electrical appliances) and Blue Bottle Coffee (as the name suggests, coffee producers). We used sprints to test the profitability of new lines of business, build the first version of new apps, improve mass-produced products, define marketing strategies, and generate reports for medical tests. Investment bankers used sprints to determine their next strategy; the Google team built the self-driving car using sprints, and high school students used the process to complete a challenging math problem.
This text is a do-it-yourself guide to help you run your own sprint and answer your most pressing questions for your business. On Monday you will analyze the problem and determine the most important point that you will focus on. On Tuesday you will sketch possible solutions on paper. On Wednesday you will make difficult decisions and turn your ideas into a testable hypothesis. On Thursday you will create a realistic prototype and on Friday you will test it on real people.
It all depends on the perfect team
We do not give advice from above, we deal with specific details. We'll help you put together the perfect sprint team from people you're already working with. You will learn highly demanding things (for example, how to get the most out of the different opinions of your team members and the vision of a leader), moderate things (for example, why your team should turn off phones and computers for three days) and very simple things ( why you should have lunch at one o'clock). You won't end up with a complete, detailed, ready-to-ship product, but you will move forward quickly and you will know exactly if the direction you are going is in the right direction.
You can discover some methods that you are already familiar with and others that are new to you. If you are familiar with Lean Development or Design Thinking, you will find that the sprint is a convenient way to use these guidelines. If your team uses what are known as agile processes, you will find that our definition of "sprint" is different but complementary. And if you've never heard of these methods, don't worry, you won't have any problems. This method is suitable for experts and beginners alike; it is made for everyone who is facing a great opportunity or idea or a great problem and somehow needs to find a start. Each step has been tested, modified, retested, and assessed over the course of our more than one hundred sprints, and fine-tuned based on the feedback we received from our growing sprint community. If the sprint doesn't work, it's not because of this text.
More knowledge - PDF download, eCourse on demand or personal advice
Offline download: Download this text as PDF - Read usage rights, Because we do not automatically submit the title of this text for privacy reasons: When buying in "interests" the title register if support is needed. After buying text exclusively Download at this URL (please save). Or for a little more directly an entire book or eCourse with this text buy, read on.
Read customer feedback and buy a book on this topic at a discount: Do you like this text and want to read more information about it? Buy the right book including this text, buy it here in two languages, as a member even with a 20 percent discount. Would you like to take a look at the book first? You can do this by previewing the book look at and then purchase on the book page.
German edition: ISBN 9783965963726
English version: ISBN 9783965963733 (Translation notice)
Your eCourse on Demand: Choose your personal eCourse on this or another desired topic, As a PDF download. Up to 30 lessons with each 4 learning task + final lesson. Please enter the title under "interests". Alternatively, we are happy to put together your course for you or offer you a personal regular eMailCourse including supervision and certificate - all further information!
Consultant packages: You want to increase your reach or address applicants as an employer? For these and other topics we offer special Consultant packages (overview) - For example, a personal phone call (price is per hour).