As a Microsoft employee working on the company's now-defunct series of Encarta CD-ROMs, Jake Knapp knew there had to be a more efficient way to create and test products. "You literally had a yearlong cycle," Knapp tells Inc. "It took you a year to find out if the thing that you made was any good, and if it would have success in the market."

That, coupled with Knapp's desire to make his own days more productive, led him to think about the perfect process for inventing something new. "I realized there was this discord between what I hoped to accomplish and what was actually happening day by day," says Knapp. "The meetings, checking email, doing 10 things at once--instead of doing one really important thing."

Knapp eventually left for Google. One weekend he traveled to Sweden to meet with some of the company's engineers with the hopes of creating a new video chat platform. Three days later, a working version of Google Hangouts was born.

"When I looked back afterwards, I thought, Those three days were worth three months, or six months, or nine months, " says Knapp. "How can you reproduce that?"

Using the Sweden experience as a blueprint, Knapp broke his technique down into a five-day method that he dubbed a design sprint. The process, outlined in his upcoming book, Sprint: How to Solve Big Problems and Test New Ideas in Just Five Days, takes teams from identifying a problem on Monday to testing a prototype with customers on Friday. It frees companies up to experiment with their ideas without sinking in excessive money or time--especially useful at startups, where these resources are often limited.

Knapp has led sprints for Google projects like search, Gmail, and Google X. Since moving to Google Ventures (now formally known as GV), he's done the same with startups the company has invested in, such as Slack, Nest, Blue Bottle Coffee, and 23andMe. Here's a breakdown of Knapp's process for designing the perfect product.

Getting started.

Before beginning, Knapp recommends putting together a team of seven people--more can slow things down, and fewer puts you at risk of not getting enough diverse opinions. The founder, CEO, or whoever else has final say should be present, as well as at least one expert from finance, marketing, customer support, tech, and design.

The sprint should take place in a room with plenty of whiteboard space. And save the texting for breaks--no devices allowed.

Day 1: Agree on the goal.

The first day's focus isn't on solving the problem, but defining it. The employees in your sprint will have different areas of expertise--about your customers, about what's economically feasible--so it's important everyone lay out the information they have. Together, the group should decide on a long-term goal for the company: What do we want to be doing six months from now, or a year from now? Then turn the focus to this sprint: What problem do you want to solve this week--and how will solving this problem help you accomplish the larger goal?

Day 2: Come up with multiple solutions.

Knapp doesn't like brainstorming. People tend to favor the craziest ideas, he says, without thinking about how they'll actually be accomplished. Or, even worse, the winner is simply the loudest voice in the room.

Instead, Knapp prefers what he refers to as parallel individual work. Each person in the sprint sits down with a pencil and paper and sketches out their idea. "It gives people a long period of quiet work when they can think through their idea," he says, "detail it out and not have to simultaneously explain it in words--which is a special skill that not everyone has." Sketching isn't a skill everyone has, either--which is why it's important to stress to employees that the focus will be on the idea, not on the artistry.

Day 3: Choose the best option and create a plan of attack.

A traditional company hierarchy rarely allows founders and employees to offer their ideas with equal weight. The design sprint does just that. The sketches are taped to a wall or whiteboard, and they all remain anonymous. The ideas are then studied in silence. "A lot of the things we do in the sprint are meant to help people who are introverted--people who aren't as comfortable or as capable of pitching their ideas--to be on a level playing field with the founder or with anyone on the team," says Knapp.

Next, each participant gets a handful of stickers they can quietly place next to the individual aspects of each idea they especially like. "Sometimes there are pieces of an idea that turn out to be really strong," says Knapp, "even if the whole thing didn't make sense." The solutions are then discussed one at a time. Notes are written around them with a dry-erase marker or Post-its, and a timer ensures all the ideas get equal attention. Each person in the sprint places a big sticker next to the solution they prefer and, using this as guidance, the designated decision maker chooses which idea--or combination of ideas--with which the team will move forward.

The beauty of the five-day sprint is that there's minimal pressure to choose the perfect solution--even if it's a disaster, you've only lost five days and not several months. And, at least you've found out what doesn't work.

Once that solution is chosen, your team should create a storyboard of what the customer experience will be like with your product, from discovery through the very end.

Day 4: Build a prototype.

One day sounds intimidating to build an entire prototype, but Knapp points out you'll already have most of what you need. If you're in the robot-making industry, for example, you probably already have a previous prototype that you can reprogram or redesign, as well as the people capable of doing so. If your product is an app, you don't need to create an entire working model--just an interface with the essential elements included. The prototype can also be marketing materials or a service, depending on what problem you're trying to solve. 

"This is where you get to be like the crew of thieves in Ocean's Eleven," says Knapp. "Everybody's got their special skill, everybody's breaking up and doing their part and coming back together. It's actually quite fun."

Day 5: Test it.

Research shows that five is the perfect number of customers on which to test your product--beyond that returns are low, since the same feedback will keep coming up. On the final day, pick five customers who will each spend an hour using the prototype and speaking with a member of your team, while the rest of the crew watches via live-streamed video in another room. The team takes notes and decides what should stay and what needs tweaking--or a complete revamping. There isn't always a finished product by this stage, but your team will know what it needs to do to get there.

"By the end of the day on Friday," Knapp says, "there's clarity about what to do next. Some solutions will totally fail, some will succeed, and often you're in the middle: You have something promising, but it needs work.

"Instead of spending months or even years on a hunch that may turn out to be wrong, you're able to answer your questions really quickly--to stop debating in the abstract, and start making progress."

Google's Trick to Crafting and Testing an Idea in 5 Days
Published on: Mar 1, 2016