Best Practices

How an Innovation Factory Scales Product Breakthroughs

Published Nov 13, 2018

It’s a bit of a seductive thought that the grass is always greener on the other side. It’s human nature to think: “if only I had x, or y, then I could really succeed, really make a difference.” In several of my previous positions at design agencies, I remember how we’d find ourselves sitting around thinking: “if only our customers would get out of their own way; if only they’d do just exactly what we tell them.” This wishful thinking crosses industries too: “if only our startup had the budget of a big company; if only our big company could be nimble like a startup.” It might be true that the grass is greener “over there,” but we are here, so how do we make this grass as green as possible?

Short, Green Grass  

We started Alpha Health with a small group of seven employees. As an innovation facility, we wanted to solve a wide range of issues in public and personal health. We experimented constantly, succeeded quickly, and failed quickly as well. We were incredibly nimble and focused on one product at a time. Communication was never an issue. Need some clarification? Lean over and tap the person next to you, they’d clear it up. Documentation wasn’t needed, just turn around and ask, or look at the wall behind my old desk, the plan was right there for everyone to see. In so many ways it was the dream, we were the greener grass.

But our goal is to do the most good, for the most amount of people; that’s too big of a job for seven people. We needed to scale; and we did, from a team of seven to 70 in two years.

Our new team was incredibly diverse and included data scientists, AI specialists, designers, product managers, developers, public health specialist, marketers, mathematicians, behavioral economists, psychiatrists, clinical psychologists, the list goes on. With this rapid growth we knew we would be able to tackle larger challenges, but first, we needed to learn how to keep innovating at high speed with all these new voices in the room.

Keeping Pace

As our team grew, so did our possibilities and potential to create products. But so did the potential to get bogged down in “big company” problems. The meetings became bigger, decisions took longer to make, big design and development by committee became more common, speed slowed down, and indecision around strategy started to creep in.

And so we were introduced to a second big challenge. First, how do we work as a team, and second, where do we focus our energy?  Should we create a product that does voice analysis for depression or Alzheimer’s detection? Can we convince teenagers to make the best decisions for their long-term health? Should we communicate a person’s health status as an embedded device? Should we create a Tamagotchi-like device as a health coach? Unlimited options, limited time.

We weren’t willing to take our foot off the gas. Prototyping fast, test fast, succeed fast, fail fast; these were standards we were built on. Not everyone is used to this type of pace or method of work. We have employees from diverse backgrounds including academia, government, and large organizations, as well as other startups who weren’t as familiar with this “lean startup” approach. It did take effort to bring everyone on board, but within a small team, this has proven to work well. Prototyping recommendation engines or even academic research has delivered results in two weeks by having designers, product managers, researchers, and developers work closely together.

Experimentation Visibility

To simplify communications and to make it clear where we are going and why, we created five OKRs. For example, “% Improvements of health,” or “Retention.” Knowing that they are just approximations, we set benchmarks, which we update once we have more data. These five OKRs are printed on a big poster in the middle of our office, a constant visual reminder to the whole team that we are going somewhere together.

Based on these OKRs we run experiments, from small online tests to improve retention, to a larger-scale research study to see if we are able to detect someone’s mental health status. To keep it crystal clear why we are doing an experiment, what our hypothesis is, and what good looks like, we use experiment cards.

On each experiment card we have: “We believe that…” with our hypothesis. “To verify that” with what the experiment is going to do. “And measure” is the one metric we use to define success and then “We are right if” with a benchmark on what good looks like.

We stick these posters on the wall and report on progress to the whole team each week and in our roadmap meetings every two weeks.

High Level and Day-to-Day Working in Tandem

As our meetings started to stretch longer, and as too many stakeholders from strategy to CEO were dragged into the details of the product, we decided to separate the strategy from the ongoing product work.

Every two weeks we have a roadmap meeting where we have research, development, strategy, the CEO, and the product team come together. We report on the hypothesis cards, our OKRs, and update the roadmap. Changes can be made by strategy and the CEO on the roadmap in that meeting, but not in between.

Nurture Autonomy

As a result of our growth, we found that our team members felt less in control and less user-centered, a natural side-effect of team growth. To improve motivation and help our employees focus their energy and incredible talent, we created autonomous pods or teams.

These smaller teams are empowered to make their own decisions. We don’t require too much documentation and the teams are kept tight enough so that personal, face-to-face communication is the most effective way to problem solve. Each pod has a product manager, designer, developer, and data scientist. These pods sit together, have their dailies and weeklies together, and their own Trello boards and backlog. We’ve tried to recreate the magic of that small group of seven, that initial “green-grass” feeling across the organization by introducing these small working groups. In our experience, this has worked great to improve communication, speed of design and development, and a sense of ownership for people’s work.

Reward Both Experimentation and Failure

It’s been my experience at the companies I’ve worked for, and on the teams that I’ve been a part of, that the most impactful and innovative experimentation always took place where exploration and playfulness was promoted; where people were nice to each other, and where there was a lot of psychological safety to speak up your mind, comment and suggest ideas. No ideas were wrong. The end result was clear, the drive to get to that end was insatiable, but the ways to get to that end result were limitless and might include a few detours along the way.

This is the feeling that drives me—at previous startups I’ve been a part of— and now at Alpha. The excitement for the product, and the motivation to take it to its limits; the challenge of attacking the problem right in front of you, and not wishing the grass was any greener.