At a recent Engineering Q&A session, I spoke a bit about the idea of an innovation budget. We’ll accomplish more in the long-term if we protect some space for innovation in addition to maintaining our focus on immediate goals. That sounds great, someone noted in chat, but what did I even mean when I kept talking about innovation?
There are a lot of ways to think about innovation, but for me I always start with the idea of a hill climbing algorithm. You’re at a point in a two dimensional space, each point in that space has a value, you want to be at the point with the highest value, but you can only see the values of adjacent points. The simplest solution is to look at the surrounding points, and move to the one with the highest value. Do this over and over, and you’ll successfully climb the closest hill. That’s delightful, but at the top a new problem arises: you can’t climb any higher, but there are probably higher points out there somewhere, you just have no idea where.
Innovation is figuring out what to do when you’re at the top of a hill.
Well, that’s the simplified, inspirational version. A more useful version recognizes that moving between some points is more expensive than others. When you’re climbing a shallow slope, it’s pretty easy to keep going: all you need is a good pair of hiking boots. As the hill gets steeper, each step takes more energy, and you might need rope or climbing shoes.
This brings us to the two fundamental innovation strategies:
- Explore around you for new hills to climb
- Optimize how you climb your current hill
Explore around you for new hills
Clayton Christensen’s The Innovator’s Dilemma tries to explain why innovative companies rarely have a second, equally innovative act. The thesis is that most businesses are structured such that they prioritize investing into existing, successful businesses at the expense of novel, unproven endeavors. The problem isn’t that these successful companies are poorly run, but rather it’s being well run that prevents companies from further innovation.
Mihaly Csikszentmihalyi’s Creativity looks at innovation through a different lens. Two key components of innovations are, one, combining learnings across multiple fields (e.g. physics and chemistry) and, two, unstructured time for thinking. Few modern workplaces create space for either of those components, let alone both.
That said, innovation clearly is possible within the modern workplace, and it’s worth thinking about our best tools for enabling innovation:
- Customer-orientation: how can we create more long-term value by prioritizing customer need over short-term financial outcome?
- Data-driven thinking: how do we use data to drive objective learning about our area and users?
- Product-thinking: how can we enable innovation by aiming and incentivizing teams properly? (Particularly love Melissa Perri’s Escaping The Build Trap on this topic.)
- User-research and design thinking: how do we understand and operate from real customer behavior and needs rather than from our assumptions about their behavior and needs?
That said, the tool I’ve found most effective for enabling exploration is a simple investment thesis. Agree as a leadership team that 20% of each team’s time should be prioritized against innovation, and make sure at least one executive is measuring and fighting to preserve that 20% as you grow.
(20% is, of course, just a made up number.)
Optimize how you climb your current hill
In Jeff Besos’ 2011 letter to Amazon shareholders, he talked about driving innovation by eliminating gatekeepers:
I am emphasizing the self-service nature of these platforms because it’s important for a reason I think is somewhat non-obvious: even well-meaning gatekeepers slow innovation. When a platform is self-service, even the improbable ideas can get tried, because there’s no expert gatekeeper ready to say “that will never work!” And guess what – many of those improbable ideas do work, and society is the beneficiary of that diversity.
Broadening the idea, there are huge opportunities for innovation that are expensive to access but not inherently expensive. A/B tests are valuable because they allow you to know with confidence that one approach outperforms another, but they’re at least equally valuable because they derisk execution. With an effective experimentation platform you can safely allow teams to launch ideas that you’re not sure will succeed. By reducing risk, the untenable becomes acceptable.
There are even cases where process prevents obviously good ideas from moving forward. For example, when I joined Uber it was a multi-day task to provision a new service in our service oriented architecture. This meant that many folks who wanted new services simply weren’t able to get them provisioned. The move to self-service pulled the infrastructure team out of the decision loop: if a team believed a new service was necessary to make progress, they could take on the provisioning cost even if the infrastructure team had other, higher priorities.
This model is focused on maintaining innovation within an existing business rather than the creative innovation fostered within a new business. There’s a lot more to say about innovation beyond this one mental model, but hopefully it provides some food for thought.