Innovation Journal: Customer In and Technology Out


A few years back, Amazon founder and CEO Jeff Bezos was receiving an award as Entrepreneur of the Year from Stanford GSB. Several of us joined our classmate and friend, Steve Kessel, now leading the new Amazon Go stores, at the awards ceremony. Jeff was asked how an online bookseller managed to create both the Kindle and Amazon Web Services (AWS), now a $13B business by itself and the leading Cloud Platform — a really mindbending achievement. He said simply, “We’re committed to innovation at our core, and we do both kinds, customer in and technology out.”

When I mentioned the idea again at our annual CEO Summit, I got a lot of questions about what it meant and realized it’s an important concept more founders should practice.

Customer in. The Kindle came from talking to customers and thinking about their needs. Amazon had avid readers, was worried about the digital transformation of that business, and focused on creating something that its existing customers (highly active readers) would value. That process typically starts with customer discovery, understanding their priorities, and empathizing with their problems such as lugging multiple heavy books on a trip because most avid readers like reading lots of different things simultaneously.

Technology out. AWS is fundamentally rooted in technical innovation. Amazon had solved key problems around cloud computing at scale for itself. Having done so, they thought, “Other companies will have similar challenges if they don’t already. Let’s try selling our platform to companies.” Even though it was a very different business from Amazon’s e-commerce business, they were leveraging a strength and competence they already had that gave their cloud platform key advantages like economies of scale, willingness to operate in low margin markets, and an eye towards automation. These are investments every web stack infrastructure benefits from but only the largest companies can afford to make.

These examples provide a useful framework for thinking about start up innovation and how founders can put the puzzle pieces together to solve big problems in unique ways. There are significant opportunities in both approaches.

Most but not all SaaS companies start with the customer in approach. Brandon Levey, Founder and CEO of Stitch Labs, was running a t-shirt business and started building software to manage his inventory. He is an example of a founder having deep domain experience, understanding an emerging problem, imagining the product that might solve this problem, and gathering the resources to make it happen.

Most infrastructure companies are technology out and begin with a technical innovation. appOrbit started with a team of container experts looking for the right application to apply their unique knowledge of the burgeoning container ecosystem. They came across the opportunity to enable DevOps for all applications (both new and legacy apps) that drives their business today. They are an example of a founding team looking for a place to apply new capabilities enabled by new shifts in the technology stack and massive hybridization of the existing tech stack.

Customer in and technology out can be combined in really interesting ways. Alation CEO Satyen Sangani’s years of experience trying to make sense of enterprise data assets combined with CTO Venky Ganti’s machine learning expertise resulted in an innovative approach which lets business users find, understand, use and govern enterprise data assets without active human metadata management or documentation. They are an example of how new data and algorithms around predictive analytics, Artificial Intelligence, and Deep Learning affect every category of company but are only meaningful if you find the right, human way to apply it and make this incredible technology accessible.

If Amazon has shown the world one thing, it’s the power of both customer in and technology out, and that they can live in the same company, applied against a number of different businesses. The important thing is to understand your priorities based on how you start, then rapidly iterate based on how the market is meeting your product. Then, be smart. Don’t lose your way trying to do too much too soon.