Failing smarter and learning faster in engineering

Software development has been evolving. When I started in the industry, working at companies like Microsoft, we would bet many person-years of development and many millions of dollars into the development of products that would sometimes be hits and sometimes be total duds. We were building blind. This was partly due to our waterfall processes, but also to how software was distributed and marketed. A dud for a smaller company could mean the end of the line.

The cost of failure was incredibly high. Over the years, we learned how to take some of that risk out by switching to agile software development and now Lean. Working this way we can learn quicker, and take smaller risks. However, there are other things we can do in how we architect our software or roll it out that can also reduce the technical and product risk and help us fail smarter and learn faster. In this talk, I speak about my experiences building waterfall products at Microsoft, building agile and lean at Adobe, Spotify and Avvo; and I give real architectural, cultural and organizational tools you can use to make your projects and team more failure-safe.