About a year ago we started working with the lean startup process.
If you are not familiar with it the main ideas are: create target metrics, experiment in the market with minimum viable products, iterate often. It’s called build-measure-learn. If you miss the target metrics then you value the learning (you knew more than you did before) and if you keep missing the target metrics you need to pivot (start doing something very different).
We liked the idea of a more scientific approach to product development. It also promises to limit the timeframe and costs of developing products. It lets you feed winning ideas with new iterations and kill those ideas that don’t fit the market. As a team we felt like it was a realistic approach to product development (most products fail because of lack of market fit rather than poor execution) and would reduce some of the pressure on the development team of creating a winning product. It has a clear lifecycle with feedback loops built in.
So, how’s it working out for us? Not all product development we do fits into the lean startup approach. In fact, our biggest product has been a multi-year, multi-million dollar effort that only recently hit the market. It took too long and was too far removed from the market feedback throughout its development period to use build-measure-learn. There was no way of doing a minimum viable product. We have, however, been able to put the lean startup approach in practice around the launch and communication of that product and around other products.
It has been harder to implement than we thought.
The metrics we set often need to be adjusted. It’s useful for us to ask the questions about how a product will perform. Once we are getting market feedback we typically need to add two or three new target metrics and remove at least one metric that turned out to be unimportant.
It is challenging to throw a minimum viable product (MVP) out to the market. They don’t make you look very good. Our MVPs are still a little more advanced than we would like. In other words, they cost more time and effort to get the same market feedback as a less developed MVP.
With tighter hypotheses we feel like we can be learning more. We found that there has been too much room for interpretation of results. In some cases the results are too ambiguous to add to our store of working knowledge.
We’re working the process. On balance we’re happy with the lean startup approach. We’re making it work for us. We’ve benefitted greatly despite the initial challenges and we’ll continue to practice the lean startup approach. It’s not easy but the benefits in terms of clear feedback loops from the market and less pressure on the team are significant. We’ve got a bunch of products in the pipeline and we’ll be using the lean startup approach as we develop each of one.