A few weeks ago, I talked to Corey Innis from Neo about careers in the tech industry and what makes a great developer. He gave me great pieces of advice as well as a list of books to read. I decided to write a book summary for myself every time I finish one os the books he recommended. This one is about “The Lean Startup” by Eric Ries.
The author attempts to answer how entrepreneurs can build sustainable companies around a new set of products. Eric finds that the right approach to answer the question is in the Lean Startup methodology. This is inspired by Toyota’s lean manufacturing system, which advocates a just-in-time-production. Contrary to the idea of following a business plan word by word, Lean Startup concepts encourages entrepreneurs to focus their efforts on product validation. Entrepreneurs need to use rapid customer learning to accelerate product development cycles. This cyclical process of repeated customer learning and product development is called Build-Measure-learn feedback loop.
The lesson here is that if put into practice, this methodology will allow anyone to identify better what customers find valuable in the products and therefore are willing to pay for.
When entrepreneurs initiate product development, they should ask if a product “should” be built, not if it can be built. In order to check whether a product should be built, one runs validated learning experiments to assess customer demand for it. Progress then is measured in terms of the customers’ feedback gathered from validated learning, not projected milestones. How can someone accomplish this? Eric Ries recommends the following:
*Continuous deployment — Deploying small batches into production often in order to get feedback more quickly, and in smaller chunks.These are done via minimum viable products (MVPs); they have only the minimum amount of features required for it to be deployed to produce validated learning.
*Falsifiable hypothesis — Creating predictions for each change that can be proved correct or incorrect
*Actionable metrics — focus on a small set of well defined metrics to measure, deliver them in simple reports the entire team can understand, and link everything back to people
Eric specifically suggests that teams:
*Use cohort analysis
*Use A/B testing
*Measure people and their actions(make sure you don’t use vanity metrics)
Upon completing an experiment, a startup should modify and/or optimize its product hypotheses and then test them with a new experiment. If you think this sounds like a scientific experiment, it certainly is.
Once an entrepreneur tests the assumptions he/she can either iterate or pivot. A pivot is a strategic decision to change course, be it the target customer or another market opportunity.
All in all, this was a great read and now I understand that adopting the Lean Startup method as the main engine of a company as opposed to only a short-term experiment is necessary to achieve a sustained culture of innovation.