Build-measure-learn (BML) is a process of building a product, measuring consumer metrics and learning from them to better respond to customer needs and improve the product for the ultimate sustainability of the company.
BML generally relies on releasing a minimum viable product (MVP), one that satisfies market demands without going overboard on features that customers don't want. After release of the MVP, customer feedback and metrics can be used to improve future versions of the product.
In a BML process, generally a market and its customers are researched to form hypotheses of what will work for an MVP. BML is designed to avoid less effective methods such as rushing to market with a “release early and often” approach or spending too much time trying to perfect a product, only to find there is no customer need. Once the market is known, the MVP must ensure that the development and manufacturing will result in the desired quality for the customer through building management processes.
Once the product is built, customer feedback and metrics are used to verify hypotheses, validating learning about customers and the market. This verified knowledge is then used again, repeating BML for improved versions of products with less effort wasted on things customers don't care for.
BML, validated learning and minimum viable product (MVP) are all concepts from “The Lean Startup,” authored by entrepreneur and co-founder of IMVU, Eric Ries. This method attempts to apply practices similar to scientific methods to regiment success in startup companies.