In today’s fast-paced world, software developers, product designers, and product managers face multiple challenges, including the need to deliver new value to customers quickly and respond to the demand for better customer experiences. To address these challenges, development teams must adopt a continuous, rapid evolution approach and utilize a key concept from the scientific world: hypotheses.

To deliver what customers expect at the speed they demand, teams must hypothesize and make data-based decisions quickly. By experimenting early and often, teams can gain valuable feedback on what works for customers and discard features that provide little benefit or hinder their goals.

Traditionally, user requirements were captured as user stories where a persona requests something to achieve their goal. However, a hypothesis is a statement that predicts a benefit that can be measured. 

For example, “We believe that [function] will lead to [outcome] and this will be proven when [measurable condition].” 

This approach, with its roots in Lean UX, requires teams to focus on metrics that truly matter, rather than vanity metrics, which are often superficial and easily manipulated.

Teams should conduct split tests with existing features to determine which option is best or even build multiple prototypes and compare them when developing new features. The goal is to continuously improve with the minimal amount of work required. In some cases, teams don’t even have to write code to test their assumptions; they can use prototypes, surveys, interviews, and other means to gather necessary data. Teams should maintain a backlog of hypotheses ranked by risk and impact that is regularly revisited.