In our previous post, we discuss the difference between Minimal Viable Products (MVP) and Minimal Marketable Features (MMF). Here we are going to look at how Lean Discovery Practices help us validate our MVP to ensure we are building the right thing.
Eric Ries, the author of Lean Startup, reminds us “The big question of our time is not can it be built, but should it be built?”
To know that, we have to go through quick build, measure, and learn cycles and with each cycle, we have to decide whether to persist or pivot based on new learnings. We have to accept the fact that our so-called requirements are really hypotheses that are only proven true when we get real data that validates whether what we build is actually usable and valuable.
We have a target solution in mind, and through successive iterative deployments, we validate our assumptions. We get there by applying some qualitative analysis (before end user deployment), as well as quantitate analysis (after user deployment) and with each step, we decide to pivot or persist until we reach our desired solution which will likely end up different from the original target solution we had in mind. Through validated learning, we are able to build a more valuable product.
The solutions we are building need not only be viable from a business perspective and feasible from a technical perspective. The solutions also need to be usable from a customer perspective. To do that, we need to understand what end users need and use data to drive decisions. The solutions need to address the whole user experience while making it simple and intuitive to use.
To do that we apply Lean Discovery practices rooted in Lean UX to help us quickly take a concept, validate it internally, prototype it, test it externally, and learn from user behavior.
Techniques such as a Value Proposition Canvas make it explicit how we are creating value for our customers by mapping customer gainers and pain points to product features that relieve these pain points and increase their product engagement.
Test cards help us set up the concepts or hypothesis along with the tests, metric and success criteria to validate. It follows the format of:
Hypothesis: We believe that …
Test: To verify that, we will …
Metrics: And measure …
Criteria: We are right if…
This way each requirement is really set up as a hypothesis that later becomes a requirement or a feature if the hypothesis is proven true. This, in turn, leads to an MMF.
Problem interviews and solution interviews, whether in person or via surveys, help us validate internally that we are solving the correct problem and the solutions proposed properly address these problems.
Creating personas help in understanding the different user groups, their demographics, behaviors, and needs. Creating journey maps through the entire process from beginning to end help in understanding specific pain points for each persona.
Storyboarding, sketching, and building prototypes help in validating the solution proposed. Running usability testing on our prototypes and, later on, our product ensures the feature is meeting its intended purpose.
Feedback from these learnings is applied into the next cycle and is used to determine the next MMF to work on.
It’s important to note that Lean Discovery is not a phase we do at the beginning of a project, but it’s an ongoing activity of discovery feeding an ongoing activity of delivery as well as validating the ongoing results of delivery and determining what to build next. This is referred to as Hypothesis Driven Development (HDD). In our next blog, we will look at how we can make HDD a reality by applying Agile Delivery techniques that allow us to quickly build and deliver a feature to further validate its viability, feasibility, and usability.
Lean Discovery Practices is the fifth in an 8-part series “Succeeding with Digital Service Delivery” from Excella Software Development Lead Fadi Stephan.
Part 1: What is Digital Service Delivery
Part 3: Is Agile the Answer?
Part 5: Lean Discovery Practices
Part 6: Agile Delivery Practices
Part 7: A DevOps Mindset