Discovery using data

A multitude of tools, and data is available today to help product managers in the decision making process.

Discovery questions should focus on understanding user workflows and behavior, and data can be collected from diverse sources to identify challenges.

Let’s consider the business problem we are trying to achieve is “increased customer retention”.

In this case, the set of focused discovery questions would look like:

• What does the workflow of our core user look like?
• What do our sucessfully retained users do in the product?
• Where do users we don’t retain drop out of the product?
• What are customers actually trying to accomplish?

It is indeed harder to arrive at answers due to multiple answers to each questions.

Methods includes talking to internal teams, understanding product usage data, and direct customer feedback. It will often be a continuous motion.

Product analytics/usage data: Helps understand if and how people are using your product, which features they do and don’t engage with, where users drop out of key workflows, and what different segments of users do differently.

To explain using the case we use (increased user retention).

• What features do retained customers engage with the most?
• Do you see usage patters in the rest of your user base?
• Where did users who churned drop out of workflows?

Direct customer feedback:

• Check for dip in feature with high engagement.

In app guide is a useful method to collect targeted feedbacks to understand the cause.

How to conduct user research

User research methodologies

• Generative user research: Also called exploratory research, where PMs conduct this research during the discovery phase to get the lay of land and gain a contextual understanding of the problems faced by users such as how much of a pain it is, what kind of solutions they’ve tried before, and how they are solving it now? — User interviews: Asks open ended questions in one-on-one interviews — User surveys: Collect input from a larger group of users at once — Card sorting: Shows how users categorize and organize information or concepts and can be used to design an information architecture, workflow, menu structure, or navigation within a product.

Each of these generative user research types can be used based on the information we have on the problem. User interviews when the ambiguity is more to card sorting when clarity is more. User surveys should be asked when you have a specific idea or question that you want to ask at scale.

It is also important how you deliver these surveys. In App surveys are often the best. Try giving it in the particular feature or context.