Making an Engineering Team's Work Estimates Reliable as a Walmart Product Manager

You are a product manager working with a development team at Walmart. How would you make sure the engineers size the work they are assigned accurately? Explain what you would do before, during and after estimation, which parts are your responsibility and which belong to the engineers, and how you would tell whether the estimates are getting better. ```hint Diagnose before prescribing Ask what makes this team's estimates wrong today, and separate the causes that come from your inputs as product manager (scope, acceptance criteria, shifting priorities, deadline pressure) from those that depend on the engineers' technical judgment. ``` ### Clarifying Questions - Are the team's estimates missing today, and if so, do they run consistently over or under, or is this a new team setting up its process? - Are estimates used only for sprint planning, or are they turned into dates promised to business partners, for example work that must ship before the holiday shopping peak? - Does the team estimate in relative units such as story points or in days, and does it keep a history of planned versus completed work? - How much of the team's time goes to unplanned work, such as production incidents or urgent requests from other teams? ### What a Strong Answer Covers - A diagnosis of why estimates miss before any fix is chosen, with the causes owned by product separated from those owned by engineering - A clear division of roles: engineers size their own work, while the product manager supplies clear scope, acceptance criteria and stable priorities and never bargains estimates down - Concrete practices for breaking work down and for handling unknowns, dependencies and unplanned work - A feedback loop that compares estimates with actuals at team level, without turning the measure into a target people game - How uncertainty is communicated to stakeholders who need dates ### Follow-up Questions - A senior business stakeholder needs a firm launch date before the team has finished discovery. What do you give them, and how? - Several sprints into your new process, the team still finishes well short of what it commits. What do you examine first? - One engineer's estimates are consistently much lower than the rest of the team's. How do you handle it?