Week 13 Day 6: The Danger of Only Valuing What You Can Measure

Yesterday we emphasized measurable value connections. Today is the counter-argument. Some of the most important work a team does -- mentoring junior developers, building team culture, maintaining documentation, reducing cognitive load -- has no direct revenue metric. If your value framework only rewards quantifiable impact, you create a team that optimizes for what is easy to measure while neglecting what is hard to measure but essential.

I learned this lesson after implementing a value-based prioritization system that was too rigid. The team started declining all work that could not be tied to a business metric. Documentation updates? 'Cannot quantify the value.' Helping the new hire get up to speed? 'That is not sprint work.' Investigating a subtle performance degradation that had not yet affected customers? 'No measurable impact yet.' Within two quarters, the team's code quality declined, new hires took twice as long to become productive, and a slow-building performance issue exploded into a customer-facing outage. The system was working exactly as designed -- and it was destroying the team. The fix was to create an explicit 'investment' category alongside revenue, margin, and overhead. Investment work has no immediate measurable return but builds capacity for future value. Documentation, mentoring, exploration, relationship-building. I allocated 20% of each sprint to investment work with no ROI requirement. The team's long-term velocity increased even though short-term output metrics dropped.

The tension between measurable and unmeasurable value is a central theme in management theory. Drucker, often misattributed as saying 'what gets measured gets managed,' actually warned against measurement-driven management in his later work (Drucker, 2001), noting that 'the most important things cannot be measured.' Muller (2018) in 'The Tyranny of Metrics' documents how measurement fixation has degraded performance in education, healthcare, policing, and business by incentivizing surrogate measures at the expense of actual outcomes. In software engineering, Goodhart's Law -- 'when a measure becomes a target, it ceases to be a good measure' -- manifests when teams optimize for velocity points, lines of code, or tickets closed rather than the outcomes those metrics were designed to proxy. The 20% investment allocation from level_2 has precedent in Google's '20% time' (Steiber and Alange, 2013) and 3M's '15% culture' (Govindarajan and Trimble, 2010), both of which produced outsized innovation returns by explicitly protecting unmeasurable exploration from metric-driven optimization. Kaplan and Norton (2004) addressed this in their 'Strategy Maps' framework by including 'learning and growth' as a foundational perspective that enables future financial performance without requiring current financial justification.