Share with your CFO
Modernization capital is failing a test most finance teams aren’t running. Economist Sergey Kyunttsel, writing in CFO.com, argues that efficiency and facility upgrade projects routinely pass payback thresholds while quietly destroying value, because simple payback compresses operating hours, asset life, maintenance costs, and verification risk into one number that nobody owns after commissioning. His illustrative case: a $2 million lighting retrofit approved on a two-year payback delivers roughly two-thirds of forecast savings once real-world schedule shifts and unbudgeted replacement costs surface.
What this means for your business
The companies most exposed here are those running high volumes of mid-size modernization projects, where each individual approval feels too small to warrant deep scrutiny but the aggregate capital commitment is substantial. The structural problem Kyunttsel identifies isn’t weak analysis at approval, it’s the absence of anyone who retains ownership of the economic outcome after the equipment ships. Engineering closes the project, procurement closes the purchase, and finance closes the budget line. Nobody closes the loop on whether the value showed up.
The argument holds, and the mechanism is more corrosive than Kyunttsel’s relatively tame lighting example suggests. When AI-driven infrastructure upgrades, data center retrofits, or automation deployments go through the same procurement-dressed-as-capital-allocation process, the value erosion compounds. Assumptions about utilization rates, model performance, and workload patterns are far more volatile than lighting schedules, and the baseline against which you’d measure savings is moving constantly. A modernization framework designed for HVAC and LED projects is structurally unprepared for the assumption sensitivity of AI infrastructure investment. The five questions Kyunttsel proposes are sound governance, but they were built for a simpler asset class than the one now consuming the largest share of enterprise capital budgets.
The practical falsification condition for this framework is accountability assignment. If your capital approval process requires a named owner of the economic outcome before funds are released, and that owner is still reporting against the original business case twelve months post-commissioning, the drift Kyunttsel describes becomes visible while it’s still correctable. If your process doesn’t require that, the lighting retrofit story isn’t a cautionary tale from industrial facilities. It’s a preview of your next AI infrastructure post-mortem.
Based on reporting from Why modernization projects pass the payback test and still lose value, originally published 2026-07-21 09:13:00.
