This week alone gives you four industries proving it.
HBR is warning that most AI strategies are built on urgency, not readiness. Tennant just disclosed an ERP rollout that cost them roughly $30M in sales and dropped their stock 23%. PitchBook is flagging a record exit overhang in private equity, with megafunds masking a backlog of stuck portfolios. New research on AI adoption says pilots work fine, they just die at scale, blocked by tribal knowledge nobody documented.
Four industries. Four headlines. One problem.
Every one of them is a business that deployed something — a model, a system, a fund strategy — on top of a human layer that was never built. Nobody trained the people. Nobody mapped who actually holds the process. Nobody did the unglamorous work of making the organization ready before the technology arrived.
So the AI stalls. So the ERP breaks live production. So the portfolio company can't exit, because scale was never the same thing as health.
None of this is a technology failure. It's the same failure wearing a different badge. You can call it governance. You can call it change management. I call it the human layer — the work that happens before the software shows up, that determines whether the software ever gets to deliver anything at all except the bill.
The tools keep changing. The failure mode hasn't moved in decades.
Dominic Luzi · Chief Answer Grape, Management Alchemy