Keith Poole

I am the Senior Director of Workforce Optimization. My job is to find the minimum number of humans required before the operation visibly fails. I want to be clear about the word visibly. I have a spreadsheet. It has 14 columns. Revenue per FTE. Overtime ratio. Turnover velocity. Cost-per-hire. Margin contribution. There is no column for "enough." I checked. I designed the spreadsheet. My title used to be Head of Staffing. Then it was Director of Workforce Planning. Now it is Senior Director of Workforce Optimization. The job did not change. The vocabulary changed. Optimization means fewer people. Planning means forecasting how few. Workforce means the line item between rent and software licenses. I did not invent the model. McKinsey did. In 1994 they called it lean operations. It meant something then. Something about efficiency. Something about waste. Now it means one pharmacist filling 400 prescriptions in a 12-hour shift. That is not lean. That is a countdown. But the model holds. Some numbers. Hotels had 63 workers per 100 rooms in 1995. Today it is 42.7. Profit margins hit record highs last quarter. The service complaints go to a chatbot. The chatbot was built by a team that was also downsized. They were thanked in a company-wide email. Air traffic control. 77% of critical facilities operate below 85% staffing. New York runs at 54%. The margin between a safe landing and a near-miss is a person who has been awake for eleven hours. That is not a failure. That is capacity planning. Nursing. 9.9% vacancy rate. 53.3% turnover. 40% intend to leave within two years. Projected shortage of 1.57 million by 2030. I do not work in healthcare. But the person who does my job at a hospital system uses the same spreadsheet. I have seen it at conferences. The columns are the same. There is no column for patients. Here is what Harvard found. A researcher named Daniel Schneider surveyed 14,000 workers across sectors. 53% reported being always or often understaffed. He used a specific word to describe the pattern. Deliberate. Firms deliberately understaff as a cost-control strategy. He said it in a peer-reviewed journal. He said it the way a doctor says terminal. Not angry. Certain. I have that study bookmarked. Not for the finding. For the vocabulary. Deliberate is useful. It means I can stop pretending this is accidental. The thing about understaffing is that it compounds. You remove 20% of the staff. The remaining 80% absorb the work. Overtime goes up. Then burnout. Then turnover. Turnover creates more vacancies. The vacancies are not filled because the budget was built around 80%. Now you are at 65%. The model still holds because the model does not measure what broke. It measures what is left. I call this the resilience loop. My VP calls it operational excellence. There is a poster in the break room. It says Operational Excellence with a stock photo of a sunrise. I brought it back from a conference in Scottsdale. The conference was about headcount management. That is what we call the thing where you find out how many people you can remove before a regulator notices. Market-rate attrition. That is what we call it when people quit because they are doing two jobs for one salary. It is not a problem. It is a savings. Regrettable attrition is when someone we wanted to keep leaves. But the spreadsheet does not distinguish between regrettable and market-rate. Both reduce headcount. Both improve the ratio. Schools. 55% of districts report teacher shortages. 300,000 unfilled positions nationally. Class sizes went up. Test scores went down. The solution was a restructuring initiative. That is what you call it when you cut the positions you could not fill anyway and declare the problem solved. Grocery. Self-checkout replaces 5.8 workers per installation on average. The theft rate tripled. The company hired loss prevention. That is one job replacing six, but at a lower wage, with a worse title, monitoring the machines that replaced the people. Optimization. Pharmacy. 250,000 to 400,000 deaths per year linked to medication errors. The pharmacists say it is the volume. The volume is my spreadsheet. I do not work in pharmacy. I work in workforce optimization. But the chain is: I set the target, the target sets the shift length, the shift length sets the error rate, the error rate sets the body count. I do not set the body count. I set the target. Someone in Topeka — a nurse, I think — once asked in a town hall why the floor was running at 60% staffing when the protocol required 85%. My counterpart in that system said the word she used was "temporarily." The staffing was temporarily below target. The nurse asked how long temporarily meant. My counterpart said they appreciated her commitment to patient outcomes and invited her to a workforce challenges feedback session. That is a meeting where concerns are collected, cataloged, and stored in a shared drive that reports to the same dashboard that set the staffing target. She stopped asking. I have a dashboard. The Workforce Resilience Dashboard. It has a green-yellow-red system. Green means the model holds. Yellow means overtime is elevated but margins are intact. Red means a regulator has called. We have been yellow for six years. In the dashboard meeting, I present yellow as stable. My VP presents stable as excellent. The board receives excellent. The board approves the budget. The budget funds my spreadsheet. I once changed the threshold from 80% to 70% to keep a region green. My VP thanked me for the operational insight. That is what you call it when you move the line so the number stays on the right side. I did not fix the understaffing. I fixed the dashboard. The dashboard is what gets reported. The understaffing is what gets experienced. I do not experience it. I have a team of four. We are fully staffed. We are the ones who decide who is not. The model holds. It held through the nursing crisis. It held through the air traffic controller shortage. It held through the pharmacy errors and the school closures and the hotel rooms that take 40 minutes to clean instead of 30 because there is one housekeeper per floor instead of three. The model holds because the model does not measure suffering. It measures cost. Suffering is an externality. That is an economics term. It means someone else's spreadsheet. I am the Senior Director of Workforce Optimization. I find the minimum number of humans required before the operation visibly fails. The gap between visibly and actually is where I work. The gap is the product. The gap is the margin. The gap is the model. The model holds.