Once you know to look for it, you'll see it on any job board. A company posts for an "Inventory Control Specialist." Six months later the same posting is back. A few months after that it's up again, with a slightly better salary and a new line about a fast-paced environment.
By the third posting, hiring isn't the problem. The company is trying to staff its way around a broken system.
The loop goes like this. Nobody trusts the inventory numbers. The ERP says one thing, the shelf says another, and complaints about stockouts and wrong shipments keep climbing. So leadership does the sensible-sounding thing and hires someone to own it.
That person walks into a job that was misdescribed from day one. The posting said "manage inventory accuracy." The real job is reconciling two versions of reality by hand, every day. They chase one discrepancy at a time, keep side spreadsheets because the ERP isn't trusted, run manual counts a working cycle count program would make unnecessary, and referee between sales, who sold what the system showed, and the warehouse, which can't find it.
None of that fixes a cause. The role was built to absorb symptoms, so that's all it can do. After twelve to eighteen months the person burns out and leaves, or gets quietly managed out for "not getting inventory under control." The posting goes back up, and a new name walks into the same trap.
The company thinks it has a turnover problem. It has a systems problem wearing a turnover costume. You can tell which with one question: if this person did the job perfectly, would the causes go away? A real inventory lead would build a cycle count program and fix the location logic, and the role would leave something behind. The reconciliation version just mops faster. When the mopping stops, the floor floods again, and no candidate can win that job.
Do the math on it. A fully loaded inventory specialist runs somewhere between $55K and $85K a year in most markets, before the cost of rehiring and retraining every eighteen months. Over five years that's a quarter of a million dollars or more, and it doesn't belong in your labor budget. You're paying a subscription for a human patch on a process problem, and the patch keeps quitting. Next to that, a serious process overhaul looks cheap.
Don't swing to the other ditch, though. Plenty of companies spot this pattern and buy their way into the same problem with new software and better branding. Sometimes new software is the answer. Just as often the fix is the process around the software you already own: a real cycle count habit, logged mispicks so patterns show up, closed sync gaps, and the veteran's knowledge written down where anyone can run it. You only find out which by looking at the floor before anyone shows you a demo.
Before you post that role a third time, pull up the last two versions and read them like an X-ray. "Reconcile discrepancies" means your data drifts daily. "Improve count accuracy" means nobody trusts the numbers. "Coordinate between sales and warehouse" means your systems don't. The posting already describes what's broken. Fix that, and the next person you hire might stay.