Most workers’ comp programs fail for a reason no one wants to put in a board deck.
When a safety initiative does not produce results, the post-mortem usually lands in one of two places. Either the workforce is “resistant to change,” or the supervisors are not enforcing it well enough. Both are wrong. Both are also expensive.
I had this conversation on Frontline Advantage with Piyush Mathur, who spent 21 years at Nielsen and built people analytics for 140,000 employees at Johnson & Johnson before founding SkillsBridge.AI. He gave the cleanest explanation of frontline behavior I have heard in years.
Picture a delivery driver. You hand him a tool that adds one minute to every stop. He stops using it.
The standard interpretation: he is resistant to change. The accurate interpretation: he wants to get home for dinner.
Frontline workers are not stubborn. They are economic actors with finite time, finite energy, and a clear understanding of what gets them paid and what gets them home. When a tool adds friction without paying them back, they route around it. When a process slows them down without making their work safer, they ignore it. When a feedback channel produces nothing, they stop using it.
The Friction Tax
Every safety program has a friction tax. The cost is invisible until you look for it.
You see it in the toolbox talk that the supervisor reads off a card because the system requires it, while the line operator nods and goes back to running the equipment one notch faster than the SOP allows. You see it in the JSA that gets signed without being read because reading it would add four minutes to a shift. You see it in the near-miss reporting form that no one fills out because the form takes longer than the near-miss did.
Each of those moments is a signal that disappeared. Each disappeared signal is a future incident that was preventable but is now in your loss runs.
The standard response from safety leadership is to add more. More forms. More training. More compliance audits. More dashboards. The thinking is that if the program is not working, the program is not loud enough.
This is the exact opposite of what works.
What Actually Closes the Loop
The fix is not louder. The fix is faster.
A frontline tool succeeds when it pays the worker back inside the same shift. That means the report is fast, the response is visible, and the feedback loop closes before the worker forgets they reported anything. Anything else trains the workforce to stop reporting.
This is what Teamforce AI is built to do.
We deploy anonymous tablet kiosks at the point of work, not on the corporate intranet, not in an app no one downloads, not in an email no one opens. Workers report hazards, near-misses, and operational risks in under sixty seconds. Operations leadership sees the signal in real time. The corrective action gets verified by the workforce itself, not just signed off by a manager. And the loop closes publicly, where the next worker can see that the last one was heard.
That is not engagement theater. That is risk reduction with a financial signature attached.
The Operating Question
Every self-insured manufacturer should ask one question of every safety, EHS, or engagement initiative on the books:
Does this tool pay the worker back inside the shift?
If the answer is no, the tool will produce signals that disappear. Disappeared signals become incidents. Incidents become claims. Claims become the line on the P&L that the CFO cannot explain to the board.
The workers are not the problem. They are telling you exactly what is wrong every shift. The system that captures the signal is the problem. Fix that, and the loss runs follow.
Watch the full conversation with Piyush Mathur, Dan Riley of RADICL, and Brad Federman of Performance Point on Frontline Advantage.
See the EBITDA Impact Analyzer for self-insured manufacturers.
Note: This post was originally published on April 29, 2026, and last updated on May 14, 2026.