← Back to Field Notes

September 5, 2025 · The Bloomfield Team

Why Your Best Employees Are Your Biggest Risk

A contract manufacturer in Michigan lost their senior estimator to a competitor in March. He had been with the company 22 years. Within six weeks, quoting turnaround doubled from 2 days to 4.5. Win rate dropped from 28% to 14%. By the end of the quarter, the company had lost an estimated $640,000 in revenue that would have converted under normal conditions.

The estimator who replaced him was competent. Experienced. He had 12 years in the industry. What he lacked was the 22 years of institutional memory about that specific operation: which customers negotiate hard, which tolerances on which geometries cause problems on their specific machines, which material suppliers quote high and which quote fair, which jobs from three years ago are the right comparables for the RFQ sitting on the desk right now.

That knowledge walked out the door.

The Concentration Problem

Every manufacturing operation has two or three people who hold a disproportionate share of operational knowledge. The estimator who has priced every job for 15 years. The setup machinist who knows exactly how the Mori runs on titanium. The quality manager who remembers every customer complaint, every corrective action, every audit finding back to the AS9100 certification in 2014.

These are the best people in the building. They are also, from an operational risk perspective, single points of failure. The Deloitte and Manufacturing Institute workforce study found that 2.1 million manufacturing jobs could go unfilled by 2030 due to the skills gap, and that estimate was made before the current wave of baby boomer retirements accelerated.

For a deeper look at what happens when that knowledge disappears, see our guide to manufacturing knowledge management.

What Actually Lives in Their Heads

Ask your best estimator how they price a job and they will walk you through a process that sounds simple. Pull the drawing, check the material, look at tolerances, estimate setup and cycle time, add overhead and margin. Straightforward.

Then watch them do it. The process they describe and the process they execute have almost nothing in common. What they actually do involves dozens of micro-decisions informed by pattern recognition built over years. They look at a drawing and immediately know that the customer's engineering team tends to call out surface finishes they do not actually need. They see a bore diameter and remember that a similar job two years ago required an extra operation because the depth-to-diameter ratio caused chatter on their Mazak. They know that this particular buyer responds to quotes over $20,000 by asking for a 12% discount, so they build that expectation into the initial price.

None of this exists in any system. It lives in context, experience, and judgment. When that person leaves, it does not transfer to a manual or a training document. The real cost of losing one experienced person extends far beyond the recruiting and onboarding expenses that HR tracks.

Why Traditional Documentation Fails

The standard response to key-person risk is documentation. Write it down. Build a procedures manual. Create training materials. These efforts are well-intentioned and almost always insufficient.

Documentation captures what people do. It rarely captures why they make specific decisions in specific contexts, which is where the operational value lives. A procedures manual can say "check historical job data when estimating setup time." It cannot replicate the estimator's ability to look at a drawing and know which of 4,000 past jobs is the right comparable, why it is the right comparable, and what adjustment factors apply based on the differences between that job and the current one.

The gap between documented process and real process is where most knowledge management efforts break down. The documentation exists. Nobody uses it because the actual decisions happen at a level of detail the documentation does not reach.

Building Systems Around Expertise

The alternative to documentation is capturing the data that experienced people use to make decisions, then building systems that deliver that data to the next person who needs to make the same decision.

When the senior estimator prices a job, the inputs they use are concrete: past job costs, material prices, setup times, cycle times, quality outcomes, customer history. Those inputs exist in the ERP, in job records, in emails, in quality logs. They are scattered across systems, but they are there. The experienced estimator has learned over 20 years where to find each piece and how to weight it. A well-built knowledge system does the same thing in seconds.

This does not replace judgment. The new estimator still needs to make the pricing decision. What it eliminates is the 20-year ramp time to learn where all the inputs are and how they connect. The system surfaces the three most comparable past jobs, the relevant quality history, the current material pricing, and the customer's order patterns. The estimator applies judgment to that information instead of spending their time assembling it.

The Risk Calculation

Take your three most critical knowledge holders. Estimate the revenue impact if each one gave two weeks notice tomorrow. Factor in the time to hire, the ramp-up period for a replacement, the productivity gap during transition, and the opportunities lost while the new person gets up to speed.

For most shops in the $8 to $25 million revenue range, the annualized risk exposure from key-person dependency ranges from $500,000 to $2 million. That is the budget available for building systems that capture and distribute institutional knowledge before it walks out the door.

The best time to build that system is while your best people are still in the building. Every day they work, they generate data. Every quote, every setup decision, every quality call creates a record. The question is whether that record goes into a system designed to make it accessible to the next person, or whether it stays locked inside one person's memory until the day they leave.