← Back to Field Notes

April 9, 2025 · The Bloomfield Team

The Estimator's Playbook: 20 Years of Quoting Lessons

In 2004, a 28-year-old estimator at a 40-person machine shop in central Ohio started his first day in the quoting office. He had spent six years on the floor running lathes and mills. The owner handed him a stack of RFQs, a calculator, and a binder of shop rates that had not been updated since 1998. Twenty years later, that estimator has quoted over 22,000 jobs, tracked his win rates across three different shops, and built a set of principles that consistently produce quotes with margins within 3% of actual job cost. Those principles have nothing to do with software and everything to do with how experienced estimators think about a part before they ever open a spreadsheet.

Price the Part You Are Going to Make, Not the Part on the Drawing

A drawing shows the finished part. The estimator needs to price the process of getting there, which includes every setup, every tool change, every handling step, and every thing that can go wrong based on the specific geometry, material, and tolerances in front of them. The drawing says nothing about the fixture that needs to be designed for that odd angle, the deburring that GD&T does not call out but the customer expects, or the first-article inspection that will add four hours to the first run.

Experienced estimators read a drawing and see the process. They mentally walk the part through the shop, machine by machine, and note every transition point where time, labor, or risk accumulates. A junior estimator reads the drawing and sees dimensions. The gap between those two perspectives is where quoting accuracy lives.

The Three-Job Rule

For any new part, find three comparable jobs from the shop's history. Not one. Three. One comparable job gives you a reference point that might be an outlier. Two gives you a range that might still miss the pattern. Three comparable jobs, with their actual costs and cycle times, define a corridor of expected performance that accounts for normal variation in setup efficiency, material conditions, and operator skill.

When the estimator in Ohio could not find three comparable jobs, he added 15% to his initial estimate as a risk buffer. Over 20 years, that buffer was too high roughly 60% of the time and too low roughly 10% of the time. The remaining 30% it was within 5% of actual cost. For the speed at which quotes need to move, that ratio was acceptable. The jobs where the buffer was too high still won at competitive rates. The jobs where it was too low were rare enough that the portfolio stayed profitable.

Setup Time Is Where Margins Die

Cycle time estimates are usually close because they are based on calculable feed rates, spindle speeds, and tool paths. Setup time estimates are where the money disappears. The standard setup time in the ERP says 2 hours for a three-jaw chuck job on the Mazak. The actual setup takes 3.5 hours because the operator had to hunt for the right boring bar, the program needed a revision to account for a material size change, and the first part came out 0.003" over tolerance requiring a tool offset adjustment and a second attempt.

The 20-year playbook says: quote setup time based on the median of the last five actual setups for comparable work, then add 20% for any job the shop has not run before. That 20% covers the surprises that experienced estimators know will happen on new work even when the process looks straightforward. On repeat jobs, drop the buffer to 10% because familiarity reduces variability but never eliminates it.

Know Your Customer's Tolerance for Price

After quoting a customer 50 times, patterns emerge. This buyer always pushes back on the first quote. That buyer accepts the price if the lead time is under three weeks. The aerospace buyer never challenges the price on AS9100-certified work but will negotiate aggressively on commercial parts. The medical device company pays a premium for documentation and traceability but expects the pricing to reflect that they are providing 18-month blanket orders with guaranteed volumes.

The playbook says: maintain a one-page profile for every active customer that includes their typical pushback patterns, their sensitivity to lead time versus price, their payment terms reliability, and the margin range the shop has historically achieved on their work. That profile informs pricing strategy in ways that a cost-plus calculation never can.

Quote the Job You Want at the Price You Need

Not every RFQ deserves a competitive price. Some jobs fill capacity during slow months and justify thinner margins. Some jobs require capabilities the shop wants to develop, and the learning cost should be partially absorbed as an investment. Some jobs come from customers who pay late, change specs after ordering, and demand requotes every time, and those jobs should be priced at a premium that compensates for the administrative burden or priced high enough that losing the bid is acceptable.

The 20-year view is that a shop's quoting strategy should be as deliberate as its machining strategy. Every quote carries an intent beyond winning the job. The intent might be to grow a customer relationship, to fill a capacity gap, to enter a new market, or to maintain margin discipline during a busy period. Quoting without a strategic intent produces a random portfolio of jobs that individually make sense and collectively create scheduling chaos, margin compression, and customer concentration risk.

For a deeper look at quoting strategy and systems, see our complete guide to AI-powered quoting.

Document Everything, Even When Nobody Asks

The estimator in Ohio kept a personal notebook for 20 years. Every unusual part, every setup surprise, every material issue, every customer complaint that traced back to a quoting assumption. That notebook contained the institutional knowledge that made his quotes consistently accurate. When he moved to his second shop, the notebook went with him and the knowledge at his previous employer walked out the door the same day.

Capturing that knowledge in a system rather than a personal notebook is the single most valuable thing a shop can do for its quoting operation. Not because the estimator is going to leave tomorrow, though statistically half of all manufacturing estimators are over 55 and within a decade of retirement. Because the knowledge becomes compounding when it is shared across the team, searchable by future estimators, and available to AI systems that can surface it at the moment of need.

Twenty years of quoting wisdom should not live in one person's head. It should be the foundation of every quote the shop produces going forward.