← Back to Field Notes

March 19, 2025 · The Bloomfield Team

The Gap Between Quoted and Actual Costs in Manufacturing

The average job shop operates with a 12% to 18% gap between what they quote a job at and what the job actually costs to produce. On an $8 million annual revenue shop, that gap represents $960,000 to $1.44 million in margin that either evaporated on underquoted jobs or was left on the table on overquoted jobs that they never won. Most shops know the gap exists. Few have measured it precisely, and fewer still have a system for closing it.

Where the Gap Lives

The cost variance between quote and actuals breaks down into five categories, and each one has a different root cause.

Setup time variance is the largest contributor, accounting for roughly 35% of total cost gap. Estimators quote setup based on what it should take. The floor experiences what it actually takes. The difference runs 20% to 60% on complex setups, especially when the operator running the job is different from the person the estimator had in mind when building the quote.

Cycle time variance contributes about 22% of the gap. Quoted cycle times come from programming software estimates or from feeds and speeds calculations that assume ideal conditions. Actual cycle times include tool changes, in-process inspection pauses, program stops for chip clearing, and the difference between theoretical and actual feed rates when cutting real material with real tooling wear patterns.

Material cost variance represents 20% of the gap. Material prices shift between quote date and order date. An estimator quoting 6061-T6 aluminum at $3.20 per pound in January might buy it at $3.85 per pound in March. On a job consuming 500 pounds of material, that is a $325 cost increase that came directly out of margin.

Scrap and rework account for 15%. No estimator quotes a 5% scrap rate unless the customer requires it. But scrap happens. A first-article failure that requires two additional setups and three more test pieces adds hours that never appeared in the quote. The cost of those hours is real and unrecoverable on a fixed-price job.

Outside services contribute the remaining 8%. Heat treat expedite charges, plating rejects requiring re-processing, and shipping costs on outside operations that were underestimated all erode the quoted margin.

Why Most Shops Cannot Close the Loop

Closing the gap between quoted and actual costs requires comparing the two numbers on every completed job and analyzing the patterns. The data for this analysis exists in most operations. Quoted costs sit in the estimating system. Actual costs sit in the ERP. The problem is that the two systems rarely talk to each other in a way that produces an automatic comparison.

In practice, connecting quoted estimates to actual outcomes requires someone to manually pull the numbers, match them by job number, and build the comparison. That work takes time that nobody has. So it does not get done, or it gets done once a year during an annual review, which is too infrequent to drive real improvement.

The shops that close this loop use their ERP data to build automated quote-to-actual comparisons that run on every completed job. When the estimator can see, on their dashboard, that their last 50 setup time estimates averaged 28% below actual on five-axis work, they adjust. When they can see that material cost estimates on stainless jobs from Q4 ran 12% below purchase price, they update their pricing sources.

The Feedback Loop That Fixes Quoting

Accurate quoting is a learning system. Every completed job produces data that should improve the next quote on a similar part. The feedback loop works in three stages.

First, the actual cost data flows from the shop floor back to the estimating function. Hours by operation, material consumed, scrap produced, outside service charges incurred. All of it tagged to the original quote.

Second, the variance is calculated and categorized. Was the setup under or over? By how much? On which machine? With which operator? Was cycle time off because the quoted speeds were optimistic, or because the job required an unplanned additional operation?

Third, the estimator uses the pattern to calibrate future quotes. A shop that has tracked quote-to-actual data across 500 jobs has a pricing database that no amount of experience alone can replicate. That database tells the estimator exactly how much to adjust for a five-axis setup on Inconel versus a three-axis setup on 6061, based on what actually happened the last dozen times the shop ran similar work.

For more on building this kind of data-driven quoting process, see our complete guide to AI-powered quoting in manufacturing.

The Margin Recovery Math

A shop doing $10 million in annual revenue with an average 15% gap between quoted and actual costs is losing approximately $750,000 per year in margin leakage on underquoted jobs. Not every dollar of that gap is recoverable. Some variance is inherent in custom manufacturing. But cutting the gap from 15% to 8% recovers $350,000 in annual margin without winning a single additional job.

That recovery comes from three improvements. Estimators quote setup time more accurately because they have actual data from similar jobs. Material costs reflect current pricing rather than stale spreadsheet data. And known cost drivers like tight-tolerance features, difficult materials, and complex fixturing get priced appropriately because the data shows exactly what those features cost the last time the shop encountered them.

What This Requires

Building the quote-to-actual feedback loop requires three things: time tracking on the floor that captures hours by operation and job, a comparison mechanism that matches actual costs to quoted costs automatically, and a format that presents the analysis to the estimator in a way they can use in real time.

The first piece, time tracking, already exists in most ERP systems. The data quality varies, but the infrastructure is there. The second piece, the comparison, is where most shops stall because it requires connecting the estimating system to the ERP in a way the software vendors did not build. The third piece, the estimator-facing analysis, is where the real value lives.

An estimator opening a new RFQ for a titanium bracket should see, before they start building the quote, how the last five titanium bracket jobs performed against their quotes. Setup time actual versus estimated. Cycle time actual versus estimated. Material cost actual versus quoted. That context turns every quote into a data-informed decision.

The gap between quoted and actual costs in manufacturing is measurable, traceable, and closable. The data to close it already lives in your operation. The work is connecting it.