← Back to Field Notes

March 10, 2025 · The Bloomfield Team

Setup Time: The Most Underestimated Cost in Manufacturing

Setup time accounts for 15% to 30% of total available machine hours in a typical job shop running high-mix, low-volume work. That range comes from data we have seen across dozens of operations, and it aligns with what the National Institute of Standards and Technology has published on small manufacturer efficiency. On a machine with a fully burdened rate of $125 per hour running 2,000 hours per year, 20% setup time represents $50,000 in non-cutting time on a single machine. Multiply that across a shop with 15 CNC machines and the number is $750,000 annually in time where spindles are stopped.

The money is real. What makes it worse is that most shops underestimate setup in their quotes by 20% to 40%, which means they are absorbing the cost twice: once in the actual production time and again in the margin erosion from quoting it wrong.

Where the Estimate Goes Wrong

Estimators quote setup time based on what the setup should take. Operators experience what the setup actually takes. The gap between those two numbers is where margins disappear.

A setup estimated at 90 minutes on a horizontal machining center might actually require 150 minutes when you account for the full sequence. The operator pulls the previous job's fixtures. They locate the tooling for the new job, which may or may not be staged at the machine. They load tools, set offsets, verify the program, run a first piece, inspect it, make adjustments, and run a second piece for confirmation. Each step adds minutes that the estimate did not capture because the estimate was based on the setup itself, not the setup plus all the work that surrounds it.

And that 150 minutes assumes nothing goes wrong. A missing tool, a worn insert that needs replacement, a drawing revision that nobody flagged, a fixture that does not match the setup sheet from last time, all of these add time in increments of 15 to 45 minutes per interruption.

The Operator Variance Problem

Two operators setting up the same job on the same machine can produce setup times that differ by a factor of two. One operator ran this part six months ago, remembers the fixturing approach, knows the offset values are close to what they need, and has the muscle memory of the process. The other operator is running the job for the first time, working from a setup sheet that may or may not reflect how the part was actually produced last time.

That variance rarely appears in the quote. The estimator uses a single setup time that represents an average case, which means half the jobs come in over estimate and the other half come in under. On a low-margin job, the overs eat the profit entirely.

For a broader look at how scattered information drives cost overruns, see our guide to production visibility in manufacturing.

Tracking Setup Time as a Separate Line Item

The fix starts with measurement. Most ERP systems allow setup time to be tracked independently from run time, but fewer than half the shops we talk to actually use that field consistently. When setup time is bundled into total job hours, the data that would reveal the problem never surfaces.

Break it out. Track estimated setup versus actual setup on every job. After 60 to 90 days, you will have enough data to see the patterns. Which machines have the highest setup time variance. Which part families consistently exceed their estimates. Which operators complete setups fastest and what they do differently.

That data becomes the foundation for three improvements: better quoting, better scheduling, and better training.

Reducing Setup Through Standardization

The shops that have driven setup time down by 30% or more share a common approach. They standardize the repeatable elements and document the job-specific variables.

Standardized tooling kits for common part families eliminate search time. If a shop runs 200 different part numbers but 80% of them use the same 15 tool assemblies in various combinations, prebuilding those assemblies and storing them at the machine or in a tool crib with clear identification saves 20 to 30 minutes per setup.

Setup sheets that include photographs, offset values from the last successful run, and operator notes on what to watch for transform a new-operator setup from a 3-hour problem-solving exercise into a 90-minute documented procedure. The knowledge that used to live in one person's head becomes available to everyone on the team.

Quick-change fixturing systems represent a capital investment, but the payback on high-frequency setups is typically under 12 months. A shop doing 6 setups per day on a single VMC that reduces average setup from 2 hours to 45 minutes recovers 7.5 hours of spindle time per day on that machine alone.

The Quoting Connection

Every improvement in setup time accuracy flows directly into quoting accuracy. When your estimator can pull actual setup data from the last three runs of a similar part, differentiated by operator and machine, the quote reflects reality instead of an optimistic assumption.

That is the difference between a quote that wins at 28% margin and delivers 28%, and a quote that wins at 28% margin and delivers 19% because setup ran 40% over estimate. Over the course of a year, accurate setup data in the quoting process can recover 5 to 10 points of margin on setup-heavy work.

For more on how quoting accuracy connects to operational data, see our complete guide to AI-powered quoting in manufacturing.

The Scheduling Multiplier

Setup time accuracy also transforms scheduling. When the scheduler knows that a job requires 2.5 hours of setup based on actual data rather than a 1.5-hour estimate based on optimism, the schedule reflects what will actually happen on the floor. Jobs stop stacking up behind machines that are still in setup. Due dates stop slipping by a day here and there.

The shops that schedule by gut feel are usually shops where setup time estimates are unreliable. Fix the setup data and the scheduling problem gets smaller by itself.

Setup time is the most underestimated cost in job shop manufacturing because it sits in the gap between what the quote assumes and what the floor experiences. Closing that gap requires measurement, standardization, and a system that connects floor data to the quoting and scheduling processes that depend on it. The data already exists in most operations. The work is capturing it and putting it where it can do the most good.