Job shop profitability: know which jobs make money

The problem

In a job shop, every job is a little different, so averages are especially misleading. A few large or complex jobs can carry the month while dozens of small ones quietly lose money, and the estimates behind them keep getting reused.

What to measure

Job-level margin vs estimate, customer contribution after erosion, and the categories (usually labor and setup in job shops) where estimates miss most often.

How MarginGuard helps

Export a job-cost report from your ERP, upload it, and MarginGuard ranks every job and customer by margin dollars, shows what drove each miss, and keeps a list of fixes your team can work through weekly.

Next step: sign up, load the demo, or compare plans.

How to get started

1. Step 1
   Export last quarter's jobs

   From JobBOSS², E2, ProShop, or any ERP.

2. Step 2
   Upload

   Mapping and readiness check in a few minutes.

3. Step 3
   Review

   Top losing jobs, customer rollups, and category breakdown.

Typical job-shop findings

Setup underestimated on short runs, rush and expedite work not priced in, a few customers whose engineering changes erode margin, and material price increases between quote and purchase.

Why averages mislead in high-mix shops

A shop can report a healthy 25% gross margin while a third of its jobs earn under 10% and a handful lose money. The profitable repeat work and a few large jobs carry the average. Because each job is different, the losers don't look alike at first glance; they only become visible when you group them by customer, part family, or cost category and rank by dollars.

A realistic first review

Import last quarter's closed jobs. Look at total erosion, then the ten jobs that lost the most dollars against estimate. For each, note the category that drove it. Usually two or three patterns cover most of the erosion: a setup standard, a customer, or material pricing. Pick one, assign an owner, and check the next month's jobs. That single loop is worth more than a perfect dashboard nobody reviews.