I remember the day clearly. A 55-person metal fabrication shop, run by two brothers who really knew how to weld, was staring at a job they thought was their best one. The CFO came to the weekly meeting with a printed job card, looked up, and said: “We billed 80 grand for this job, and I still can’t tell you if we made or lost money on it.” That’s the moment ERPNext Consulting gets invited into the room.
The problem wasn’t the workers, it was the data trail
The shop had a fine reputation in the town, but its costing was a disaster. Materials were pulled from the rack without any record, and the welders’ hours were marked down on paper at the end of the shift from memory. A job that looked profitable on the quote sheet was actually bleeding 11% of its value once every bolt, gas refill, and overtime hour was counted.
The fix started with three small habits, not a big dramatic rollout. Every material pull got logged against a job number. Every shift clocked labor hours into the same record. And scrap metal was weighed and tagged at the bin. That gave the brothers their first real picture of job-level profit in years. ERPNext Consulting has worked with plenty of shops like this, and the pattern is always the same: the data already exists inside the business, it’s just scattered across memory, notebooks, and spreadsheets.
What changed when the numbers finally lined up
- Quotes started from actual historical costs, not gut feel, which lifted the average profit margin on new work by 6.4 points within two quarters.
- The purchasing manager could see which steel grades were turning into scrap, so he stopped over-ordering the soft sheet stock that the painters kept rejecting.
- The owners finally had a weekly one-page report showing every open job, its costs to date, and its remaining budget. A meeting they used to avoid became the shortest one on the calendar.
Was it tedious in the first few weeks? Yes. The welders complained, the office manager complained, and one foreman swore he would quit rather than type a digital entry.
But the brothers held the line for a month, and the grumbling faded. The foreman who threatened to quit ended up being the one who showed new hires how to use the screens.
The real win appeared six months later
They took on a large, awkward architectural job that had gone through three quoting rounds already. The CFO opened the historical cost data on a similar job from the previous year, saw the hidden corners where that earlier job had lost money, and re-priced the new one accordingly. That single decision recovered more than the implementation fee.
If you run a shop that quotes jobs without knowing your true costs, do this tomorrow: pick the three biggest jobs from last year and reconstruct what they really cost you. If you can’t do it in an afternoon, that’s your first signal. The goal is not to get a nicer report, it’s to get a number you can trust before you promise a price to a customer.