You find out a job lost money a month later
The parts shipped in March. The margin arrived in April, as a surprise.
The lesson, stated plainly
3 things that decide this
- 01Weever's manufacturing research finds about half of manufacturers still lean on spreadsheets or paper alongside their main system, which is where job costing goes to lag.
- 02When labor and material land in the books at month-end, quoting corrects a month behind reality, and the same losing job gets quoted again.
- 03The fix is not a 12-month ERP project. It is live capture on the floor feeding the quoting spreadsheet you already trust.
The traveler knows the truth and tells nobody
A job travels the shop as a paper packet. Operators write times and quantities on it, mostly accurately, mostly at the end of a shift. The packet returns to the office when the job ships, and someone keys it into a spreadsheet, and eventually the accounting file. Only then does anyone learn what the job cost.
By that time the estimate that priced it is stale, the machine time it consumed is unrecoverable, and the customer who got a bargain is asking for the same price on the next order. Weever's research on factory operations puts roughly half of manufacturers in this position: a main system of record, with paper and spreadsheets doing the real work beside it.
Why the ERP did not fix this
Many shops bought software for exactly this problem and got a different problem. Fabrication-industry writers at Eziil, echoing years of forum threads, describe metal fab plus monolithic ERP as where 12-month implementations go to die: modules for everything, configured for nobody, with consultants billing through the second year.
The shop's response is rational. People route around the system that slows them down, back to paper and Excel. The ERP becomes an expensive ledger while the floor runs on the packet, and job costing stays a month behind.
- 01Floor data enters systems at shift-end or ship-time, never as work happens.
- 02Quoting reads last quarter's costs because that is the freshest number available.
- 03Repeat work inherits the errors: the underquoted job becomes the underquoted contract.
Capture at the machine, cost while the job runs
The workable version is small. Operators log start, stop, and quantity on a screen at the machine, in seconds. Material draws scan against the job. A dashboard shows cost against estimate while the job is still on the floor, so a job drifting over can be caught, repriced, or at least understood before the next quote goes out.
This is deliberately not an ERP replacement. It is the layer between your spreadsheets and the ERP you do not want to buy, connected to whatever already runs accounting. We build operations systems in exactly this shape: WAIQ, a business operations platform we shipped, turns scattered operational data into execution its teams act on daily.
Operations data, made actionable
What owners and ops managers ask
01Will operators actually use screens on the floor?+
Yes, when the interaction takes seconds and replaces writing, not adds to it. The tools that fail ask operators to serve the software. A start-stop-quantity tap at the machine survives because it is faster than the pen it replaced.
02Do we have to abandon our spreadsheets?+
No, and you should not at first. The spreadsheet embodies how your shop actually quotes. The right first step feeds it live numbers instead of month-old ones. Retiring it later is an option, not a prerequisite.
03How is this different from buying a job shop MES?+
Products like production tracking suites are worth evaluating, and for standard workflows they can fit. The gap shows when your routing, pricing, or customer requirements do not match the product's assumptions. Custom capture built around your actual travelers avoids forcing the shop to change how it works.
