CRM for manufacturers and machine shops
In custom manufacturing the job usually goes to the first credible quote, not the lowest. Shops lose work they never knew they were competing for, because the RFQ sat behind whoever was estimating and the cost model lives in that person's head.
Where the generic CRM breaks:
RFQs arrive as email. Unstructured, unqueued and invisible, so nobody can see how many are waiting or how long they have waited.
No part concept. A repeat part re-quoted from scratch loses the most accurate estimate available, which is what it actually cost last time.
Quoting depends on one estimator. When the cost model is judgment rather than data, two estimators quote the same job differently and neither can be checked.
Job costing arrives after shipment. Knowing a job lost money once it has shipped is history, not management.
The data model that actually fits:
RFQ. Structured intake with drawings, quantity and required date, queued and visibly ageing so turnaround becomes a measurable number.
Part. The thing you actually make, carrying its quote and cost history, so repeats are priced from evidence.
Cost model. Material, setup, run time and margin as maintained data rather than estimator judgment, which makes quoting fast and consistent.
Job with live cost. Actual hours and material against quoted, visible during the job rather than at shipment.
Our verdict: Most shops under fifty people do not need full MRP and use a fraction of it when they buy it. Build quoting, part history and job costing. Quote turnaround alone changes win rate, and it is the cheapest of the three.
Why does quote turnaround matter so much?
Because buyers frequently award to the first credible quote. Shops that start measuring turnaround are usually surprised by how many RFQs they lost without ever being compared on price.
Can quoting be automated for custom parts?
Partially. Material and standard operations calculate reliably; unusual setups and tight tolerances need an estimator. The goal is getting the routine majority to a draft the estimator adjusts, not removing them.
Is live job costing realistic in a small shop?
It is, once floor time is captured against the job rather than the day. That is a floor discipline change as much as a software one, and it has to be solved alongside the build.
Do we need a full ERP?
Rarely at this size. Quoting, job costing and scheduling pay for themselves; the rest can wait until volume genuinely demands it.
Last reviewed 22 August 2026