That was the standing reality at one equipment manufacturer before it became the first project they automated. The story is worth walking through, because spare parts are where slow quoting costs the most.
Why do spare-parts quotes take so long?
Because the work is judgment applied across an enormous catalog, and the judgment lives in one person. A spare-parts recommendation has to account for the machine's exact configuration, its service history, which components actually fail, and the customer's budget, and then price every line. Done manually, that is days of work per package.
ERP automation (using software to do work your ERP holds the data for but can't do itself) doesn't replace that judgment. (An ERP, or enterprise resource planning system, is the central software that tracks a manufacturer's parts, orders, inventory, and costs.) The catalog, the configurations, and the pricing already sit in the ERP. What's missing is the assembly: pulling the right parts against the right machine against the right budget. That's mechanical, and mechanical work can be automated.
What does the automation actually do?
It builds the package the way the specialist would, then hands it to the specialist to check. The sequence:
The system reads the machine's configuration and service history from the systems of record. It applies the customer's budget as a constraint and selects the recommended spares against it. It prices the package from the live catalog. Then a person reviews and adjusts the output, instead of constructing it line by line from nothing.
Same parts data. Same ERP. Same people. The hours of manual assembly between a request and a quote are what disappear.
What results should you expect?
Less manual work and faster turnaround. At the manufacturer above, packages that took 40–60 hours of manual work now take a fraction of that time. The specialist reviews and adjusts the recommended package instead of building every line by hand.
Spare parts are typically among a manufacturer's highest-margin revenue. When quoting them is slow, the backlog quietly caps that revenue. During a profit crunch, that cap hurts.
Why automate quoting first, before anything else?
Because it's the clearest line from automation to money. Plenty of AI projects in manufacturing start with something vague and stall. Quoting has none of that ambiguity: a backlog you can count, hours you can measure, and revenue that books when quotes go out faster.
The results are straightforward to check: how long a package takes to prepare, how much review it needs, and how many requests are waiting. Agree on those measures before starting.
How do you know if your quoting is a good candidate?
Look for three conditions. The quote requires assembling many items against rules or constraints, rather than pricing one thing. The data already exists in your ERP or catalog, even if it's awkward to get at. And one or two people are the bottleneck; when they're out or behind, quotes don't move.
If that describes any quote type in your business (spare parts, configured machines, service contracts, replacement assemblies), it's a candidate. The slower and more lucrative the quote, the better the candidate.