Should quoting live inside your ERP or beside it

The ERP salesperson has a good line and it is worth taking seriously. You already own the rates, the routings and the material costs, so why would you buy a second system to hold copies of all 3 and then spend a year keeping them in step.

That argument is correct about the data and wrong about the work. The ERP quoting module is genuinely the right answer for a large share of what arrives, and genuinely unable to handle the rest, and the useful question is which share is which on your floor. The buyer’s guide frames the wider choice this sits inside.

What an ERP quoting module actually does

It prices from the master data you already maintain, which is exactly what you want when the part exists. Your work center rates are current because production uses them. Your material costs are current because purchasing uses them. Nothing has to be exported and nothing drifts.

For a release against a part you have run 40 times, this is not just adequate, it is better than any external tool could manage. The ERP knows the standard cost, the last actual cost, the current material price and what the last order of 5,000 did on the floor.

An estimator answering that request outside the ERP is rebuilding an answer the ERP already has. Anyone shopping for quoting software before exhausting this option is about to buy a second copy of something they own, and that is a conversation worth having internally before any demo gets booked.

The modules have also improved. Most mid-market ERPs now carry a quote object that can hold a bill of materials and a routing before either is released, which was the obvious gap 10 years ago and is no longer the whole story.

There is a cost argument as well and it is not small. The module is usually either included or a modest add-on against a licence you already pay, which puts it an order of magnitude below anything else on the market. On a floor where quoting is slow rather than broken, that difference buys a machine.

The one thing to check before assuming it is free is user count. Some vendors licence quoting per named user, and if your sales head and 2 estimators all need access, the modest add-on stops being modest.

The part number that does not exist yet

A quote is the exercise you run before the part number, the routing and the bill of materials exist, and the ERP is built on the assumption that all 3 do. That is the structural problem and no configuration fixes it.

Most modules deal with it by making you create the part first. You invent a number, attach a provisional routing, build a bill of materials from a drawing, and price against it. If you win, the record was useful. If you lose, and you lose most of them, it stays.

Most quotes never become parts.

A floor quoting 120 requests a month and winning 30 percent creates several hundred dead part numbers a year. The item master fills with things that were never made, planners stop trusting search results, and somebody eventually runs a cleanup project that nobody enjoys.

There are ways around it. Some ERPs support a quote-only item type that never reaches the item master. Some let you price against a template part and adjust. Both work, both need discipline, and both are worth asking about specifically rather than assuming.

Where the ERP quoting module stops

The second limit is the package, and it is the one that costs the most hours. No ERP reads a scanned print. No ERP opens a customer workbook and finds the volume schedule on the third tab. No ERP notices that the quality spec calls for a first article that adds a day of inspection.

All of that still happens in an estimator’s head, and the ERP is the place where the conclusion gets typed. That typing is not the expensive part of quoting. The reading is, and the module does not help with it.

The third limit is who the screen was designed for. ERP quoting screens are built by people thinking about planners, and a planner needs completeness before a record is valid. An estimator needs a number by Thursday.

The gap shows up as 14 mandatory fields between an estimator and a figure they already know, which is why estimators at floors with a perfectly good module still keep a spreadsheet open beside it. That spreadsheet is the real system and the ERP is where the answer gets filed afterwards.

The spreadsheet beside it is the tell.

If that is happening in your building, the module is not being rejected on features. It is being routed around because the shape of the work does not fit the shape of the screen, and no amount of training changes that.

What the integration actually has to carry

Most of the argument about ERP against standalone is really an argument about how good the integration is, and the answer depends on 4 things moving. Burdened rates by work center, material cost and current stock, approved supplier lists for outside processing, and the won quote going back as a routing and a bill of materials.

Rates and material are the easy 2 and every vendor will claim them. Ask how often they refresh, because a nightly pull means January’s rate change lands in quoting the same week rather than whenever somebody remembers.

The supplier list is the one that gets skipped and it matters more than it looks. If quoting does not know which processors are source approved for this customer, an estimator will quote a plating line against a supplier who cannot legally run the work, and the error surfaces after the award.

The write-back is where integrations usually stop, because it is the hardest direction and the least visible in a demo. Without it the quote and the job are 2 unconnected records, and auditing quoted against actual becomes a manual exercise nobody has time for.

Reading from the ERP without putting the logic in it

The arrangement that works on most floors keeps the master data in the ERP and the quoting logic outside it. Rates, material prices and stock come out of the ERP live, so nothing is copied and nothing drifts. The won quote goes back in as a routing and a bill of materials, so production inherits what was actually priced.

That last direction is the one people forget, and it is worth more than it sounds. When the quote writes back, the job that runs is the job that was priced, which means comparing quoted against actual is a report rather than an archaeology project.

The logic stays outside because the logic is the part that needs to change often. Customers change forms, you add a machine, a process moves outside. Changing that inside an ERP means a consultant and a change window. Changing it outside means an afternoon.

This is also the arrangement that survives an ERP migration, which most manufacturers go through every 10 years or so and nobody enjoys. Quoting history that lives inside the ERP is the first thing sacrificed in a migration scope meeting.

Keep the record, move the system.

There is a version of this that goes wrong and it is worth naming. If the quoting side keeps its own copy of rates and material costs, refreshed by somebody remembering to export a spreadsheet, you have built the exact problem the ERP salesperson warned you about. The rule is that master data has one home and quoting reads it rather than storing it.

The same applies in reverse. If quoting invents its own part numbering, or its own work center names, reconciliation becomes a permanent task for somebody. Whatever sits beside the ERP should speak the ERP’s vocabulary, because the alternative is a translation layer that nobody maintains after the person who built it moves on.

What to do if you already own the module

Use it, and be specific about which requests it handles. Repeat releases, reorders, and anything where the part number already exists should go through the ERP, because it is faster and more accurate there than anywhere else.

New parts, messy packages and anything needing outside processing quotes need something in front of the ERP. That is the honest split, and running it deliberately beats running it by accident, which is what a spreadsheet beside the module actually is.

Before buying anything, count a month of requests against that split. If 80 percent are repeats, the module plus a tidy spreadsheet is a defensible answer and the money is better spent elsewhere. If half the month is new work against customer prints, the arithmetic changes and the evaluation questions are worth working through properly.

The mistake is asking one screen to serve both, then concluding the ERP is useless when it did the job it was built for and not the other one.