What a manufacturing ERP has to do, how to run a selection that holds up, and what the decision really costs.
A manufacturing ERP runs the whole business on one set of records: what you sell, what you buy, what you build, what it costs, and what the general ledger says about all of it. Every system in the market can show you a clean demo. The ones that fail do so eighteen months later, over things nobody asked about during the demo: how scrap gets recorded at an operation, whether a lot can be traced through a rework, what happens when the customer changes the order after the work order is released. This guide covers what a manufacturing ERP has to do, how to run a selection that survives contact with your shop floor, how to match a system to the way you actually build, and what drives the cost. If you already know NetSuite is the direction, the NetSuite manufacturing guide goes deeper on the software itself.
Manufacturing ERP is not accounting software with a production module bolted on. The test is whether one transaction on the floor updates inventory, cost and the ledger at the same moment. These are the capabilities to hold every candidate against.
| Capability | What to look for | Why it bites later |
|---|---|---|
| Items and bills of materials | Multi-level BOMs, revisions with effective dates, phantom assemblies, alternates | Engineering changes mid-production are normal, and a system that cannot version a BOM forces spreadsheets back in |
| Work orders and WIP | Issue or backflush by component, partial completions, scrap and rework at operation level | This is where the cost of a job either becomes visible or disappears |
| Routings and capacity | Work centers, setup and run times, load by resource, reschedule without rekeying | Promise dates come from capacity, not from optimism |
| Planning | MRP that reads demand, stock, open orders, lead times and safety stock, and proposes orders a planner can edit | Without it, purchasing runs on reorder points that nobody has reviewed in two years |
| Inventory and traceability | Bins, lots and serials, expiry, FEFO picking, full genealogy both directions | A recall, an audit or a customer complaint is when you find out whether trace actually works |
| Costing | Standard, average or actual, with variances that reconcile to the ledger | Margins by product are only as good as the costing method behind them |
| Quality | Inspection at receipt, in process and before shipment, with hold and disposition | Quality kept in a separate system means nonconforming stock stays sellable |
| Purchasing and supply | Blanket orders, vendor lead times, receiving against POs, landed cost | Material cost variance is invisible unless landed cost lands on the item |
| Order management | Configurable items, make to order, partial shipments, returns and credits | Sales promises are made in this screen and paid for on the floor |
| Reporting | Numbers people trust without an export: schedule adherence, yield, margin by item, WIP value | If the answer lives in a spreadsheet, the ERP is not the system of record |
Two more that rarely appear on a requirements list and always matter: how the system handles your integrations (ecommerce, EDI, shipping, machine data) and how much of your process depends on a single person knowing where to click.
Generic requirement lists produce generic demos. “Lot traceability” is not a requirement. “Trace a finished lot back to the supplier lot of every ingredient, including material added during a rework, within one screen” is. Aim for thirty to fifty statements at that level, written by the people who do the work, not by the software committee.
Walk the floor and write down the exceptions: the job that gets split across two machines, the part that is issued from a staging bin nobody counts, the customer who sends changes by text. Exceptions are what break implementations, and they are cheaper to find now than during testing.
Analyst grids and vendor lists are a starting point, not a shortlist. Cut to three that serve manufacturers of your type and size, and be honest about whether you are a discrete, process, engineer-to-order or mixed-mode shop. A system built for one is rarely good at another.
Send each vendor the same five scenarios from your own operation, with your data if you can: quote to cash on a configured item, a work order with a component shortage, a batch with a yield loss, an engineering change on a released order, a month-end close. Watch who says “we would handle that with a customization” and count the clicks.
Ask for two references in your industry that went live in the last two years, and ask them what went wrong rather than what went well. At the same time, confirm how each system connects to the things you already run: the ecommerce store, the EDI trading partners, the shipping software, the machines on the floor.
Compare subscription, implementation services, integrations, third-party add-ons, training, and the internal time your team will spend. An honest model usually changes the ranking. Our implementation estimator shows how services scale with scope, and the implementation guide explains the phases behind the number.
Agree the phases, who from your team is on the project and for how many hours a week, what data is migrating, and what will be left in the old system. A signature that comes before this conversation buys software. A signature that comes after it buys a project.
Most failed selections come down to the wrong fit for the production model, not the wrong vendor.
| How you build | What the ERP has to handle well | Where to read more |
|---|---|---|
| Discrete, repetitive | Standard BOMs and routings, backflushing, capacity by work center, standard costing with variances | NetSuite manufacturing guide |
| Job shop and make to order | Quoting from estimates, job costing, changeovers, labor capture by job | Metal fabrication and machine shops |
| Engineer to order | Project structures, BOM revisions during build, milestone billing, actual versus estimate | Engineer-to-order guide |
| Process and batch | Formulas, yields, catch weight, lot and shelf life, co-products and by-products | CPG manufacturing guide |
| Mixed mode | Both discrete and process in one item master, without two systems and a nightly file | NetSuite for manufacturing |
If two of these describe you, say so early. Mixed-mode operations are common in food, chemicals and contract manufacturing, and they are the fastest way to discover that a system only handles one of them properly.
Every requirement list grows. The discipline is deciding which items would make you walk away. A working rule: a must-have is something that, if the system cannot do it, means a person will keep a spreadsheet. Everything else is a preference.
Score the shortlist against the must-haves only. Preferences are how vendors win on points while losing on fit.
Manufacturing ERP has three cost lines and they behave differently.
Add-ons sit alongside all three: SuiteApps or third-party modules, shop floor hardware, and any ongoing support plan after go-live. For a services range and timeline against a scope you define, use the implementation estimator.
We implement NetSuite, so read this section with that in mind. NetSuite is a cloud suite: financials, inventory, purchasing, order management and manufacturing on one ledger, extended with SuiteApps rather than separate systems. For manufacturers that means work orders, BOMs and routings, WIP, MRP, quality and costing sit next to the general ledger instead of feeding it overnight.
Where it fits well: companies that want one system across finance and operations, that are outgrowing QuickBooks or an aging on-premise ERP, and that value adding capability through apps over running a second system. Where it needs help: deep shop floor scheduling, formula and batch work, project-based manufacturing. That is what our products extend, and it is worth knowing which parts of a vendor pitch are core product and which are add-ons.
If you are comparing NetSuite against other platforms, we keep the detail on our NetSuite comparison pages rather than in this guide.
An ERP selection run by finance alone picks a general ledger. One run by operations alone picks a scheduling tool. The selections that hold up have five roles involved from the start.
Two habits matter more than the org chart: write decisions down as you make them, and keep a list of every process someone says is “just how we do it”. That list becomes the design phase agenda.
The cloud argument is mostly settled for mid-market manufacturers, and it distracts from the questions that still differ between systems.
A manufacturing ERP runs finance and operations on one set of records: items and bills of materials, work orders and work in process, planning, inventory and traceability, purchasing, order management, costing and the general ledger. The distinction from general business software is that a transaction on the floor updates inventory, cost and the ledger at the same time.
Most selections run two to four months from writing requirements to signing, assuming someone owns the project. The work that takes longest is not the vendor demos, it is agreeing internally on how you want to run, and getting reference calls scheduled.
There are three lines: the subscription, priced by the vendor on users and modules, implementation services, priced by scope, and your own team time. Services vary most between quotes, because scope varies. Our implementation estimator shows how services move with scope.
Shortlist software first, then evaluate partners for the finalists, then decide together. The partner does the work that determines whether the software fits your plant, so a strong partner on the second-best platform often beats the reverse.
Not always. The signals that you do: nobody can say what a job cost without a rebuild, inventory accuracy is below what your customers need, planning happens in one person head, or an audit or customer requirement now needs traceability you cannot produce quickly.
Yes, but check it rather than assume it. Ask to see a formula with a yield loss and a discrete work order with routings in the same demo account, using your items, and ask how costing works for each.
Running a manufacturing ERP selection, or already leaning toward NetSuite? Get in touch and we will walk through your requirements, your production model and what an implementation would involve.
