New to this topic?
We recommend reading these guides first to get the most out of this one:

What Is an ERP System?

An Enterprise Resource Planning (ERP) system is the central nervous system of a manufacturing organization. It integrates all business functions — planning, purchasing, inventory, production, quality, finance, and HR — into a single database so that every department works from the same source of truth.

If you've used SAP, Oracle, Epicor, Infor, or Microsoft Dynamics on a shop floor, you've interacted with an ERP. But most people only see their tiny slice. Understanding the whole system — and how MRP logic actually works — is what separates operators from people who can actually improve the operation.

$50B+
Global ERP market
95%
Of large mfg companies use ERP
50-75%
Of ERP implementations fail expectations
1960s
MRP origins at IBM

Core ERP Modules for Manufacturing

A manufacturing ERP typically includes these interconnected modules:

Planning modules

ModuleWhat it doesKey outputs
Master Production Schedule (MPS)Translates demand (forecasts + orders) into a build plan by time periodWhat to make, how many, which week
MRP (Material Requirements Planning)Explodes the BOM, offsets lead times, nets against inventory to generate purchase and work ordersPlanned orders, purchase requisitions
CRP (Capacity Requirements Planning)Loads MRP planned orders onto work centers to check if you have enough capacityLoad profiles, overload alerts
S&OPMonthly cross-functional alignment of demand, supply, and financial plansAgreed demand/supply plan

Execution modules

ModuleWhat it doesKey outputs
Shop Floor ControlReleases work orders to production, tracks labor/material consumptionWork order status, labor booking
PurchasingConverts MRP requisitions into POs, manages supplier relationshipsPurchase orders, receipts
Inventory ManagementTracks raw material, WIP, and finished goods across locationsStock levels, transactions, valuations
Quality ManagementInspection plans, non-conformance tracking, CAPA managementInspection results, NCR records

Financial modules

ModuleWhat it doesKey outputs
Cost AccountingTracks actual vs. standard costs by work order, product, and departmentVariance reports, product costing
General LedgerAggregates all financial transactions into the booksP&L, balance sheet
Accounts Payable/ReceivableManages supplier payments and customer invoicingCash flow, aging reports

How MRP Logic Actually Works

MRP is the engine inside every manufacturing ERP. Understanding its logic is essential. Here's what happens when MRP runs:

Start with demandMPS says: "We need 500 units of Product A in Week 12."
Explode the BOMProduct A requires 2x Part B, 1x Part C, 4x Part D. Each may have sub-components — the BOM tree can be 10+ levels deep for complex products.
Net against inventoryWe have 200 Part B on hand, 50 on order arriving Week 10. Gross requirement = 1000, net requirement = 750.
Lot size the orderMinimum order qty from supplier = 500. So we plan to order 1000 (2 × 500).
Offset by lead timePart B has a 3-week lead time. Need by Week 12 → order by Week 9. This is backward scheduling.
Generate planned ordersOutput: "Release PO for 1000x Part B in Week 9." Repeat for every component at every level.
Demand (MPS)
→
BOM Explosion
→
Netting
→
Lot Sizing
→
Lead Time Offset
→
Planned Orders
The MRP calculation loop — runs for every item at every BOM level
The "nervous system" problem: MRP is extremely sensitive to data quality. If your BOM is wrong, your inventory counts are off, or your lead times are inaccurate, MRP will generate perfectly calculated — but completely wrong — orders. In practice, data accuracy below 95% makes MRP unreliable. This is why cycle counting, BOM audits, and lead time validation are not optional.

Master Data: The Foundation

An ERP system is only as good as its master data. These are the records that every transaction references:

  • Item Master — Every part number: description, unit of measure, lead time, lot size rules, safety stock, cost, ABC classification
  • Bill of Materials (BOM) — Parent-child relationships: what components go into what assemblies, quantities per, effectivity dates
  • Routing — Manufacturing steps: sequence of operations, work centers, setup time, run time per unit, tooling requirements
  • Work Center — Resources: machines, labor pools, available hours, efficiency rates, cost rates
  • Supplier Master — Vendor information: lead times, pricing, quality ratings, approved parts
  • Customer Master — Ship-to, payment terms, pricing agreements
⚠️
Lesson from defense manufacturing: At large defense manufacturers, a single aircraft program may have 300,000+ part numbers in the BOM. A wrong quantity-per on one component can cascade through MRP and create shortages that halt an assembly line for weeks. This is why Configuration Management (CM) and Engineering Change Order (ECO) processes are so rigorous — they protect the BOM integrity that MRP depends on.

From MRP to MRP II to ERP

The evolution matters because it explains why modern ERPs are structured the way they are:

EraSystemWhat it added
1960sMRPMaterial planning — "What do I need and when?"
1980sMRP IIAdded capacity planning, shop floor control, purchasing, finance — closed-loop planning
1990sERPExtended to all enterprise functions: HR, CRM, supply chain, business intelligence
2010s+Cloud ERP / ERP IISaaS delivery, real-time analytics, IoT integration, AI-driven planning

Major ERP Systems in Manufacturing

SystemTypical customerStrengths
SAP S/4HANALarge enterprise, defense, automotiveDeep functionality, global multi-site, regulatory compliance
Oracle Cloud ERPLarge/mid enterprise, process & discreteModern cloud architecture, strong financials
Epicor KineticMid-market discrete manufacturingShop floor focus, job shop support, MES integration
Infor CloudSuiteIndustry-specific (food, fashion, aerospace)Pre-configured industry solutions
Microsoft Dynamics 365Mid-market, distribution-heavyOffice integration, rapid deployment
Plex (Rockwell)Automotive, food & beverageCloud-native, built-in MES, quality

Why ERP Implementations Fail

50-75% of ERP implementations fail to meet expectations. The reasons are almost always the same:

Common failure patterns
  • Treating ERP as a "software project" instead of a business transformation
  • Dirty master data migrated into the new system
  • "Lift and shift" — replicating broken processes in new software
  • Underestimating change management and training
  • Over-customization that makes upgrades impossible
  • Go-live without adequate testing or parallel runs
Success patterns
  • Executive sponsorship with authority and budget
  • Data cleansing BEFORE migration (BOM accuracy >98%)
  • Process redesign first, then configure software to match
  • Dedicated training with role-based curricula
  • Configure, don't customize — use standard functionality
  • Phased rollout with pilot sites before enterprise-wide

ERP vs. MES: Understanding the Boundary

One of the most common points of confusion: where does ERP end and MES begin?

FunctionERPMES
Planning horizonWeeks to monthsHours to days
Scheduling granularityDaily or weekly bucketsMinute-by-minute sequencing
Data collectionManual entry or batch importReal-time from machines/operators
Decision speedPlan, then executeReact in real-time to floor events
Typical usersPlanners, buyers, financeOperators, supervisors, quality
"Source of truth" forWhat should happen (the plan)What is happening (actual execution)

The ideal is bidirectional integration: ERP sends work orders and schedules down to MES; MES sends actual production data, labor, and quality results back up to ERP. When this works well, your plan and your reality are always in sync. Read more in our MES Systems guide.

How ERP Connects to Everything Else

PLM
→
ERP
→
MES
→
Shop Floor
PLM designs the product → ERP plans production → MES executes on the floor
  • PLM → ERP: Engineering releases BOM and routing data from the PLM system into ERP for production planning
  • ERP → MES: ERP sends work orders, material allocations, and schedules to MES for execution
  • MES → ERP: MES sends actual completions, labor hours, scrap, and quality data back for costing and reporting
  • ERP → BI: ERP transactional data feeds into Business Intelligence tools for dashboards and analytics
  • ERP → Supply Chain: Purchase orders, ASNs (advance shipping notices), and EDI transactions flow to suppliers
💡
For a complete picture of how all these systems interact, see our interactive Production Systems Ecosystem guide.

Key Takeaway

Remember This

ERP is the planning backbone — it answers "what to make, when, with what materials, at what cost." Its MRP engine backward-schedules from demand dates, explodes BOMs, and generates the orders that drive purchasing and production. But ERP only works when master data is accurate and processes are disciplined. An ERP can't fix broken processes — it will just break them faster.

Interactive Demo

Explore ERP modules and see how integrated data flows between production, inventory, purchasing, and quality.

⚡
Try It Yourself
ERP Module Explorer
▼
Click modules to activate them and see how data flows between integrated ERP modules. Compare the benefits of integrated data vs siloed systems.
ProductionPlanningInventoryPurchasingQualityMaintenanceFinancials
Integrated vs Siloed Impact
45 min
— 0 min vs baseline
Data Entry Time
8.5%
— 0.0% vs baseline
Error Rate
72h
— 0h vs baseline
Reporting Delay
25%
— 0% vs baseline
Visibility
Ready for the full knowledge check? Test your understanding with guided scenarios and data export.
PROTake the Pro Knowledge Check →
Free forever · Every feature included

Stop reading, start modeling

Model your process flow, run simulations, optimize staffing with TOC math, and test your knowledge with 107 interactive checks — all in one platform.

Open Workbench → Pro Knowledge Check

Take this to a room

The running order

For anyone on an ERP project or living with one. They should leave knowing why implementations fail and where the ERP/MES boundary sits.

7 beats · 13 min
  1. 1

    Between half and three quarters disappoint

    Fifty to seventy-five per cent of ERP implementations fail to meet expectations. That is the context for everything else.

    • Ninety-five per cent of large manufacturers run one, so this is not a fringe risk.
    • The failures are not usually technical.
    • They are data, process and change management.

    Ask the room How does our current system get described on the floor?

  2. 2

    What MRP actually does

    Six steps, and knowing them makes every planning conversation clearer.

    • Master schedule, BOM explosion, netting against on-hand and on-order.
    • Lot sizing, lead time offset, then planned orders.
    • Every one of those steps trusts a piece of master data. That is where it goes wrong.
  3. 3

    Master data is the foundation

    BOMs, routings, lead times and item master. If these are wrong, everything computed on them is wrong.

    • Clean the data before migration, not after. Target BOM accuracy above 98 per cent.
    • Dirty data migrated into a new system produces a more expensive version of the old problem.
    • This is the single most common root cause of a disappointing go-live.

    Ask the room When did we last audit a routing against the floor?

  4. 4

    Do not lift and shift

    Replicating a broken process in new software buys you the broken process at higher cost.

    • Redesign the process first, then configure the software to match.
    • Configure, do not customise - over-customisation makes upgrades impossible.
    • A software project framing guarantees a software-shaped result.
  5. 5

    The ERP and MES boundary

    Worth drawing on the board, because most arguments here are boundary arguments.

    • ERP plans: what to make, when, from what, at what cost.
    • MES executes: what is actually happening right now on the floor.
    • PLM to ERP to MES to shop floor. Each layer answers a different question.

    Ask the room Which of our questions are we currently asking the wrong system?

  6. 6

    Why implementations fail

    Six patterns, and none of them is about the software vendor.

    • Treated as a software project rather than a business change.
    • Dirty data migrated. Broken processes replicated.
    • Change management and training underestimated. Over-customisation.
    • Go-live without a pilot.
  7. 7

    What success looks like

    The success patterns are the mirror image, and they are all decisions made early.

    • Executive sponsorship with real authority and budget.
    • Data cleansing before migration. Process redesign before configuration.
    • Role-based training curricula, not a generic course.
    • Phased rollout with pilot sites before the enterprise.

    Ask the room Which of those are we missing right now?