Most plants have an ERP. Many also have an MES. Yet expediting, schedule churn, and missed priorities still show up every week. That's not because the tools are "bad." It's because they were built for different jobs—and neither one is designed to run execution end-to-end. ERP and MES are not interchangeable ERP and MES sit on different layers of the operating stack. ERP manages the business record: customer orders and order changes; inventory balances and locations; purchasing and supplier commitments; costing, financial postings, and traceability for accounting; master data (items, routings, BOMs, work centers, calendars). MES monitors what happens on the floor: machine and line status; production counts and yields; labor and downtime events; process parameters (where applicable); shop-floor activity and timestamps. Both are valuable. Neither is sufficient to keep execution aligned when reality changes. What ERP actually does (and what it doesn't) ERP systems are designed for structure, standardization, and traceability. They are excellent at answering: - What happened? (transactions, confirmations, receipts) - What exists? (inventory positions, open orders, WIP as recorded) - What was planned? (work orders, planned dates, planned quantities) But ERP is not built to answer: what should happen next—right now—given constraints and new signals? Because ERP logic typically assumes batch updates and periodic reconciliation, clean master data and stable routings, planned sequences that won't be overridden multiple times per day, and decisions made outside the system through meetings, calls, and tribal knowledge. As a result, ERP becomes the ledger of intent and outcomes, not the engine of day-to-day coordination. The latency problem ERP is built around batch cycles. MRP runs once a day or less. Confirmations are posted at end of shift. That latency means the ERP picture of the factory is always hours behind operational reality—which is why experienced planners often stop trusting ERP during the day and work from their own spreadsheets instead. What MES actually does (and what it doesn't) MES sits closer to production reality. It provides high-resolution visibility into execution: - Actual cycle times and stop reasons - Progress against work orders at the line or machine level - Scrap and rework capture - Real-time machine signals where connected to equipment MES is strong at answering: what is happening right now on this asset, line, or order? But MES often struggles to coordinate decisions that span functions: re-prioritizing production when demand changes mid-shift, allocating constrained materials across competing orders, synchronizing packaging with upstream process and warehouse capacity, managing tradeoffs between changeovers, service level, and labor availability. In other words: MES can tell you the plant is falling behind. It typically won't decide, assign, and orchestrate the recovery plan across planning, production, quality, and logistics. The gap between ERP and MES is where execution breaks Between ERP (planning and transaction layer) and MES (execution-data layer) is a missing capability: operational decisioning and workflow orchestration. When that layer is absent, common failure modes emerge: plans don't reflect constraints because the constraint data doesn't flow into the sequence; the shop floor runs the "next job" based on local optimization rather than enterprise priority; material gets staged late because shortage signals don't translate into tasks with owners; quality holds stop flow without a coordinated path to disposition and recovery; expediters become the integration layer—calling across departments to manually coordinate what a system should be doing. This is why "we have ERP and MES" can still coexist with daily firefighting. Each system may be functioning as designed, but the operation is not being coordinated as a system. What the missing coordination layer must provide A practical coordination layer typically: - Ingests real-time inputs from operations and demand (production events, inventory signals, labor availability, equipment status, quality dispositions) - Applies business rules and constraints dynamically (service priorities, changeover logic, material allocation policies) - Prioritizes actions continuously (what to run, what to expedite, what to defer) - Orchestrates workflows across teams (who needs to do what, by when, and in what sequence, with escalation when deadlines are missed) Where the coordination failure is most costly At the order-change interface When customer orders change, the response must span sales, planning, production, procurement, and potentially quality. ERP holds the order. MES holds the production state. Neither system natively coordinates the cross-functional response. A coordination layer that handles order changes structurally—parse the change, assess the impact, route to affected functions, collect responses, update ERP—delivers clear, measurable value. At the quality-production boundary A quality hold affects both MES (stop processing this lot) and ERP (hold the inventory, update the order status, notify customer service). Without a coordination layer, the hold exists in one system and takes hours to propagate to the other—or never does. Plants that manage quality holds informally frequently discover that ERP shows released material that quality has held, or vice versa. That divergence creates risk. At the maintenance-production boundary Unplanned maintenance events interrupt production sequences. In most plants, these coordination flows are manual: maintenance communicates by radio, production adjusts informally, ERP and MES get updated when someone has time. A coordination layer that ingests maintenance events, assesses the impact on the production sequence, and surfaces a recommended resequencing creates a tighter loop between asset availability and production planning. What changes when execution becomes system-driven When execution is structured and system-driven rather than person-dependent and informal: - Decisions become faster and more consistent—the same logic applies across shifts, sites, and supervisors - Data stays aligned across teams—one operational picture, not competing versions - Workflows are predictable, auditable, and repeatable—which means you can improve them Measuring the impact of added coordination Constraint-to-response time: how long from when a constraint is identified to when the floor has a revised sequence? Without a coordination layer, this is often hours. With one, it should be minutes. Exception repeat rate: how often does the same type of exception occur repeatedly without resolution? Repeated exceptions signal that decisions are being made without addressing root cause. A coordination layer captures the root cause with every exception, making pattern analysis possible. Cross-functional SLA compliance: when an exception affects multiple functions, does each function respond within its expected timeframe? With a coordination layer, every exception has owners and time expectations—and compliance is measurable. These metrics only become available when exceptions are captured systematically. That's what a coordination layer provides: not just faster resolution, but the data to understand and improve resolution over time. ERP and MES both remain essential. They just need the coordination layer that connects them to each other—and to the people making decisions between plans and outcomes. Measuring the impact of added coordination When a coordination layer is added between ERP and MES, the impact shows up on metrics that were previously hard to measure. Constraint-to-response time: how long from when a constraint is identified (shortage, hold, downtime) to when the floor has a revised sequence? Without a coordination layer, this is often hours—because the signal travels through informal channels. With a coordination layer, it should be minutes. Exception repeat rate: how often does the same type of exception occur repeatedly without resolution? Repeated exceptions signal that decisions are being made without addressing root cause—which typically means decisions are informal and untracked. A coordination layer captures the root cause with every exception, making pattern analysis possible. Cross-functional SLA compliance: when an exception affects multiple functions, does each function respond within its expected timeframe? Without a coordination layer, SLAs don't exist for cross-functional exceptions. With one, every exception has owners and time expectations—and compliance is measurable. These metrics only become available when exceptions are captured systematically. That's what a coordination layer provides: not just faster resolution, but the data to understand and improve resolution over time. ERP and MES both remain essential. They just need the coordination layer that connects them to each other—and to the people making decisions between plans and outcomes. A practical path to improving manufacturing execution Improving execution doesn't start with a platform decision. It starts by isolating where coordination breaks down. Identify high-leakage processes: target areas where decisions repeatedly move outside systems—production planning adjustments made informally during the shift, order promising and change management handled through calls and emails, inventory movements and material substitutions resolved through verbal authorization. Reduce manual coordination: replace ad-hoc communication with structured workflows that route decisions to the right owner automatically based on event type and severity, capture the decision rationale and supporting context, and update downstream teams automatically when a decision is made. Align functions through a single execution layer: connect sales, production, and procurement actions so tradeoffs are made once, transparently, using shared rules—not negotiated repeatedly in side channels. ERP and MES both remain essential. They just need the coordination layer that connects them to each other—and to the people making decisions between plans and outcomes.