What is MRP, really? (plain English)

MRP stands for Material Requirements Planning. It is a method, and now, a category of software, for answering one practical question: to deliver what I have promised, what do I need to make, what do I need to buy, and when.

The jargon is older than the software. It dates to 1960s American manufacturing, and much of the terminology is still inherited from that era. But the problem is the same as the one you have right now: you owe customers finished goods, you have some stock, you have some lead times, and you need a plan.

The three inputs

Every MRP calculation, from the simplest to the most sophisticated, takes three things and produces a plan.

1. Demand, what you need to ship. That includes confirmed sales orders, subscription replenishments, and a forecast for anything not yet ordered. Demand has a date.
2. Supply, what you already have, plus what you are already making. On-hand stock and work-in-progress. Supply also has a date (it is available from today, or on the date the in-progress batch completes).
3. Bills of materials, how much of each component goes into each finished product, so that a demand for finished goods can be translated into a demand for components.

The one equation that matters

Everything else in MRP is a consequence of that equation. The complexity comes from doing it at every level of the bill of materials, for every date in the planning horizon, across every SKU in the business.

A worked example

Imagine a cosmetics manufacturer making a velvet matte lipstick. The BOM, simplified, is:

• 1 × base formulation (itself a sub-assembly)
• 0.2g × iron oxide red pigment
• 1 × lipstick tube
• 1 × cap
• 1 × carton

Monday's demand: 500 units promised to a retailer for Friday delivery. On-hand finished stock: 120 units. Work-in-progress: 0.

MRP resolves this as follows:

1. Finished-goods gap: 500 − 120 = 380 units needed by Friday.
2. Component demand (exploding the BOM for 380 units): 76g pigment, 380 tubes, 380 caps, 380 cartons, and 380 × base formulation.
3. Base formulation: itself a sub-assembly with its own BOM, which explodes into its own component demand. MRP walks down the tree recursively.
4. For each component, check stock. If short, generate either a production order (if you make it) or a purchase order (if you buy it), timed to land before the production run that needs it.

That last step is where time-phasing comes in. MRP does not just answer “what do I need” but “when do I need it”, backing the date off by supplier lead times and production lead times.

Where MRP tends to go wrong

• Bad master data. If your BOMs are wrong or your stock figures are stale, MRP produces a confident plan for the wrong business. See our guide on building BOMs properly.
• Over-forecasting demand. The forecast is the weakest input. A system that treats forecast demand as firm demand will over-purchase.
• Ignoring work-in-progress.If you don't record running batches, MRP will plan a second one.
• One plan for every product. Fast-moving SKUs want reorder-point planning. Slow-moving SKUs want order-by-order planning. Treating them the same wastes stock.

How MRP differs from the neighbouring things

• Inventory management tracks stock levels. MRP plans what to do about them.
• ERP includes MRP plus accounting, HR, CRM, warehouse, and more. For small manufacturers, this is usually too much surface area. See MRP vs ERP.
• MES (Manufacturing Execution Systems) sit between the plan and the shop floor, barcoding, operator instructions, real-time quality. MRP tells you what to run; MES helps you run it.

Who actually needs MRP

The honest answer: any manufacturer with a bill of materials more than one level deep, more than one active SKU, and more than one supplier. Below that threshold, a spreadsheet works. Above it, spreadsheets silently produce wrong answers.

You know you have crossed the line when you can no longer hold the whole picture in your head, but your spreadsheet still says you can.

Stokka's take on MRP

Stokka calls it the production plan. One click builds a weekly capacity plan from your open production orders and real routings. An at-risk board flags the must-start weeks. An exceptions list shows what needs attention. The underlying maths is MRP; the interface is owner-operator-sized.

Multi-level BOMs explode through sub-assemblies. Reorder suggestions factor in supplier lead times. A run in progress counts as supply, so the plan never tells you to make something you are already making. The next layer is on the roadmap: one screen that nets demand (committed orders, subscriptions, forecast) against supply, with the gap by SKU and by date.

Further reading

Next, read MRP vs ERP to decide which shape fits your operation. If your BOMs are the bottleneck, start with how to build a BOM. If you are still on spreadsheets, read the MRP switchover.