The supply chain planning process is a repeating loop of six steps: forecast demand, build a supply plan against real constraints, explode that plan through the bill of materials, convert the result into purchase orders, execute and monitor, then review what actually happened and feed it back. Each step runs on its own cadence — the forecast monthly, the supply plan weekly, execution daily — and the loop is only as good as its slowest, stalest step.
Below is what each step produces, who owns it, how often it should run, and where it typically breaks.
Step 1 — Demand forecast
Produces: expected volume by SKU and period. Cadence: monthly, refreshed weekly against actual orders.
Start from a statistical baseline built on shipment history, then adjust for what history cannot know: promotions, listings won or lost, new customers, seasonality that has shifted. Customer forecasts, where you get them, are an input rather than an answer — most retail forecasts are optimistic and should be tracked for bias like any other source.
The measure that matters is forecast accuracy at the level you plan, plus bias. Accuracy tells you how noisy the forecast is; bias tells you whether it is wrong in a consistent direction, which is the more expensive problem because it quietly inflates every downstream order.
Step 2 — Supply plan
Produces: a production plan the factory can run. Cadence: weekly, with a rolling 8–13 week horizon.
This is where the forecast meets reality: line capacity, shift patterns, changeover time, labour, shelf life and storage. The output is which SKU runs on which line in which week, in a sequence that minimises changeovers without letting any product age past its window.
Where capacity does not cover the forecast, the gap gets escalated with options rather than silently absorbed — build ahead, add a shift, move volume, or short the customer knowingly. The distinction between this step and the forecast is covered in supply planning vs demand planning.
Step 3 — Material requirements (MRP)
Produces: net requirements by material and date. Cadence: weekly, and after any material change to the production plan.
The arithmetic is mechanical and always the same:
- Explode the production plan through the bill of materials to get gross requirements.
- Subtract stock on hand that is actually usable — not blocked, not expired, not allocated elsewhere.
- Subtract stock on order that will land in time, using the supplier’s realistic lead time rather than the contracted one.
- What remains is the net requirement, which is then rounded up to the MOQ or pack size and offset backwards by the lead time to give an order date.
Worked through one line item: a bakery needs 900 kg of butter across three weeks of production. It holds 240 kg, of which 40 kg is reserved for a different run, so usable stock is 200 kg. There is an open order for 250 kg due in week 2. Net requirement is 900 − 200 − 250 = 450 kg. Butter ships in 25 kg blocks with a 200 kg MOQ, so the order is 450 kg — 18 blocks. Lead time is 10 working days, and it is needed at the start of week 3, so the purchase order goes out this week, not next.
Run that for every component and the output is the buying list.
Step 4 — Purchasing
Produces: confirmed purchase orders with dates. Cadence: daily.
Raising the order is the fast part. The slow part is everything after: getting an acknowledgement, getting a confirmed delivery date, noticing when the confirmation says something different from the order, and chasing the ones that go quiet.
An order without an acknowledgement is not a supply commitment, it is a hope. In practice the useful discipline is to treat three things as separate states — sent, acknowledged, confirmed with a date — and to chase on the transition that has stalled rather than on a blanket weekly follow-up of everything outstanding.
Step 5 — Execution and exception management
Produces: decisions, fast. Cadence: continuous.
This is the step that determines whether the plan survives contact with the week, and the one most planning documentation skips.
Typical exceptions, in rough order of how often they occur: a supplier confirms a partial quantity; a delivery slips; goods arrive and fail QC; a line breaks down and a run moves; a customer pulls an order forward; a stock count corrects downward. Each one invalidates part of the plan.
The discipline that works is triage by consequence, not by size. The question is never “how big is this shortage” but “which run does it stop, and when”. A 40 kg shortfall that halts Thursday’s line outranks a 2-tonne shortfall on a material not needed for six weeks. That ordering is obvious in retrospect and surprisingly hard to maintain at 8am with 30 unread supplier emails.
Step 6 — Review
Produces: corrected master data and a better plan next cycle. Cadence: monthly.
The review closes the loop, and it is the step most often dropped when the week is busy. Three things are worth reviewing every month:
- Forecast accuracy and bias, by category and by customer, so you know which sources to trust.
- Supplier on-time-in-full and actual lead times versus the lead times in the system. This single comparison, done monthly, removes more shortages than any amount of extra safety stock — because most “supplier problems” are really stale master data.
- Expedites and their causes. Every expedited order is a planning failure with a traceable origin: late forecast, stale lead time, missed exception, or a genuine supplier failure. The mix tells you what to fix.
A month in practice
For a mid-sized food manufacturer, the loop typically lands like this:
- Week 1: actuals close. Forecast refreshed, accuracy and bias reviewed. Supplier performance pulled for the S&OP pack.
- Week 2: S&OP meeting — one agreed plan across commercial, operations and finance. Capacity gaps and their options are decided here, not in the corridor afterwards.
- Weeks 1–4, every week: supply plan rebuilt for the rolling horizon, MRP run, buying list issued, long-lead items ordered.
- Every day: confirmations chased, exceptions triaged, the schedule adjusted for what actually happened.
The monthly rhythm gets the attention because it involves meetings. The daily loop is where service level is won.
The bottleneck is the loop speed, not the maths
None of the arithmetic above is difficult. The reason planning fails is latency: the plan is rebuilt weekly, exceptions arrive continuously, and the gap between them is filled with stock, expediting and phone calls.
Shorten that gap and most of the symptoms go away. That is the design goal behind Leio — it connects to your ERP and mailbox, keeps the material plan live against the production schedule, raises and chases the purchase orders, and puts the exceptions that threaten a run in front of the planner while there is still time to act. The planner decides; the loop keeps running.
If your planning cycle is currently as fast as whoever has time to rebuild the spreadsheet, book a demo — for background on the terms used here, start with what is supply chain planning.