1. Supply Chain
  2. Features
  3. Demand planning
Feature · Demand planning

Demand planning

Build forecasts by product, warehouse and time bucket under named versions, run MRP and act on replenishment suggestions.

OverviewReplenishment Suggestions
Supply Chain · Demand planning

Replenishment Suggestions

Replenishment plans

New
Create PO2
PRD-8114WRH-110
470
03 Oct
PRD-2790WRH-130
130
24 Oct
Add
Create transfer2
PRD-7671WRH-110
20
22 Oct
PRD-4516WRH-150
260
27 Oct
Add
Create production order1
PRD-7021WRH-110
370
09 Oct
Add
Do nothing1
PRD-2296WRH-150
800
22 Oct
Add
Example screen with sample data

What you can do

Plan from one agreed number

Every buying and production decision starts from the forecast. The Demand Planner records demand per product, per warehouse, per time bucket: the statistical forecast, the sales team’s override and the consensus figure everyone agrees to plan against. Actual demand is recorded beside it so accuracy can be measured. Forecasts sit under a named version, such as the baseline, a sales scenario or the locked plan, and each version moves from Draft to In Review, Approved and Locked. Older versions are Superseded, never overwritten.

Time buckets are weeks, months or quarters with a start and end date and a fiscal period. A bucket is Open while planners can change it, Frozen when the near term is fixed, and Closed once it is past. MRP runs against a forecast version over a horizon and produce replenishment suggestions: create a purchase order, a transfer or a production order, or do nothing, each with the net requirement, the need-by date and a suggested supplier.

How the work is organized

  • Consensus per bucket. Statistical, sales and consensus figures sit side by side for each product, warehouse and bucket.
  • Versions locked before MRP. MRP runs against an approved or locked version, so the plan behind each suggestion is known.
  • Suggestions reviewed, then actioned. The planner reviews each suggestion and actions it, ignores it, or lets it expire.

What to agree during setup

Agree the bucket granularity for each horizon, for example weeks for the next quarter and months beyond. Agree the freeze window. Agree who may override the statistical forecast and who approves the consensus version. Agree the MRP horizon and how often it runs. Decide how expired suggestions are reviewed.

Check the handoff on an actual case

For March, the statistical forecast for a 5 kg bag of resin at the Dallas warehouse is 1,200 units. Sales expects a large customer order and overrides to 1,600; the consensus meeting agrees 1,500. The planner moves the version to Approved and then Locked. The MRP run that night, over a 90-day horizon, finds 400 on hand, 200 allocated and safety stock of 300, and suggests a purchase order for 1,600 from the primary supplier, needed by February 20.

The planner reviews the suggestion and actions it. The buyer’s purchase order then carries the reference back to the plan. In April, actual March demand comes in at 1,430, and the accuracy figure for the bucket is recorded against the locked version.

What your team sees

  • Replenishment Suggestions — New and reviewed suggestions with product, warehouse, net requirement, action, need-by date and supplier.
  • Forecast Versions — Versions by status with type and lock date.

Follow the work through

Review the data model before loading history.

What it keeps track of

Data model
Demand forecasts
Forecasted demand per product per warehouse per bucket under a version.
Forecast versions
Named forecast iterations: baseline, S1, S2, locked.
Forecast buckets
Time buckets against which forecasts are stored.
MRP runs
Execution-level record of a material requirements planning run.
Replenishment plans
MRP and reorder-point-driven net demand suggestions.

Who uses it

Permissions
Demand Planner
Builds forecasts, runs MRP and reviews suggestions.
Buyer
And Inventory Manager act on purchase order and transfer suggestions.
Admin
Manages access.

Set up Supply Chain your way.

Start free, then use Proto to change fields, steps and approvals before your team goes live.

How to get started