The beverage order is ready. The bottles are not.

One formulation can create several planning problems

A base drink sold in two bottle sizes shares a formulation but uses different packaging and may run at different speeds. Keep the approved formula separate from each pack’s bill of materials. Connect both to the finished product and preserve the versions used by an actual run.

In an illustrative planning exercise, an order needs 5,000 bottles but only 4,200 matching caps are available. The planner needs a visible shortage and the expected replenishment date. Quietly reducing the order or borrowing a superficially similar cap moves an unresolved decision out of view.

Ask who approves a component substitution and what open orders it affects. The purchasing team can identify an alternative; the relevant product and quality owners decide whether it is acceptable. The application should record that decision and its scope.

Put changeovers on the schedule

A schedule that assigns every minute to production has left no room for the work between runs. Document the line’s capabilities, operating calendar and changeovers with the people who run it. Include the time and resources needed for the plant’s approved cleaning and setup process, rather than borrowing assumptions from a software demonstration.

The planner should be able to move an urgent order and see which commitments shift. Preserve the earlier schedule for comparison. If every revision overwrites the previous plan, a report can show excellent adherence simply because it is measuring against the latest version.

An existing ERP may already own production orders and material records. A separate planning tool can be useful when sequencing is the gap, but agree how an approved schedule returns to the ERP and how later changes are acknowledged. Two editable schedules create extra work at precisely the point where a single decision is needed.

Close the run in units that production and finance both recognise

Record material issued, actual liquid output, filled units, packaging rejects and material returned. Compare planned with actual quantities without forcing every difference into one process-loss field. A broken bottle, an incorrect opening count and a late posting require different follow-up.

When one run fills several pack sizes, retain the connection from its inputs to each output. GS1’s traceability standard provides a reference for those transformation relationships. The inventory transaction and the lot history should describe the same physical event.

A reversal is worth demonstrating. If someone corrects the packed quantity, the stock balance, production record and downstream allocation should all agree afterwards. A tidy original transaction is the easy part of the evaluation.

Source: GS1 · Section 5: data recording requirements, R24. Reference details

A forecast is useful only before the commitment is locked

Packaging orders, production runs and distributor replenishment operate on different horizons. Decide which decision a forecast is supposed to inform. A prediction made after packaging is committed cannot improve that purchase, however accurate it looks afterwards.

Keep distributor orders, consumer sales where available, promotions and stock-building events distinct. Compare any forecast with a simple recent-history or seasonal baseline at the horizon the team can act on. Record overrides so the team can distinguish a model error from a deliberate commercial decision.

A useful first demonstration takes an order through a shortage

Choose a product family with two pack formats. Delay one packaging receipt, revise the sequence, record a partial run and correct an output quantity. Then show what sales, purchasing and dispatch see. This exposes more of the implementation than an ideal run with every component available.

Before expanding the scope, check whether planners spend less time reconstructing the schedule and whether the reasons for missed runs are recorded consistently. Those observations will help decide whether the next investment belongs in better master data, a planning application or more advanced optimisation.