Supply Chain & Operations
Supply chain planning technology
Planning fails quietly. A forecast nobody believes, a supply plan built in a spreadsheet beside the planning system, and an S&OP meeting that reconciles numbers instead of making decisions.
Why planning systems get abandoned
Planning tools are among the most frequently bought and least frequently used enterprise systems. The pattern is consistent: the tool is implemented, the planners keep the spreadsheet, and within a year the spreadsheet is authoritative again.
It usually comes down to trust. If the system cannot explain why it produced a number, a planner who is accountable for the outcome will not defend it — and will quietly maintain a parallel model they can defend.
What planning actually covers
- Demand planning
- Statistical baseline, promotional and event overlays, new product introduction, and the consensus process that turns a forecast into a commitment.
- Supply planning
- Constrained supply against capacity, materials, lead times and supplier reality rather than nominal lead times nobody meets.
- Inventory planning
- Safety stock by service target and demand variability, not by a flat weeks-of-cover rule applied to every SKU.
- Capacity planning
- Rough-cut and detailed capacity across plants, lines, labor and storage.
- S&OP / IBP
- The cadence that reconciles demand, supply, inventory and finance into one plan people are accountable to.
- Scenario planning
- What happens under a demand shock, a supplier failure or a capacity loss — answered before it happens rather than during.
Where planning gets its data
A planning system is only as good as what feeds it, and it feeds decisions in every direction:
- ERP — orders, item master, BOM, lead times, costs, open supply.
- WMS — actual on-hand and location-level availability, which is often materially different from the ERP figure.
- MES — production capability, actual yields, and where capacity really goes rather than where it is assumed to go.
- TMS — transit times and cost-to-serve, which change the answer to "where should this inventory sit".
- Commercial systems — pipeline, promotions and customer commitments.
- External signals — supplier confirmations, POS or sell-through, and lead-time volatility.
Implementation approach
- Segment before modeling
- Not every SKU deserves the same planning treatment. Volume and variability segmentation determines what should be forecast statistically and what should not be forecast at all.
- Measure forecast quality honestly
- Bias and accuracy at the level decisions are made, not at an aggregate that flatters.
- Make the model explainable
- A planner must be able to see why a number moved. Unexplainable output is unused output.
- Design the cadence, not just the tool
- S&OP is a decision process. Software supports it; it does not create it.
- Close the loop
- Compare plan to actual and feed the difference back, or the model never improves.
Common failure modes
- Forecasting everything. Intermittent demand does not respond to statistical forecasting and should be handled differently.
- Safety stock by blanket rule. Uniform weeks-of-cover overstocks stable items and understocks volatile ones simultaneously.
- Unconstrained supply plans. A plan that ignores capacity is a wish.
- S&OP as a reporting meeting. If no decision is made and no trade-off is accepted, it is a status update.
- Planning on ERP inventory alone when the warehouse knows better.
Next step
Talk about your planning environment
The useful conversation is usually about which decisions are being made badly, not which tool to buy.