Enterprise Intelligence Group
CPG demand forecasting: a larger order is not automatically stronger consumer demand
Before changing a CPG forecast, separate consumer sales, customer orders, shipments and network inventory. Use five practical checks with commercial and planning teams.
A customer can order more while shoppers buy the same amount. Your planning process needs to distinguish the two.
For a CPG business, sell-in records sales into a customer or distribution channel. Sell-through describes onward sales, often retail sales to consumers. Customer orders and supplier shipments provide additional signals, but they are not interchangeable with either measure. Agree the definitions used in your business before combining them.
A customer order increase might reflect a promotion, a new listing, an inventory build, a delivery-calendar change or a response to earlier supply disruption. Treating every increase as a lasting change in consumption can distort the next planning decision.
Four aligned evidence panels
Not a causal flow — four signals to put alongside one another for the same item, account, unit and time period.
Consumer sell-through
Onward sales, often retail sales to consumers.
Customer orders
What the customer or distribution channel has ordered.
Supplier shipments
What has actually shipped against those orders.
Network inventory
Stock already held downstream in the network.
Same item, account, unit and time period.
For a retail-focused view of how demand estimates connect to planning and replenishment, see Retail demand forecasting: connect the forecast to planning and replenishment.
Five checks before changing the forecast
1. Align the evidence. Compare the same item, customer, unit and period. Check pack conversions, data latency, cancellations and returns. A case-based shipment total cannot be compared directly with an each-based consumer sales series.
2. Separate the signals. Put consumer sell-through, customer orders, shipments and network inventory alongside one another. Mark unavailable customer stock or POS data as missing, not zero. Where coverage is partial, state which customers and channels the evidence represents.
3. Follow the promotion beyond the order. Distinguish the retailer's loading period, consumer event and post-event stock position. An early shipment supports execution; it does not by itself establish incremental consumption. Ask whether the next order reflects demand or stock already held downstream.
4. Expose commercial assumptions. Record distribution gains, listing dates, customer commitments and expected event participation. Name the commercial owner and the evidence behind each adjustment. Keep an aspiration separate from a confirmed order or agreement.
5. Make one coordinated decision. Customer teams, demand planning and supply planning should agree what has changed, which forecast or order assumption needs revision and who approves the response. Preserve the earlier forecast so the intervention can be reviewed using consistent definitions.
These are operating requirements, not an argument that existing software cannot support them. For example, RELEX describes combining retail POS and sell-in information with account-level planning. The consulting question is how the available signals, configuration and responsibilities work in your business. RELEX Demand Planning & Sensing reference Platform references describe published capabilities, not EIG partnership or endorsement.
A useful output from the next account review
Produce a single decision record: account and item group; observed change; proposed explanation; missing evidence; inventory implication; approved action; owner; review date.
AI can prepare that record from authorized sources and flag inconsistencies. It should not create customer commitments or fill an unseen inventory position with a confident estimate presented as fact.
EIG’s Forecasting & Replenishment Performance Review examines a defined workflow and turns the findings into prioritized improvement work.
Bring one planning challenge to EIG
Discuss a planning performance review
The EIG Library · Insights · AI Maturity Assessment · Contact