How to Build a CPG Retail Sales Forecast for a New Launch
A useful CPG retail sales forecast starts with the number of stores expected to sell each SKU, the weeks those stores will be active, and a realistic units-per-store-per-week assumption. Build it by SKU and retailer before combining totals. This keeps a distribution ambition from becoming an unsupported revenue forecast.
The basic retail demand equation
Expected retail units = active stores × selling weeks × units per store per week. Use a weekly schedule when stores open in waves. A chain authorization for 200 locations is not the same as 200 locations selling for a full year.
Illustrative 13-week launch: 100 stores sell one SKU for all 13 weeks. At 3 units per store per week, expected movement is 3,900 units. At 5, it is 6,500. At 7, it is 9,100. These are scenarios, not category benchmarks.
Choose a defensible velocity assumption
Start with comparable results from the same brand where possible. Match retailer format, package size, price, geography, and promotional support. If the evidence comes from a very different channel, explain the limitation and widen the forecast range. Avoid selecting the best launch week as the ongoing baseline.
For a new item with no history, write down the assumption and the date you will revisit it. Separate regular-price movement from a temporary promotion lift. Do not apply a strong event week across the entire year.
Model a rollout instead of multiplying by 52
Create one row per SKU per week with planned live stores, baseline velocity, event adjustment, and resulting retail units. For example, 50 stores in weeks one through four and 100 in weeks five through thirteen create 1,100 store-weeks. At five units per store per week, demand is 5,500 units, not the 6,500 implied by a fully live 100-store launch.
If velocity already reflects normal stockouts, applying another availability discount can count the same loss twice. Decide whether the model estimates observed sales or unconstrained demand and label it accordingly.
Keep shipments separate from consumer sales
The first distributor order includes inventory to fill the pipeline. It is not proof of consumer demand. Build a separate inventory bridge: opening inventory plus receipts minus expected sales equals ending inventory. Then calculate replenishment needs against the desired ending stock, accounting for case packs, order timing, and lead times.
Similarly, shelf-price retail dollars are not the brand's revenue. Estimate brand revenue using the brand's invoice price and applicable deductions. Use the margin waterfall to connect those measures.
What should the buyer see?
- Projected retail units and dollars by SKU for the requested period.
- Store counts, rollout timing, and the evidence behind velocity.
- A low, base, and high scenario with consistent assumptions.
- The sampling or promotional support associated with the base case.
How often should the forecast change?
Review actual movement weekly during launch and reforecast when distribution, availability, or velocity meaningfully diverges. Preserve the original forecast so forecast error remains visible. A forecast should support inventory decisions and buyer conversations, not be rewritten to make every result look planned.
Use our velocity calculation guide to keep the inputs consistent.
Put this to work for your brand. CPG Consulting helps emerging brands turn retail data and commercial plans into usable sales tools. Explore our services or book a call.