Analyst Insight: Consumer packaged goods companies strive for supply chain optimization, but many still face a fundamental challenge: accurately predicting true shopper demand. Modern retailers provide detailed trading data, but over-reliance on sales figures alone can distort your picture of the market. True demand can’t be measured directly — it must be inferred from met demand, plus an estimate of unmet demand. CPGs that master this inference will achieve superior planning, execution, and profitable growth.

Achieved sales only hint at true demand, but fail to capture it completely. Sales volume data reflect what was actually sold, not the potential volume if shopper needs were fully met.

This gap, unmet demand, stems from friction throughout the supply chain. Demand is missed due to:

• Availability issues. Sales are throttled when products are out of stock in-store, unavailable at depot or subject to manufacturing or sourcing issues. Even when stock is present, demand can be missed if there is no stock on the shelf, stock is damaged, or the shopper can’t find the available items.

• Distribution and ranging. Demand is missed where stores choose not to list a product (by design).

• Competitive pressures. Competitor activities, such as promotions or price reductions, can persuade shoppers to try alternatives, temporarily cannibalizing demand for a CPG’s product.

Using relevant supply and demand signals, one can estimate the counterfactual "true demand." This presents an alternative to recorded facts, modeling what could have been sold had supply limitations and negative external factors been removed, which is especially valuable during periods of promotional activity.

Historic demand decomposition provides a long-term breakdown of achieved sales volumes to explain the underlying “base” demand, factoring in long-term trends, seasonality, events, stock shortages, promotional uplift and cannibalization.

The true demand challenge requires a foundation of rich data and sophisticated analytical techniques, often housed within a demand signal repository (DSR), to enable counterfactual simulation. With a digital twin for each product, it is possible to map actual sales against various demand drivers to replay achieved performance and explore what true demand might be, adjusting for variables such as availability, promotions, and seasonality.

The process of historic decomposition utilizes these true demand indicators to produce more accurate demand forecasts. A detailed variance analysis generates a robust and realistic model of baseline demand (sales volumes expected without promotional activity) before incorporating factors such as price elasticity, depth of cut, and distribution. 

Modern demand intelligence solutions leverage machine learning and artificial intelligence to rapidly build accurate baseline models from rich, detailed data. Machine learning and AI are increasingly incorporated into DSRs to discover, highlight, and automate best practices, moving CPGs beyond simple spreadsheet analysis towards predictive analytics.

With the emergence of agentic AI, CPGs can now assimilate hundreds of analyses of past promotions and current conditions to forecast both true demand for upcoming events, such as promotions, new product launches, or price changes, and achieve uplifts, based on historic execution limitations.

Resource Link: https://www.gocrisp.com

Outlook: Rapid, regular forecasting will help CPGs respond quickly to demand shifts in 2026. Utilizing daily demand intelligence, CPGs can identify when demand exceeds plans and act rapidly to pull forward future planned shipments, adjust production runs, and explore sourcing options. The result is true agile execution, whereby sensing demand shifts and automating needed supply responses, brands can maximize promotional results, maintain optimal inventory levels, and balance the risk of waste against lost sales opportunities.