The Key to On-Shelf Availability? Consumption-Based Forecasting

Oct. 9, 2023

By Manish Ghosh, Blue Yonder Corporate Vice President, Industry Strategy

Recently I saw some disturbing statistics. While CPG companies boast on-time, in-full (OTIF) delivery rates of 93.2% to their retail customers, they’re only achieving on-shelf availability (OSA) rates of 75.4%.

We’re all familiar with the dangers of having stockouts at the shelf level. If CPG companies don’t have the right products on the shelf, shoppers have plenty of other options today, including competitors’ products or private-label products — which accounted for sales of $199 billion in 2022.

According to recent NIQ data, private-label food sales increased 16% from 2020 to 2022. And their growth is only accelerating. In February 2023, private-label sales rose by 13%, outpacing an overall market growth rate of 5%.

How can CPG companies combat the private-label threat? Obviously a critical element is having their products available at the shelf. And that means addressing the glaring disconnect between the products they’re successfully delivering and the products that consumers are actually looking for.

Why Traditional Forecasting Methods Fall Short

The problem is that traditional forecasting approaches lack visibility to the store shelf — and they don’t reflect what’s happening there. Most CPG organizations are forecasting demand based on data about shipments into the distribution center (DC), or from the DC to the store.

But that doesn’t tell the full story about what happens when products land on the shelf. Individual stores have mixed results — overstocks on one SKU and out-of-stocks on another. And there are also variations across stores, based on local differences in consumption. A forecast based on aggregate shipping volumes will never capture and reflect the lost opportunities, store by store, that are typically captured by private-label products or competitors.

CPG companies are also using point-of-sale (POS) data for forecasting, but that’s an equally flawed approach. Transactions at the cash register don’t reveal the choices and trade-offs made by shoppers in the aisle, the products available to them at the time of sale, or the role of stockouts in influencing the ultimate buying decision.

Today’s retail aisle is incredibly fast-moving and fast-changing. Forecasts based on historical shipments or POS data offer only a limited perspective of today’s complex consumer behaviors.

Consumption-Based Forecasts Are 30% More Accurate

So what’s the alternative? Today’s volatile and localized markets demand an approach that considers current real-world consumption patterns, not shipping data or POS data.

CPG companies need a way to understand actual consumption-based demand at the level of the individual store — not just for their products, but for competitive products, including private label. They need to forecast at the SKU level by understanding consumption at the store level. They also need to make shelf adjustments — such as aligning their assortments and space plans with store-specific layout and preferences — outside the bi-annual planogram generation process.

To ensure availability, this consumption-based forecast needs to be shared back through the supply chain, so replenishment processes are accurate and fast. The supply chain needs to be driven by actual shopper behaviors at the store level.

According to research by Pensa Systems, consumption-based planning methods result in a forecast that’s 30% more accurate than one generated by POS data alone. That’s an enormous difference, and one that’s certain to drive a dramatic improvement in today’s availability rates that hover around 75%.

How Do We Get There? Technology Is the Only Solution

While CPG companies face a difficult challenge in matching product availability to fast-changing, localized demand, the good news is that today’s advanced technology solutions are up to this challenge. The level of localization, insight and responsiveness described here might sound impossible, but there are robust solutions in category management, demand planning and replenishment that are designed to deliver it.

Enabled by artificial intelligence (AI) and machine learning (ML), today’s new generation of retail planning and forecasting solutions are capable of ingesting data about SKU-level and store-level consumption, provided by a third party specializing in retail-aisle observation. This data includes SKU- and store-level availability, planogram and display compliance, shelf share, competitors’ product availability, category adjacencies, product positions, and changing consumer behaviors and availability levels based on time of day, day of the week, promotions and other attributes.

After AI- and ML-enabled planning and forecasting engines ingest this data, they apply proprietary algorithms to transform all those granular insights into an extremely accurate, forward-looking forecast. Category management solutions transform the insights into highly accurate space plans that capitalize on every inch of space in every store. Dynamic replenishment planning engines ensure a continuous flow of the right products to the right location, driving high availability.

The Result: Full Category and Location Visibility

By adopting a new, consumption-based forecasting approach, CPG companies can see exactly how their products are performing, store by store, shelf by shelf, and SKU by SKU — as well as how their competitors, including private labels, are performing. By arming themselves with full category and location visibility, CPG companies can maximize their inventory turns and sell-through rates — as they match their high OTIF performance with equally impressive OSA performance.

While this might sound futuristic, advances in data science and data availability, forecasting algorithms and cognitive planning processes place this capability within the reach of every CPG company. Pensa Systems data show this approach has been proven to increase sales by 12% per year as CPG companies learn to intelligently course-correct their shelf plans to reflect changing real-world shopper behaviors.

The private-label threat is real and growing for the world’s consumer packaged goods companies. But so is their ability to counter this challenge by adopting a new, consumption-based approach to forecasting that’s designed to provide visibility to the retail shelf.