Albertsons
AI demand forecasting for fresh and automated shelf replenishment
Albertsons completed in October 2025 the rollout of Afresh's AI fresh demand forecasting solution across the fresh departments of its 2,257 stores; the platform claims typical results of 25% less shrink, 3% more sales, 7% higher inventory turns, and more than 100 million pounds of food waste avoided.
Objective
Predict demand for each fresh product department by department to order the right quantity, maintain availability, and reduce spoilage on short-shelf-life products.
The deployment
Albertsons Companies operates 2,257 stores in the United States under the Safeway, Albertsons, Jewel-Osco, Shaw's, Vons, and ACME banners. Since 2022, the retailer has been deploying Afresh's AI solution in its fresh departments. The system factors in promotions, featured displays, seasonality, and holidays to predict demand and recommend order quantities to store teams, on products where spoilage is structural. In October 2025, Afresh announced the completion of the rollout across all of the network's fresh departments (bakery, deli, meat, seafood, produce), complemented by a DC Forecast component that gives buyers daily forecasts at the distribution center level.
Results Proof C
Deployment documented by an Afresh press release naming Albertsons (quotes from Albertsons' Chief Merchandising Officer) and by established trade press (Grocery Dive). The shrink, sales, and turns percentages are aggregate typical results for the Afresh platform, not isolated for Albertsons, hence no level B on impact.
How it works
Documented architectureThe stack in detail
- plateforme Afresh Fresh Replenishment SaaS demand forecasting and replenishment solution dedicated to fresh departments.
- outil Afresh DC Forecast Component that gives buyers daily forecasts at the distribution center level.
- llm Modeles ML de prevision frais Afresh's proprietary machine learning incorporating promotions, featured displays, seasonality, and holidays; the exact algorithm is not published.
- infra Flux ventes et inventaire Albertsons Connection to sales and fresh inventory data from the 2,257 stores, a condition for reliable recommendations.
How it runs, concretely
For ops teams-
1Fresh data ingestion Afresh platform
Sales, inventory, promotions, featured displays, seasonality, and holidays feed the model by department.
-
2Demand forecasting AI model (Afresh)
The model predicts demand for each fresh product over the replenishment horizon.
-
3Order recommendation fresh department manager
The system proposes an order quantity to the department manager, who approves or adjusts it.
-
4Distribution center forecasting buyer / Afresh platform
DC Forecast provides buyers with daily forecasts to set upstream orders to the stores.
Predicted fresh demand by product and department. If fresh inventory is mis-recorded or sales data is noisy, the recommendation becomes wrong and spoilage or stockouts rise.
How your customers perceive this type of use
Sourced studiesC'est la famille la moins acceptee : 68% des Americains jugent inacceptable un score financier personnel calcule par algorithme et 67% l'analyse video automatisee d'entretiens d'embauche (Pew Research, 2018). La demande d'explication et de recours est massive : 83% veulent savoir quelles donnees l'IA utilise et 91% veulent pouvoir corriger des donnees erronees (Consumer Reports, 2024). A l'echelle mondiale, seuls 46% se disent prets a faire confiance aux systemes d'IA et 70% jugent une regulation necessaire (KPMG / Universite de Melbourne, 2025).
Acceptance conditions
- Transparence sur les donnees utilisees : 83% des Americains la reclament (Consumer Reports 2024)
- Droit de correction des donnees erronees : 91% le demandent (Consumer Reports 2024)
- Explication de la logique de decision : 44% des consommateurs sont plus enclins a utiliser un agent IA si sa logique est clairement expliquee (Salesforce 2024)
- L'acceptabilite depend du contexte de la decision : 50% des Americains jugent equitable un score de risque criminel pour la liberation conditionnelle, contre 32% pour un score financier applique aux consommateurs (Pew Research 2018)
Red lines
- La decision opaque et sans recours sur l'emploi, le credit ou le logement : 45% tres mal a l'aise pour l'embauche, 39% pour le pret, 39% pour le logement (Consumer Reports 2024)
- Le scoring des personnes a partir de donnees comportementales : 68% le jugent inacceptable pour les offres financieres (Pew Research 2018)
Sources: Pew Research Center 2018 · Consumer Reports 2024 · KPMG / Universite de Melbourne 2025 · Salesforce 2024
How to replicate
Inference, not sourcedData prerequisites
- reliable fresh sales and inventory data by department and store
- promotions and featured-display calendar
- shelf-life attributes for fresh products
Org prerequisites
- department managers ready to follow order recommendations
- a disciplined fresh inventory process
Possible stack
- Afresh
- RELEX, Blue Yonder, Invafresh (specialized fresh forecasting)
- in-house forecasting
The plan, step by step
- Step 1Make fresh inventory reliable for a high-spoilage category (produce, for example) and connect the sales data.Deliverable: Clean sales and stock data for the pilot category
- Step 2Configure the platform: fresh catalog, promotions, featured displays, product shelf lives.Deliverable: Forecasts computed in shadow mode, compared with actual orders
- Step 3Launch the in-store pilot: order recommendations approved or adjusted by the department manager.Deliverable: Shrink and availability measured against control stores
- Step 4Extend to the other fresh departments and stores, train the teams, add forecasting at the distribution center level.Deliverable: Rollout plan + shrink and turns dashboard by department
First step: Make fresh inventory reliable for a high-spoilage category (produce, for example), then connect a forecasting engine that recommends order quantities to the department manager.
The same pattern, proven in other industries
Every Man Jack
predictive replenishment timing (predicted next order date)
Maersk
predictive monitoring of refrigerated cargo (customer visibility)
T-Mobile US
Intent-driven decisioning platform: an engine reads customer intent and sentiment in real time, decides the next action, and can execute it, beyond a simple chatbot
Sources
- S1 Afresh Completes AI-Powered Fresh Replenishment Roll Out Across All Albertsons Companies Fresh Departments Interested party archive pending
- S2 Afresh expands AI-powered tools for fresh Established press archive pending
- S3 Afresh Completes AI-Powered Fresh Replenishment Roll Out Across All Albertsons Companies Fresh Departments (Afresh) Interested party archive pending
An error, newer info, a source?
This page lives on its accuracy. If a figure has moved, if the deployment has changed, or if you have a higher-quality source, tell us. Every sourced correction is verified before publication.