Customer Story

How a Southwest US regional grocer built shelf intelligence into daily operations

A 100+ store, full-line supermarket chain with a community-store focus — and a Dunnhumby comparison that changed the merchandising team’s minds.

Customer
Southwest US regional grocer
Scale
100+ stores, full-line supermarket
Entry use case
Promotional demand forecasting

The problem

The chain’s merchandising team was running weekly promotions across dozens of categories with forecast accuracy that varied wildly — sometimes 60% under, sometimes 300% over. Store-level allocation was chain-forecast divided by store count, which produced simultaneous stock-outs at high-velocity locations and markdowns at low-velocity locations across the same promotion.

Existing analytics investments (including a large incumbent retail-BI platform) produced dashboards but didn’t close the loop back into the next planning cycle. Post-promo reviews were arguments about weather and execution, not decompositions of what actually happened.

Why Akuret

The team piloted Akuret’s promotional demand forecasting alongside the existing incumbent. Within two months the merchandising leadership reported that Akuret’s per-store forecasts were consistently tighter than the incumbent’s on the same promotions — and, crucially, decomposed the forecast variance in a way that made the next-cycle correction obvious.

The customer shared a full year of historical sales data to support the model, indicating a level of trust that only builds when a vendor’s outputs are consistently better than incumbents on the customer’s own data.

What changed

  • Per-store allocation replaced chain-average allocation for the SKUs Akuret modeled.
  • Post-promo attribution moved from an argument to a data-driven review, with variance decomposed into baseline drift, lift model error, cannibalization, and execution gap.
  • The merchandising team gained a defensible answer to CPG partners asking about promotional execution.

“Their forecasts held up better than the incumbent’s on our own data. That is what got our attention.”

— VP of Merchandising, Southwest US regional grocer

Reported outcomes

Customer-reported outcomes across the Akuret customer base include:

  • Sales lift of 1–3% per store on categories where Akuret is applied
  • 1.5–3 hours of daily labor saved per store on manual shelf checks and audit work

All figures above are customer-reported and vary by chain, category, and rollout scope. Not audited external claims.

Why we don’t name this customer publicly

Our customers view their operational intelligence as competitive advantage. Where a customer has authorized public naming, we say so. Otherwise we publish anonymized studies to protect their edge — because the moment a competitor sees how you improved, the advantage narrows.

Want the same result in your stores?

A 30-minute walkthrough — no slides, real data.

Request a demo