Back to Case Studies
A century-old retail merchandising leader puts vision AI in the hands of its field associates
Finding planogram drift and hidden out-of-stocks in the aisle, with guidance associates can act on before they leave the store
Case study at a glance
- Customer: The nation’s leading in-store retail services provider, with more than a century in business
- Industry: Retail and consumer packaged goods
- Challenge: Shelf problems that cost sales went unseen until after associates left the store
- Solution: Seekr explainable AI and vision language models (VLMs), deployed at the edge
- Results: Rolling out to more than 4,000 associates, with every store visit an AI-verified, auditable record
The customer
The customer is the nation’s leading in-store retail services provider. It began more than a century ago as a makeshift newsstand and grew into the largest distributor and merchandiser of entertainment products in the country, then expanded across grocery, health and beauty, hardware, impulse merchandise, gift cards, and consumer electronics.
Its business is the last 100 feet of retail: the shelves, endcaps, and aisles where purchase decisions happen. More than 4,000 full-time in-store associates execute that work for the world’s largest consumer packaged goods (CPG) brands across more than 40,000 retail locations.
The challenge
Out-of-stocks cost the global retail industry about $1.2 trillion a year, according to IHL Group. Add overstocks, the other side of the same broken visibility, and inventory distortion costs retailers about $1.73 trillion. The same research finds that 72% of out-of-stocks trace not to upstream supply chain failures but to in-store ordering and replenishment practices. The problem, and the fix, sit at the shelf.
That matters because the majority of brand-level purchase decisions are still made in front of the shelf. For the customer, the job has always been getting the right product to the right place at the right time, every time. What changed is the precision and speed that “every time” now demands.
For associates, the hard part wasn’t fixing the shelf. It was knowing where to look. A planogram can drift a few facings at a time, and an out-of-stock can hide behind a row that looks full. An associate who misses a problem may not see that shelf again until the next visit.
For decades, CPG brands have invested heavily to see how their products are actually executed in store. Traditional audits delivered that visibility after the fact, once the associate had moved on and the chance to fix the problem had passed.
The Approach
An enterprise AI partner, starting at the shelf
The customer named Seekr its enterprise AI partner and chose the last 100 feet of retail as the starting point.
Vision language models with retail-specific intelligence
Seekr’s VLMs see what associates see. They spot where a planogram has drifted and where an out-of-stock is hiding behind a row that looks full.
Deployed at the edge, where the work happens
The applications run in associates’ hands, so guidance arrives while the associate is still in the aisle.
Prioritized, explainable guidance
The applications show associates which fix will protect the most revenue, with clear, actionable guidance on where to focus, so they can act before leaving the store.
An auditable record of every visit
Each store visit becomes an AI-verified record of what was right, what was wrong, what was fixed, and why, giving the CPG brands the customer serves insight, accountability, and confidence that their products are represented exactly as intended.
The results
Seekr’s solution, built on SeekrFlow is rolling out to the customer’s more than 4,000 associates. The customer expects it to significantly improve merchandise quality resolution, with associates identifying and correcting issues in real time, the first time.
The rollout delivers:
First-visit resolution: More in-store issues found and corrected on the initial visit instead of requiring a return trip.
Verified execution: Every shelf becomes a measurable performance surface for the CPG brands the customer serves.
Service margin: Less administrative overhead, so associates spend more of each visit on judgment work.
Why it worked
The customer started from its associates’ expertise, not from the technology. It considers its associates the best in the industry and wanted technology that matches that expertise. The AI sees what associates see and tells them where to focus, which is the difference between auditing a shelf and improving it.
It also put the AI where decisions actually get made. The edge has long been where mission-critical decisions happen in industrial operations and defense. Retail execution works the same way: the decision is made in the aisle, so the analysis has to be there too.
What’s next
The shelf is the starting point of a broader enterprise AI partnership. With every store visit producing verified, structured data, the customer can deliver in-store visibility to the brands it serves at a scale that wasn’t possible before.
Accelerate your path to AI impact
Book a consultation with an AI expert. We’re here to help you speed up your time to AI ROI.
Request a demo
