Balancing the Risks & Rewards of an Agentic System with the 5/15/80 Model

As previously shared (here and here) as part of the launch for Agentic Systems for Commerce (ASC), the current state of digital commerce presents fundamental challenges that brands must confront:

 

• Rising costs without growth: Managing millions of dynamic shelves and escalating retail media spend has created spiraling costs, but not incremental demand—forcing brands to do more for the same shoppers.

• Structural inefficiencies: Legacy operating models—fragmented teams, disconnected tools, and misaligned incentives—cannot keep pace with the machine-speed demands of real-time commerce.

• Agile competitors seizing share: Private labels and marketplace sellers are growing faster by focusing narrowly, moving leaner, and exploiting digital shelf opportunities where incumbents are spread too thin.
   

The 1% challenge

Despite the growing importance of winning in digital commerce channels, most brands are actively managing only a fraction of their digital commerce operations at any given time:

 

• Updates to product page content are infrequent and typically cover just 10% of their portfolio

• Digital commerce teams and investments are generally focused on just the top 3-5 retailers

• Execution playbooks center almost entirely on product detail page (PDP) optimization, while dozens of other critical performance opportunities remain unaddressed
   

We call this the 1% Challenge because most brands are, at best, actively managing just 1% of their total commerce operations. Meanwhile, smaller, digitally native brands are systematically exploiting these blind spots at scale by adjusting product listings in real time, conquesting competitive out-of-stocks, and responding to emerging demand signals as they happen.
 

This challenge is allowing these smaller brands to capture disproportionate market share while established brands struggle to keep pace. Solving the 1% challenge has become critical for legacy brands and requires leaning heavily into automation for commerce operations.
 

Learnings from other industries

The path forward isn't unprecedented. The supply chain and manufacturing functions made this transition over 20 years ago, which I was fortunate to witness firsthand while leading the design and implementation of HEB’s supply chain system during that period.

​

Back then analysts were managing around 120,000 products weekly for warehouse orders, which had become impossible to do effectively. They had hit the "systems inflection point" for supply chain operations, as so many had across other industries.

 

The one consideration we had to plan for was that no system is perfect. Making a mistake with a high-volume product, such as Coke or Pepsi, could result in an extra truckload of it, which wasn’t acceptable. Making a mistake with a long tail product, such as crushed red pepper, may mean a store receiving an extra case of it, which wasn’t a concern. Finding the balance between the benefits of automation while managing risk intelligently was the key to the success of the system.

​

Just as supply chain operations reached a systems inflection point 20 years ago, commerce is now at the same juncture—requiring a model that balances risk and reward at scale.

​

The key is balance. Successful automation requires a model that thoughtfully balances the risks and rewards, ensuring human oversight where it matters most while allowing systems to handle routine operations at scale.

​

The 5/15/80 model

The model used then and still applicable today is the 5/15/80 model. This structured approach ensures experts are enabled and in charge of the highest value portion of the product portfolio, while the system leans heavily into automation in the long tail, creating the optimal balance between system enablement and automation.

 

• Top 5% of products: High-value SKUs should be managed continuously with expert-driven, system enabled strategies.

• Next 15% of products: Secondary SKUs should be periodically managed in an expert review & approve model based on system recommendations.

• Bottom 80% of products: The long tail should be fully automated with system optimization, ensuring no revenue opportunity is missed.