Ask This Question When Investing in an AI-Powered Retail Operations Platform

Retail is the most complex it's ever been and a multitude of factors have pushed business leaders to prioritize artificial intelligence implementation at HQ and in-store.

What I hear now from retail leaders is an urgent mandate to get their workforce AI-enabled. The data backs this, with 97 percent of retail organizations planning to increase spending in this area over the next year.

I believe that retail leaders can use AI to address many of the major constraints of the moment. What's missing is a reliable way to tell whether a tool will solve your specific problems or manufacture new ones.

Here’s how to ask the right questions to find out.

Rethinking Key Evaluation Questions

Early on, the promise of AI was that its outputs would lead to faster teams. That meant, in evaluating new AI-powered solutions, retail leaders would usually ask questions like, "What can this tool do? How fast can it help us go?"

However, for retailers, the most important question is no longer what an AI tool can do or make, but instead: Can this AI solution understand the context of our business well enough to improve how work gets done?

The 'Data Swamp' Challenge

The challenges frontline teams face aren’t generic. An AI feature that looks impressive in a demo but fundamentally misunderstands how your business operates will produce the same wrong answer over and over.

AI should either help you make a decision or help your frontline teams act on one.

Tools can be trained on documentation. However, if the technology isn't learning from your operations, then it’s only as good as the documentation it’s been handed. If an employee handbook is five years old, fine. If a returns policy changed last month and the tool doesn’t know it, frontline teams are left amplifying bad information to customers.

The same failure shows up in prioritization. A tool ranking tasks without knowing your business will sort by what it can see: due dates, task age, whoever submitted it. It can't see that the endcap reset matters this week because the promo drops Thursday, or that the safety notice is the one thing a district manager will ask about. Therefore, it surfaces a non-urgent reminder and buries the thing that had to happen before opening.

Google Cloud called this phenomenon a “data silo” or “data swamp.” It writes that "many retailers struggle with fragmented, low-quality data that prevents AI from providing accurate, real-time insights or making effective decisions.”

AI That Improves How Retail Work Gets Done

If the best question to ask is whether an AI tool improves how work gets done, your next question is how.

Every retailer’s how will be different based on their operational needs, but the criteria should center on the following:

• Context — does it know your business? Can the solution use what it knows about how you operate to deliver relevant guidance and support?
• Specificity — does it know who it's talking to? Can the solution tailor that guidance and functionality to the employee’s role, location, permissions, and situation?
• Real-time data — does it know what’s current? Can it access and support work based on up-to-date information about scheduling, inventory, communication, and other operational data?
• Measurable outcomes — can you prove it worked? Can you connect the work that the tool supports to improved operational outcomes like better execution, higher reporting accuracy, and increased employee engagement?
• Improvement — does it get better on its own? Does the AI solution you’ve chosen get smarter, learning from how you operate and scaling over time?

For retailers developing their AI investment plans, the best approach is to prioritize tooling that uses real conditions within your business to solve specific operational retail challenges. The right features, integrated into your relevant business context and retail operations, can make all the difference.

Melissa Wong is the CEO and co-founder of Zipline, the AI-powered operations platform for field teams with brick and mortar brands.