AI/ML

The New Retail Tech Benchmarks

By Greg Buzek Nov 18, 2025 6 Min Read

Most conversations about retail tech benchmarks start in the wrong place.

Leaders ask: “What new technology should we buy?”

They research solutions. They compare vendors. They evaluate features. They build business cases around the latest capabilities.

And then they’re confused when the implementation disappoints.

Here’s why: They’re optimizing for the wrong metric. They’re shopping for features when they should be benchmarking against standards and true needs.

The organizations winning right now aren’t the ones with the newest features. They’re the ones hitting clear benchmarks—measurable standards for autonomy, efficiency, and business impact that define successful implementation.

What changed? Everything.

Do You Have What You Need for NRF? Let Us Help

Let’s look at this week’s new installation headlines. One overriding theme? The benchmarks used to evaluate retail tech are changing fast.

Today there are three core benchmarks that separate aspirational technology conversations from production-level implementations:

1. The Autonomy Benchmark

Autonomy means systems that act independently without human intervention or approval. It’s the difference between AI that “suggests” and AI that “does.”

This is a fundamental distinction.

Traditional AI systems are advisory. They analyze data. They make recommendations. A human sees the recommendation and decides whether to take action.

Agentic AI, which is now hitting production environments across retail, acts independently. It sees an opportunity and takes action without waiting for human approval.

In this week’s news, Google is building features that monitor prices across retailers and automatically purchase items when they hit your target price. No “email me to ask permission.” No “get this approved by a manager.” The AI sees the condition is met and executes the purchase.

Levi’s created AI agents operating inside Microsoft Teams that handle corporate tasks autonomously. These agents manage schedules, process requests, route approvals, and execute actions without asking for permission at each step.

Debenham’s Group deployed agentic AI to manage sales and pricing operations. The system monitors market conditions, competitor pricing, inventory levels, and demand signals, then adjusts pricing in real-time to optimize margin and velocity. Autonomously.

This is the autonomy benchmark: systems that act independently at production scale. And it’s arriving faster than most organizations expect.

2. The Efficiency Benchmark

Efficiency benchmarks are quantifiable. They have numbers. They’re replicable. They’re measurable. Not “this should save time” but “this saves 400,000 work hours per year.” Not “this might be faster” but “this reduces operational time by 85%.” Not “this could increase throughput” but “this processes 1,500 units per hour.”

Currys deployed electronic shelf labels across their store network. Traditional approach: store associates manually update prices every night. New approach: price changes push to digital labels instantly. Result: 400,000+ work hours saved annually.

Revisn, a hotel in Raleigh removed their front desk and replaced check-in with mobile-first kiosks. Instead of waiting in line for an associate, guests check in instantly through an app. The operational improvement: 85% faster check-in times.

A pub in Leeds installed self-serve beer tap systems. Instead of bartenders managing pourouts, customers pour their own drinks (and the system meters the amount). Throughput increased to 1,500 pints per hour. Several formats in the US already do this as well, seeing incredible profitability.

If you can’t measure it with data, it’s not a benchmark, it’s an expense.

3. The Impact Benchmark

But business impact is what really matters and we see this as well in this week’s news. 

Ahold Delhaize unified five brands (Food Lion, Stop & Shop, and others) under one Prism platform. The integration allowed them to consolidate operations, improve data quality, optimize inventory across the network, and enhance customer experiences across all brands simultaneously. The impact benchmark: serving 26 million customers per week.

Tennis Giant took a different path. They focused their integration narrowly: connecting their pro shops (physical locations) with their online sales channel through one unified system. The impact benchmark: over 100% sales growth in four years.

The impact benchmark asks: “What changed in our business because of this technology?”

If you can’t answer that question clearly, the technology wasn’t benchmarked to real business outcomes.

Join one of the most exclusive events NRF Weekend – Nearly 100 Retailers already signed up

Production-Level Benchmarks Hitting Retail Right Now

Here are the production-level benchmarks already hitting retail floors that hit the news this week.

Automation & Labor Displacement

Ulta Beauty is testing a system from 10Beauty that performs complete five-step manicures—removing old polish, filing nails, painting new color, full service—for $30 per service. It works. Customers use it. The economics work.

Delivery Speed Becoming Baseline

DoorDash has autonomous robots cruising the sidewalks in Miami delivering groceries and Old Navy launched their entire apparel collection on DoorDash for same-day delivery.

Superdrug partnered with Just Eat to deliver beauty products in 30 minutes. Thirty minutes from order to beauty products in your home. That saves a night out.

Target is piloting a sorting center in Cleveland designed 100% around next-day delivery.

These are production systems delivering measurable outcomes which are the new the baseline expectation. Same-day delivery isn’t a luxury anymore. Speed isn’t an innovation. It’s an expectation. Efficiency isn’t a nice-to-have. It’s a standard.

The Leadership Benchmark—Pushing Back on Vendor Price Increases

There’s a story in this week’s newsletter that deserves deep attention because it reveals a fundamental shift in how retail technology strategy is evolving.

A CTO at Kingfisher, the parent company of major European retailers like B&Q and Screwfix, faced a situation that many large retail organizations encounter.

Their existing SAP system was aging. A major software vendor approached them with an upgrade proposal. The pitch was compelling. The price tag was staggering. Their technology costs would increase by a factor of 20.

This CTO said no. “Don’t just hand me a price list that jacks up my costs 20 times. Show me the value.” That simple statement shifted the entire dynamic. It was asking the vendor to justify their premium based on actual business value, not just capability upgrades. When the vendor couldn’t credibly defend that 20x increase, the CTO forged his own path.

Here’s what he did:

First, he kept the legacy SAP system. It wasn’t the bottleneck. It was working. Replacing it wasn’t the problem. Maintaining it was manageable.

Second, he moved that system to Google Cloud. Modern infrastructure, managed services, reduced operational overhead. Significantly lower total cost of ownership than on-premise hardware.

Third, he hired a third-party support vendor, Rimini Street, to manage the legacy system. Professional support at a fraction of what the original vendor was asking.

Fourth, and most important, he took all the money he saved and reinvested it into building modern AI and personalization tools on top of the existing system.

He didn’t rebuild the foundation. He built the future on top of what was already working.

Better outcomes. Lower costs. More control. Custom solutions instead of vendor-dictated upgrades.

Sometimes it’s not about having the latest technology. It’s about having the confidence and analytical clarity to push back on vendor proposals, evaluate alternatives, and make decisions based on actual business value rather than vendor roadmaps, recognizing that you don’t need the vendor’s upgrade path to stay competitive. You may be able to keep legacy systems, optimize their infrastructure, invest in modern tooling on top, and achieve better outcomes for less money.

The Question That Changes Everything

“Stop Asking About Features. Start Asking About Benchmarks.”

With all this change happening so fast, with agentic AI and new platforms and new delivery models, what’s the one benchmark that your business really needs to hit right now?

It’s a different question than most organizations are asking.

Different organizations have different benchmarks they’re behind on. But asking the question… really asking it…honestly assessing the answer…is where better technology strategy starts.

Retail technology benchmarks have fundamentally changed.

Efficiency isn’t about hoping technology helps, but rather implementing systems where you can point to specific metrics and demonstrate impact.

Explore IHL Group’s latest research on retail technology adoption, shelf intelligence, and inventory management. We track what’s actually being deployed across retail segments, not what’s in beta or planned for someday.  Several free or paid research reports to choose from. 

Visit ihlservices.com to dive deeper into retail transformation, or explore the complete analysis in the latest episode of Retail Reality Check: