How retailers are missing AI’s biggest opportunity: Operational efficiency

(Source: Inside Retail)

Artificial intelligence discussions in retail are largely focusing on customer-facing applications such as personalisation, chatbots and recommendations. But the greatest opportunities may sit further behind the scenes, where AI can improve the operational decisions that keep retail businesses moving.

Michael Guantiero, VP and head of solution consulting, Asia Pacific, Japan and India, with Infor, talked with Amie Larter of Inside Retail for her podcast series Retail Untangled, where retail industry experts share business hacks that help them succeed.

He stressed that retailers need to look beyond visible AI applications and focus on connecting the operational backbone of their businesses.

From the customer experience perspective, “the value is not so much in the chat box,” he continues, rather the daily operational decisions that keep the business moving.

“That is where we see the value: How we help improve operational aspects and connect the information flow behind the scenes. It is about shortening the time between the signals and the action that needs to be taken, which, behind the scenes, helps improve the overall experience.”

While many retailers have invested in AI capabilities, Guantiero said fragmented systems and inconsistent data are preventing businesses from achieving the full benefits.

“We are seeing pockets of AI, but disconnected systems and uneven data make it a little bit difficult to bring it together,” he said.

“Most have some AI capability, but because of these disconnected systems, it is hard to really get the efficiency companies are looking for. Teams can see the issue, but they cannot really act fast on it to get to where they need to be.”

Data fragmentation has been a long-running challenge for retailers, with enterprise resource planning, warehouse management systems, supply chain and finance platforms often operating separately.

Guantiero said progress is being made, but retailers do not necessarily need large-scale technology replacements to unlock improvements.

“It is improving, but it is just not improving fast enough,” he said.

“You do not need a rip and replace; it is about how quickly we can identify areas where we can connect the highest-value data points and workflows to really get some significant outcomes and productivity gains.”

AI can improve forecasting

One area where AI is already showing clear potential is forecasting and replenishment, particularly as retailers continue to manage the challenges of excess inventory and stock availability.

Guantiero said the key is connecting customer demand signals with operational execution.

“Historically, you would say the customer experience and the demand picture have always been separate from the replenishment aspect of it,” he said. “Now with the pace of technology and AI, we are bringing that picture together, which means data needs to be more aligned, so we can optimise what actually happens in that flow.”

However, Guantiero warned that AI cannot solve underlying operational problems without the right foundations.

“AI will not fix that on its own,” he said.

“There was this initial perception that we would dump everything into a data lake or data warehouse, and AI would figure it all out. What AI does not know is the context and the dependencies that you have in your overall supply chain process to make sense of it in a proper way.”

The rise of agentic AI, where systems can sense information, make decisions and take action using available tools, could accelerate operational improvements. But Guantiero said retailers need trusted data and clear business rules before deploying these technologies at scale.

“Agentic AI senses, and it kind of decides and acts based on the tools it already has available,” he said.

“In retail, it only really works if it is grounded in trusted data and clear business rules. Every decision that touches inventory, margin, fulfilment and the customer experience has to be fully connected.”

When pilots don’t deliver

For retailers looking to move from AI pilots into wider adoption, Guantiero said governance and practical use cases will be critical.

“The challenge that some companies are facing is that they are running a pilot. But the pilot is not necessarily representative of what the real world looks like,” he said.

“It is about identifying use cases that truly represent how the business runs, something that is scalable.”

As customer expectations continue to rise, Guantiero said the impact of better-connected operations will ultimately be felt through improved service levels, which is where Infor measures performance.

“When you walk into a store today and you want to buy a dress or a blouse, if they do not have it in the store, your next question is do you have it in another store you can transfer?” he said.

“If a store were telling you today, ‘I cannot see that’, you would look at them as if that was strange.”

Looking ahead, Guantiero believes the pace of change will continue to accelerate as agentic AI creates new ways for retailers to operate.

“Everybody is expecting things to happen a lot faster,” he said.

“I think the evolution is the agentic era is bringing this to life more easily, and it is creating new business models faster than ever anticipated. We are still figuring out what these new business models are and what the new retail model is that brings this to life.”

  • Listen to the full podcast online to hear Guantiero and Larter unpack the opportunities AI brings to retailers in this region, as well as what they might be consistently missing.

bci-logo

Asia's Top 100 Retailers, ranked by brand conviction. The results might surprise you.

Recommended By IR

You have 7 articles remaining. Unlock 15 free articles a month, it’s free.