How Can Retail Leaders Harness AI While Keeping Customers at the Heart?

AI can help retailers reduce costs, improve decisions, and personalize experiences at scale. What can leaders do to ensure those capabilities translate into genuine customer value rather than merely being expensive gimmicks?

For retail leaders, AI promises better forecasting, faster service, lower operating costs, and personalization at a scale that was previously difficult to achieve. Yet technological sophistication doesn’t automatically make a retailer more customer-focused. Automation can remove frustrating delays, or make it harder to reach a real person when one is needed. Personalization can be genuinely useful, or start to feel intrusive.  

With so many possible uses for AI, leaders must decide which investments matter most. A retailer struggling with product availability may need better forecasting before advanced personalization. Another may create more value by fixing recurring service problems than by adding recommendations at checkout. 

“Customer focus becomes an investment filter,” explains Margareta Glass, Managing Partner at Signium in Munich. “It helps leaders put investments in the right order: choosing technology that can advance the business now, keeping an eye on systems that may become useful later, and ruling out those likely to become expensive gimmicks.”  

The best way to use AI: Giving customers a reason to return  

Retailers invest considerable effort in attracting and converting new customers, but retention often receives far less attention. Gartner’s 2026 CMO Spend Survey found that awareness and conversion accounted for 62.6% of total media spending, while loyalty and retention received less than 15%.  

AI may widen this gap. It’s easier to measure whether technology helped make a sale than whether it helped build a lasting customer relationship. However, Gartner also found that the most AI-mature marketing organizations allocate a larger share of their budgets to loyalty and retention. 

For retailers, AI can help make the customer experience easier and more responsive. Practical applications include:  

• Using forecasts to help keep the products customers want in stock 

• Answering routine questions quickly and directing more difficult problems to the right person 

• Recommending products that reflect a customer’s interests and needs, rather than repeatedly sending generic offers 

• Connecting information across different channels so customers don’t have to explain the same problem each time 

• Identifying recurring complaints, returns or service failures so the retailer can address the cause 

Glass suggests that customer retention may be the true value of AI in retail: “Customers may not remember the AI system itself. They will remember whether the retailer understood what they needed and made the experience easier. Retail leaders must ask whether their use of AI is helping to create an experience that makes customers want to return.”  

How does Otto Group’s AI-first approach put the customer first? 

German retail and services company Otto Group has adopted an “AI-first” direction. In the Otto app, customers can explain what they need via text or voice rather than relying on conventional search terms and filters. The AI shopping assistant asks follow-up questions and draws on data covering more than 18 million products to recommend suitable options. 

The company says voice users average 11 interactions per session, compared with 2 for text users, while the average order value among assistant users is 37% higher than with conventional search. Otto has also developed a separate service assistant for questions about orders, deliveries, returns, and invoices. Routine inquiries can be answered within seconds, while more complex matters are transferred to human service employees. 

“Achieving that balance is one of Otto’s most remarkable wins,” says Glass. “Technology should support people, not replace them. Customers still value personal advice, empathy and authentic service. Leaders must decide where AI can handle an interaction, where it should help an employee and where a person should remain directly involved.” 

Stop collecting AI tools: Four questions to ask instead  

In a McKinsey and ICSC retail report, 45 of 50 executives interviewed had considered introducing an agentic commerce tool, yet fewer than five had a strategy agreed at board level. If the leadership team hasn’t decided what it wants AI to achieve, different parts of the business might head in different directions. Marketing may introduce one tool, customer service another and operations a third, without anyone considering how those choices fit together.  

The CEO and executive team must decide what AI is for, where to invest, and where to set limits. They don’t need to understand every technical detail, but they must know enough to judge whether each application is solving the problem it was introduced to address. 

Four questions can help retail leaders make those decisions: 

1. What customer outcome are we improving? 

Explicitly define the problem, the intended benefit, and how customers should experience the difference. 

2. Where must real people remain involved? 

Decide where AI can act, where employees should supervise it, and how customers can escalate a complex or sensitive issue to a real person. 

3. What needs to change for AI to work?  

Introducing an AI tool is seldom as simple as plug and play. Retailers must put the right foundations in place before, during and after implementation: 

• Reliable, accurate data for the AI system to use 

• Employees who are trained to use the tool 

• Clear responsibility for the decisions the system informs 

• New ways for teams to work together when the tool affects different parts of the customer experience 

4. How will we know whether it’s working? 

Leaders must decide in advance what each AI application is expected to improve. Across the business, success should be measured through stronger customer retention as well as cost savings and sales.  

Glass elaborates: “If AI saves money but leaves customers waiting longer or choosing not to return, it’s not delivering the intended value.” 

The golden rule of AI in retail: Focus on what customers need 

“AI is not only a technology transformation; it’s a leadership transformation,” concludes Glass. “To apply AI effectively, retailer leaders must make deliberate choices about where it can add real value and keep the customer need at the heart of those decisions.” 

That focus should also shape how retailers talk about AI. “Customers rarely care whether a retailer uses AI,” says Glass. “AI is just a tool. Retailers should position themselves as making customers’ lives easier. Consider a carpenter. He uses a hammer on every project, but he’s not going to talk about hammers all day. He’s going to talk about the quality of his finished product. Customers care about the finished result, not the tool used to create it.” 

AI is becoming a routine part of retail, but deploying more of it is not a meaningful measure of progress. The leadership task is to decide where AI has a place, and to ensure that the complexity behind the scenes makes the customer’s experience simpler.