From quick wins to securing executive support, tech leaders from top retailers in Australia and New Zealand unpack their AI journeys and how they plan to leverage a technology that is expected to profoundly transform retail in the coming years.

AI is “poised to become the driving force” in retail growth, with AI accounting for a whopping 78% of growth by 2029, IHL analysts estimate.

Because of AI’s enormous potential, we devoted the latest edition of the Logic Retail Leaders Forum to the topic. Logic’s Andy Winans, Chief Client Officer, Sam Caltabaino, VP for APAC, and Tony Rost, Global Managing Director for Cloud & Managed Services, hosted the event. They were joined by an impressive roster of retail tech leaders from Australia and New Zealand, including:

• Scott Coppock, CIO, Country Road Group
• Edwin Gear, CIO, The Warehouse Group
• Craige Pendleton-Browne, Interim CTO, David Jones

The forum participants had a lively and frank conversation about the current state of AI at their organizations, challenges they face in bringing AI use cases to market, and the next steps they plan to take. Here are six takeaways from the discussion.

1. You have to get the data right first

All three participants agreed that clean data and solid data governance are fundamental to any future AI initiatives. And like retailers around the world, they have been on a long, multi-year journey in order to support data-driven retail. In fact, they are well on the way to making that happen and already have the data to leverage for specific AI use cases.

However, the journey continues, and it will take time to get all relevant data from across the organization AI-ready. “It is going to take us another few years to get there,” said one participant.

2. Make AI bite-sized, at least at first

Except for a few very large organizations with deep pockets, most retailers have been relatively slow to adopt AI in a significant way. All three forum participants said that, while the C-suite is enthusiastic about the future of AI, it can be difficult to secure funding for AI initiatives with so many other competing needs. “It can be hard to gain traction for an AI project, especially when you’re competing against the option of opening four more stores,” explained one leader.

While the participants see the value of adopting a big-bang, holistic AI strategy, they said tech leaders were more likely to gain executive buy-in by starting with a limited number of well-chosen use cases that can quickly provide measurable ROI. “Being able to demonstrate the value even in a small area can really open doors,” said one executive.

3. ITSM is a great use case for building AI momentum

Several of the retail leaders consider IT service management (ITSM) as the ideal starting place for their AI journey. It is a relatively mature use case, making it easier to demonstrate potential value based on existing case studies. Working with an AI consultant, one leader explained that AI could result in autonomous resolutions in more than three-fourths of ITSM incidents handled by support staff today, and slash resolution times from days to hours or minutes.

In addition to rapid ROI, ITSM is a way to create momentum for AI because it can provide value in a highly visible way across the organization. “You have the potential of getting positive feedback from your entire user base, and that can give you a lot of energy and momentum,” he explained.

Potential ITSM use cases include:

• 24/7 chatbot support. AI-powered chatbots are improving quickly. They are a great way to quickly resolve common issues and keep down ticket numbers.
• Ticket prioritization and routing. Not only can AI often identify issues immediately, it can also route tickets to the right resource and even define the severity of an issue.
• AI-driven, human-in-the-loop resolutions. AI-driven chatbots may not help users solve more complex issues, but in the hands of specialists, they can be a powerful tool for speeding resolutions and minimizing P1s and P2s.

4. Inventory use cases promise even bigger ROI

By winning the confidence of the C-suite with a successful ITSM use case, the leaders said they would then make the case for much more impactful AI use cases, especially around inventory.

“When you have humans managing inventory using rudimentary algorithms, you have a place where you can realize big savings,” argued one participant. “I think you would gain a lot of interest if you say, ‘Now let’s use AI to deal with the fact that we are purchasing tens of millions of dollars worth of inventory more than we should.”

Impactful use cases mentioned by participants included:

• Inventory and price optimization
• Store-level size curves
• Targeted offers

5. Partner with providers, especially the big cloud providers

Because AI talent is scarce and expensive, all three participants believe that instead of struggling to build an in-house team, partnering with providers is a great way to go—particularly the big cloud providers. “Surprisingly, it is the big guys that are a lot more innovative and willing to work on joint opportunities when it comes to AI,” said one participant.

In addition to having large libraries of pre-built AI offerings, the large cloud providers offer microservices and metered billing. That makes them a cost-effective and flexible option, especially when compared to third-party providers who generally just repackage what the big providers already offer.

6. AI will speed the transformation of retail from a “push” to a “push-pull” model

In the digital age, retail has increasingly moved from a “push” model—the retailer making big merchandising bets on an assortment of products and pushing those products in front of all its customers—to a “push-pull” model, in which customer demand signals take the lead in defining what retailers offer. AI is likely to accelerate this process, both by massively increasing insight into individual customer preferences and by enabling retailers to deliver extreme personalization of offers and experiences at scale.

“We are on a journey of modernizing our systems and our infrastructure to get to the point where our business is truly demand-based,” said one leader. “We are beginning to truly have common data models for decision support and all kinds of higher-order functions.”