By Graeme McVie
Global Managing Director
Data

Feeling pressure to jumpstart gen AI initiatives? It’s time to take a step back and define what gen AI can and cannot do, how it’s different from other forms of artificial intelligence, and why retailers still need an overarching data and analytics strategy that prioritizes opportunities that drive the most value.

The wow factor of generative artificial intelligence (gen AI) tools like ChatGPT has been undeniable. “It’s impossible not to get excited,” as InfoWorld puts it. Naturally, retail boards of directors want their organizations to come up with a genAI game plan. But in a world of limited IT resources, should gen AI take precedence over more tried-and-true data and analytics approaches?

I do believe AI in general will transform retail. However, at present, real-world retail use cases specifically for gen AI, while impressive, remain limited. More importantly, gen AI doesn’t enable retailers to leapfrog the data and analytics challenges they faced before the new technology came along. If you haven’t deployed AI capabilities to tackle core merchandising, supply chain, marketing, and operations questions, you probably need to start there.

The problem is that all the hype around gen AI has also sown confusion about what gen AI is, what it can and cannot do, and how it is different from more “traditional” AI/ML (machine learning) approaches that have been around for a while—some of them for decades.

In short, tech leaders need to right-size organizational expectations around gen AI. They need to clear up confusion between gen AI and other forms of AI. And they need to help their organizations understand that gen AI will never be a substitute for a value-focused data and analytics strategy—which could include anything from filling gaps in basic KPI reporting all the way to AI-driven predictive and prescriptive analytics.

Here’s the good news. By focusing on their core data and analytics strategy first, retailers can not only deliver more tangible benefits to the business in the short term but also prepare their organizations to power truly game-changing AI use cases over the longer term.

What GenAI Can (and Can’t) Do for Retailers

GenAI isn’t magic, no matter what the bleeding-edge evangelists would have you believe. Here’s how it works. By leveraging large language models (LLMs), it can train on huge amounts of unstructured content, and then generate new content outputs as a result. That instant generation of content, often with impressive results, has created lots of excitement in the mainstream, and in the boardroom.

Wasting no time, retailers are already using gen AI to power some very interesting use cases.

• Marketing content. Gen AI can be used for marketing content ideation, including personalizing marketing communications to individual customers at scale.
• Consumer sentiment analysis. Gen AI can analyze vast quantities of customer reviews and social media content to understand opinions, wants, and needs.
• Generation of additional product attributes. Gen AI can identify and tag products with descriptive information that enhances merchandising, promotions, personalization, and more.
• Productivity enhancements through AI chatbots. Gen AI can power self-service chatbots that help speed the resolution of customer inquiries.

However, in each case, gen AI is an accelerator, not a miracle pill. And in each case, the output from gen AI tools still requires meaningful human oversight.

Unraveling Confusion Between Gen AI & Other Forms of AI

So far, gen AI use cases for retail have been exciting but limited in scope, largely deriving value from productivity enhancements or cost savings. By contrast, other forms of AI—for example, those that use structured data generated by barcode scanners—have the potential to materially improve a retailer’s top and bottom lines across core retail business processes, including:

• Merchandising decisions from regular pricing, promotions and markdown to assortment and space planning.
• Supply chain decisions from demand and replenishment forecasting to inventory allocation.
• Marketing decisions including personalized marketing and predicting customer churn.
• Store and digital operations applications including product recommendations, predicting drivers of shrink, and improving customer experience.

Creating a Holistic Data and Analytics Strategy

With so many innovations happening so quickly, it can feel a little overwhelming to chart the right course with data and analytics. To ensure you’re sailing in the right direction, it’s important to develop a strategy that can serve as a compass for the journey.

For starters, retailers must ensure that their data & analytics strategy is aligned with
the company’s overall business strategy. Then, you need to identify the major business decisions that deliver value. With that, you can determine the contextualized insights to deliver insights you want to deliver to decision-makers, including enterprise-wide capabilities that meet cross-departmental needs.

Once you answer these high-level questions, you can determine the data and analytical capabilities that will deliver the insights you need. Finally, you can apply a framework that allows you to create a prioritized roadmap for delivering those contextualized insights that takes into consideration the resources & talent you have already, or can acquire, to ensure that you consistently deliver value over the short and long term.

Yes, gen AI can produce immediate and impressive results, and its potential use cases will only continue to grow as the technology matures. However, retailers and their boards need to understand its limits and keep their focus on the kinds of data and analytics strategies and capabilities that can truly impact their top and bottom lines.

In his capacity as Global Managing Director of Logic’s Data practice, Graeme helps our clients leverage their data and analytics to make decisions that ensure competitive success. Graeme is an established thought leader with deep domain knowledge in the retail and consumer brands sectors. He’s an expert in advanced analytics in the areas of merchandise planning and optimization, shopper centricity, retailer/supplier collaboration, and customer segmentation. In his 25-plus years of consulting experience, he has worked with top-tier clients such as Target, CVS, Best Buy, IKEA, Macy’s, Kraft, and Coca-Cola.