Find the original article published by Chain Store Age here.
If given the option, who wouldn’t want a faster car?
Let’s say a group of auto-aftermarket producers developed a set of spark plugs that could be added to an engine to automatically boost the vehicle’s torque. Instantly, an old jalopy could race off the line. Those components would be revolutionary, but would the car still run the same? Might the spark plugs cause new problems in the engine?
Alternatively, what if a car owner simply bought a newer automobile with a faster engine with next-gen spark plugs already built in?
Retailers can look at the evolution of AI in much the same way. Want to boost the performance of supply chain, pricing and assortment optimization? The first phase approach has been to bolt on predictive analytics modeling and a generative AI co-pilot to a legacy platform.
The second phase is implementing a native AI-powered platform with generative AI applications included, and unified data and analytics as the engine.
Retailers are investing more in AI and generative AI. Coresight Research estimates that the combined generative AI hardware and applications market will total just shy of $80 billion this year and is growing at a compound annual growth rate (CAGR) of 31.1% (reaching $235.5 billion in 2028).
As spending on AI continues to rise, here’s a look at the second phase of AI implementations and what it means for retailers.
Unified data is the key to the second phase of AI
As generative AI technologies, specifically, became more mainstream over the last few years, and AI technology became more affordable, retailers rightfully jumped head-first into trying out applications. However, the data that feeds GenAI applications has not been centralized, and usage has been untethered.
For instance, a marketing department might test a generative AI tool for automating advertising text for promotions, while a supply chain team tries out a different solution for inventory safety stock adjustments. What’s more, one tool could be in the cloud and the other working off a desktop, each using different data. Retailers have often overlooked the importance of implementing unified data and setting clear governance rules around the use of the data.
Sticking with the analogy: a car running on a legacy engine with disparate add-on components attempting to make it go faster could end up on the side of the road. A legacy platform with disparate components layered on can create data silos that stall insights. These are among the reasons that nearly a third of generative AI projects will be abandoned due to poor data quality, according to Gartner.
Retailers moving more into a serious adoption phase of AI capabilities first need to centralize the use of AI, develop rules on how users can access and leverage the insights, and ensure all the data goes to one single source of truth.
Product information intelligence anchors the second phase
The second phase of AI, characterized by native AI platforms, begins with product information intelligence as the anchor, with unified components like generative AI integrated from the start. Product and shopper data are then harmonized across merchandising, marketing, pricing, assortments, and other applications.
Benefits of a second-generation, native AI approach include the ability to:
Streamline data management across retail business users into one single source of truth, which eliminates data silos and latency.
Enhance collaboration between retailers and CPGs with real-time results on product inventory issues, demand forecasting, and promotional and pricing strategies in one unified place.
Remove app fatigue on a tech architecture, where IT teams overload a legacy platform with disparate apps.
An Accenture study finds more than 90% of the retailer executives surveyed expect to scale up investment in AI within the next five years, and three-quarters of respondents see generative AI as being instrumental to their revenue growth.
To manage expectations and meet this demand, retailers will benefit long-term from a more native approach with AI, working directly from product information intelligence.
Defining the second phase of AI for retailers
The second phase of AI encapsulates retailers exploring a native AI platform that contains all the product data that retail applications need. Layering AI on top of legacy solutions was a quick, easy fix in the first wave of generative AI implementations, but this approach falls short of realizing the true benefits of AI.
As applications and teams leverage AI to grow margins and enrich shopper experiences, the intelligence and results of those optimizations must be stored in this same centralized platform to inform and reinforce future decisions.
Unified, native platforms help improve the accuracy and responsiveness of AI applications, shorten the time to develop new AI applications, and address the problems of compliance, data security, and managing IT budgets. Retailers are increasingly realizing the benefits of this approach; after all, who doesn’t want a faster, more efficient car?