How retailers can benefit from big data and artificial intelligence
The role of AI in retail
Although artificial intelligence (AI) has a reputation for being one of the 21st century’s most cutting-edge technologies, its history dates back centuries. Our fascination with creating machines that can think as we do began with the ancient philosophers.
Today, AI permeates every aspect of our lives, driven by the expansion in processing power and the resulting explosion of data managed by every organization and individual. Voice assistants, chatbots, personalized retail recommendations, and predictive text are all examples of AI playing a role in some of the most relied upon day-to-day processes.
As for the retail sector, there are many reasons it’s primed for AI. The most important of these reasons is retailers’ ready access to data, not just from transactions but — in an era of digital commerce — also from browsing habits, factors affecting purchase decisions and how offers drive behavior.
One purpose of AI is to identify patterns in data that would be near impossible for humans to see. The extraordinary power of AI systems to sift through different data sources to find relationships between them leads to equally impressive insights that enable retailers to make more informed decisions, reduce costs and waste, optimize resources, and improve customer experiences and loyalty.
There is a virtuous cycle of AI that exists in retail, and it starts with automating existing manual processes. It then expands understanding of customer behaviors, incorporates this knowledge into processes and, finally, delivers results. This cycle continues as new insights are built into future strategies, and as they build repeatable models that provide even more results as they test and learn.
Managing the ‘what-ifs’ in retail with AI
AI is also great for calculating the impact of ‘what-if’ scenarios. Based on the information a retailer already has about the sale of goods in a particular store, for example, it can confidently predict what would happen if the store reduced prices or introduced special offers at a certain time of day.
Similarly, AI can help stores analyze which products sell better in-store vs. online within certain scenarios, including weather conditions, social media sentiment, or based on customer personas. The ideal from here is, of course, personalization, which grocery retailers have become particularly skilled at during the lockdown.
Not only can online shoppers view their regular shopping list and set up weekly deliveries based on their favorites, but the retailer can nudge the customer to buy previously purchased items that are not on their current list.
Unsurprisingly, big data and AI rank highly among retailers’ investment plans around the world. In late 2021, research undertaken with 312 global retail leaders revealed that big data and AI will be a top priority. Nearly one in four (39%) respondents said that big data and AI will be their second most important area for investment in automation, just behind assortment management and optimization (41%).
Of course, assortment management and optimization is a process that could be improved through the application of big data and AI itself — along with myriad other processes such as in-store and on-shelf availability, demand forecasting, freshness management and optimization, monitoring store activity with IoT, product selection, development and design, and workforce scheduling.
Retail’s best first steps in AI
As the list above demonstrates, there are various opportunities for AI to improve processes within the retail sector. It’s unlikely that any retailer has access to the funds needed to completely change how they operate overnight. On the other hand, limiting automation to hyperlocal applications and processes may not deliver the best results in the long term.
Whichever route retailers follow — and it’s likely to be a middle ground between those two extremes — a vision of the role retailers expect AI to play in their business and the value they expect it to generate is vital.
As Deloitte suggests, it may be preferable to build low-risk use cases that deliver clear and immediate value, such as improved identification of repeat stock-outs of a core product. The idea is to start small but scale fast.
Deloitte continues, “Once use cases have been optimized and are routinely delivering value, they can be swiftly rolled out across the organization to maximize the value and spread the benefits. In the example described above, this could mean the indicators that best predict recurring stock-outs [of] a core product may also apply to a number of similar products.”
AI has come a long way since the days of the ancient mystics. The opportunities that today’s AI opens up for retailers are wide and varied, and they’re certain to bring huge benefits to the brands that can see — and seize — its potential.