You will be redirected back to your article in seconds
Skip to main content

Q&A: Microsoft’s Retail Strategy Director on AI’s Role in Preventing Store Closures

Artificial intelligence is a top-of-mind topic today as retailers double down on technology that can help them ward off threats from all sides.

Despite intense interest in AI and what it has to offer, some retailers haven’t been able to take advantage of the technology while one-fifth have no deployment plans on the horizon, according to a survey conducted by Synchrony, the consumer financial services company. There’s plenty of skepticism about the hype surrounding AI and a few retailers (13 percent of the 300 retail decision-makers Synchrony surveyed) have little faith that solutions currently on the market can deliver on their promises.

But in a retail landscape in which an overstored U.S. is seeing thousands of shops fold each year and tariffs could threaten thousands more, there’s no time to take a wait-and-see approach to AI and its potential to help retailers survive tumultuous times.

Sourcing Journal caught up with ShiSh Shridhar, worldwide director of retail business strategy for Microsoft, to learn how the software giant’s solutions are helping retail enterprises make sense of AI and apply it across the business to see quick wins while maximizing ROI.

Sourcing Journal: Why are retailers finally starting to get serious about artificial intelligence? What macro and micro factors are driving interest?

ShiSh Shridhar: There’s no denying that the retail industry has been changing over the past several years and that customers are driving this change. Retailers are realizing that the most successful companies are the ones adopting the latest technology, such as AI and IoT, to plan for their future. Retailers that succeed in today’s landscape will be the ones who put customers at the center of every business decision and work backwards from there, leveraging data and AI to enable intelligent retail across their organization. And I believe there are two factors that are accelerating AI across the retail industry: the ubiquity of data and the availability of solutions that empower retailers to easily make sense of that data.

Retailers have always had data from daily transactions. The challenge is that this data is one-dimensional—it only gives them insight into what transpired in the past, not giving them a 360-degree view of customer preferences and purchases across other parts of their business.

Related Stories

Today, with the advent of open data initiatives, social media and [Internet of Things], we have an explosion of data sources outside of a retailer’s own proprietary data. This data can help retailers to be more competitive, better understand customers, deliver a more personal and relevant experience, augment the capabilities of their employees and most importantly drive operational efficiencies in their business to deliver the right product at the right time, place and price.

It’s this explosion of data that’s driving retailers’ interest in AI. The combination of the retailer’s proprietary data with public and purchased data enables them to further enrich their AI models, making them far more accurate and as a result, driving much higher ROI.

However, all that data doesn’t mean much if retailers aren’t able to analyze it. Microsoft has been focused on democratizing AI for companies by delivering tools that do not require deep technical skills or programming. Data scientists can now easily build AI models with very little or no code.

SJ: AI can be applied in myriad ways across the retail organization. In what parts of the business do retailers typically want to start this journey?

SS: In my experience, the primary area that retailers have been focused on is customer centricity, or better understanding their customer. Today, more than ever, customers are in the driver’s seat, meaning their brand loyalty and expectations have changed. It’s extremely easy for customers to shop across different channels, sites, platforms, etc. With all the capabilities of AI, retailers have the potential to use the data across every area of their business to better understand the wants and needs of their customers. Combining the data from all their touch points, public and purchased, enables retailers to deliver a more personalized and relevant experience.

As the lines continue to blur between customer centricity and the idea of an intelligent supply chain, this data also informs retailers of what kind of inventory they should have based on the customer profile around the store. They can define the assortment, inventory and price because they now know more about the customer. Retailers can better engage with customers and acquire more customers because they know them better. As a result, when retailers engage in a more personal manner, the customer acquisition, retention and loyalty improve.

SJ: Where does AI make the biggest impact in retail? And how long, on average, until retailers achieve ROI from their AI implementations?

SS: One of the biggest impacts AI can have in retail is reducing operational costs. Retailers have been applying AI models to help better determine the right product and assortment at a hyperlocal level based not only on past sales data but also other influencers like local demographics, economy, events and holidays. This enables retailers to improve the profitability of their stores. One of the challenges many retailers are dealing with today is store closures. This has challenged them to be more thoughtful about placing or prioritizing stores in the most profitable areas. This is where data and AI can come in to help retailers determine location profitability.

With AI, companies like Microsoft can help retailers fine-tune the data to identify the most profitable location for all stores. By applying analytics, retailers can lift the profitability of every individual store and help reduce store closures. Optimizing the inventory, assortment and pricing in stores can drive immediate profitability for retailers by reducing out of stocks, overstocks and improving inventory churn.

SJ: What are the most common challenges to deploying AI and retailers’ most common mistakes or missteps around AI?

SS: A common challenge, and a common mistake, for deploying AI is starting with the data instead of the business challenge to solve. Data is only as good as the retailers’ ability to make sense of it, and many times so much data exists that they don’t know where to start. To combat this, it’s important to instead start with the business challenge and then ask what data they have that will help solve it.

However, embracing AI takes skills and resources. AI can be complex, and the skill sets needed to run it successfully are expensive and hard to find. Many retailers don’t have the expertise in house to manage complex analytical tasks, and as a result, data goes largely untouched save for maybe a few limited scenarios.

Today, building AI models and deploying them has been democratized and simplified—you no longer have to be a coder or have high technical expertise. So now, it’s much easier to uncover insights. By taking past transactions data and combining that with data from other sources both internal and external, you can train the AI model to identify patterns.

For example, when the weather changes, how does it impact your sales? When the demographics change in an area, do your sales increase or decrease? The more data you have about the influences around you, the more accurate your AI model is going to be. Data visualization tools enable business decision makers to access the insights from AI models without having data science skills.

SJ: What prevents some retailers from embracing AI?

SS: In addition to the challenges I already mentioned around the skills gap, the other thing that is challenging retailers is data dumping. There’s a lot of data, but not all of it is “clean” and retailers can’t use it. Microsoft has helped address data management by applying democratized tools that are easy to use and eliminating the need for programming and technical expertise required to prepare and manage the data.

SJ: Do you believe AI is table stakes yet in retail? If not, how far out into the future do you see that happening?

SS: I believe that AI is becoming ubiquitous in retail, but perhaps not in ways that are immediately evident. Today, AI is part of every tool and solution that Microsoft offers. We have AI solutions that address knowing your customers, empowering your employees, delivering intelligent supply chain solutions and more.

There’s also a lot more that can be done with AI. AI can help—and is already helping in some cases—retailers create new business models like autonomous stores, scan-and-go features, visual search and more. So, the industry has made a lot of progress, but at the same time they’ve only scratched the surface. Microsoft believes that AI will be the tool of the trade in this new retail frontier.

It’s AI—including cognitive services, machine learning, bots, etc.—that will allow retailers to drive game-changing efficiencies and extract actionable insights.