The Center for Data Innovation spoke with Navjit Bhasin, chief executive officer and founder of Newmine, a retail technology company based in Boston, Massachusetts. Bhasin discussed how Newmine helps retail organizations effectively manage and reduce product returns using their innovative data tools.
This interview has been edited.
Sujai Shivakumar: How big a logistical and financial challenge are returns for retailers?
Navjit Bhasin: Returns are the dirty little secret of retail industry. The word “returns” has had a negative connotation. Everybody swept it under the carpet and only managed the cost and logistics of returns. That’s where retailers are leaving money on the table.
Retail is a for-profit business. They make money by selling merchandise to consumers. Returns negatively impact profits. Think of a billion-dollar retailer. If their returns rate is 10 percent, that’s a hundred million dollars in returns, right? Retailers spend on marketing and advertising to generate that billion-dollar demand, which ultimately falls down to 900 million because of returns. So, right off the top, they’re writing off that hundred million dollars.
When a product is returned, there are shipping and packaging costs associated with returning that product back to the retailer. Then there are the labor costs of refurbishing it to put it back up for sale. In fashion industry where the sales life is time sensitive, the returned products may not be relevant any longer to the consumer. Retailers then have to decide if that product should be sent out to an outlet for sale at a markdown, or if it is eventually going to end up in a landfill. So, there is that reverse journey.
Returns is a complex problem to solve. It not only involves taking into account the different functional areas of the retailer combined with data and insights, but it also considers the alternate corrective actions that a retailer can take to reduce and streamline returns. In the past, we tried doing this with brute force, but it was not a scalable and sustainable process. Now, with AI we have the right technology and tools to do it at scale and automate much of the process of discovering the root cause of returns and fielding the right insights to the people who can fix it. For retailers, from a financial perspective, it’s an absolute dream logistically.
With the onset of the COVID pandemic, retailers got shut down for the initial three months. Then, with the pent-up demand, e-commerce grew through the roof. But returns also grew at the same rate. Even if the returns rate still stayed at 20 percent, that is 20 percent of a billion versus 20 percent of one and a half billion before the volume grew. Take the same equation back into the reverse logistics chain, all of those parties, all of those constituents, and you begin to see the need to manage the process a lot more.
Shivakumar: How does this challenge vary by type of retailer, type of product?
Bhasin: I can give you two extreme examples. Let’s first look at a beauty products company whose price point is anywhere from $15 to $50—that’s one extreme. And the other extreme, think of a company like a Wayfair selling a bulky home furnishings product, which can be anywhere from a few hundred dollars to thousands of dollars.
The whole process—from the consumer’s behavior to the retailer’s reactivity to those returns—is going to be very different for a $15 lipstick compared to a $4,000 couch that the consumer did not like for some reason. Now even to get that couch out of that consumer’s house, back into the retailer’s distribution center can cost a couple hundred dollars. In some of these categories, the margins aren’t that high to begin with, and that’s why retailers struggle.
We look at the impact of returns on the whole ecosystem, including suppliers and the vendors. That’s where our AI technology comes into the picture. From the moment the consumer has made up their mind to return a product, our technology starts to track any data point we can gather—whether it is coming from the customer’s interaction with the retailer, whether it is to a call center, product reviews, social media data.
The consumer cares about getting his or her refund back…or they want a replacement product ASAP. Logistics companies like FedEx and UPS are happy when there are more products being shipped in either direction. But these returns are not profitable for the retailer or environmentally sustainable. The retailer wants the customer to have a positive experience, so if a return does not go well, they fear that the emotional connection with the consumer—and with it the lifetime value of that relationship—is being lost as well.
Shivakumar: How are retailers using AI to predict consumer behavior on returns, and why is this important?
Bhasin: Retailers are missing out on leveraging AI to address one of the most expensive problems in retail today—merchandise returns. We focused our efforts on applying AI to a clear and present danger to the retailers’ profitability. It does not have to alienate the consumer. In fact, we leverage our AI tools and predictive models to understand the voice of the consumer. What is this customer is starting to talk about the product? Is she complaining that the quality is not good? Did she have different expectations based on the product description? We build models to understand the customer’s emotions. We can alert retailers to take action on products that we believe are going to have much higher return rates. The power of AI to translate the consumer emotions to a fact-based corrective action by the retailer is where the rubber meets the road.
We leverage AI-based scoring to generate what we call “KeepScore.” This score is an index of AI-based scores that benchmark product, supplier, and customer success by combining data on sales, profit, and return analytics to create a standardized measurement to predict if a retailer’s customers will keep their purchase.
KeepScore allows retailers to evaluate, at a glance, which products are meeting customer expectations and predict which products will be kept. It evaluates the customer lifetime value to predict the likelihood of the customer keeping purchases. And it includes returns in supplier performance evaluations, allowing retailers to remove products with a consistently low keep score from their assortment so that they can focus on higher-value products.
KeepScore also helps the retailer personalize marketing for customers based on their score, including discounts and special offers to high scorers. Customers are happier and so are retailers.
Fraudulent returns happen. Somewhere between 5 percent to 10 percent of returns are definitely fraudulent. So KeepScore can detect those, too, and alert the retailer.
Shivakumar: How are retailers using AI to predict and manage the type of inventory that generates higher rates of returns?
Bhasin: From an inventory standpoint, retailers can leverage two key metrics: Product KeepScore and Supplier KeepScore. Retailers are already thinking, “hey, now I know the KeepScore for this product, what should my inventory levels be when I’m starting to plan my next season?
So, retailers are starting to gauge the performance of their suppliers as well and provide constructive feedback. They can go back to their suppliers and work with them to say, “these are the products you manufacture that are selling versus these are being returned.” Getting these conversations started early is how long-term supply chain impact happens.
Getting feedback from customers early in the selling cycle is critical as it helps retailers get back to their suppliers to correct any issues that customers may be having. If I, as the retailer, get negative feedback from the first two or three customers, I can fix this before hundreds of the same product are shipped out to consumers. That’s the quick action. That’s where our AI technology comes into the play.
So, for me, it boils down into bits and bytes. If there is any piece of information, whether it is hard, transactional data or soft contextual data—even if you had called them or even given a review—I want to decipher that conversation and use it. I can make the retailer more productive with this knowledge, rather than just have them ignoring early signs of a problem upstream—something that could take a long time. And by the way, a lot of these opportunities for action go unnoticed because retailers don’t have the resources in terms of people, capital, and technologies to detect these problems.
Shivakumar: How are retailers preparing for changes in returns behavior as the pandemic ends?
Bhasin: When we started our journey as a startup, our first goal was to educate the market that returns reduction is possible. So, we spent the first two years doing it. Then COVID happened. And in fact, COVID accelerated our mission of educating the retailers on the power of insights from merchandise returns. And today, we are sitting right at a tipping point for insights-driven retail.
Retailers for the most part have always been in the backseat, in terms of technology. Our goal at Newmine is to get retailers on the front end of technology to drive profitability. Other industries spend far more on data analytics compared to the retail industry. Financial industry spends about 20 percent of their revenue on their technology budget, whereas most retailers spend at an average of 2.5 percent. They’re always squeezing the technology budget. Amazon is an outlier because Amazon is a diversified business.
Companies like Amazon and Walmart have got their shipping and reverse logistics costs under control today. But look at the other thousands of other retailers who are seeking to remain competitive as this industry goes through major transformation right now. Without the right tools, without the right mindset and, without the right buy-in at the leadership level positive change is hard to make happen. And you know what, I’m glad we’re talking to you today because it requires a major shift in the retail ecosystem.