Personalizing the Retail Experience: 6 Questions With Birdzi CEO Shekar Raman

It began with an idea his 11-year-old daughter had about locating items in a supermarket. But what started out in 2014 as a prototype, like many businesses, morphed into something else.

After being inspired while helping his daughter with her Invention Day project for school, CEO and Co-founder Shekar Raman formed Birdzi , a grocery retail solutions company that taps artificial intelligence (AI) for data-driven results. The Bridgewater, New Jersey-based company began when he and his co-workers spent weekends meeting at Panera Bread ideating and building a prototype for a store mapping and indoor navigation platform. But in a few years, they realized they were solving a consumer problem with no business backing. Enter the pivot.

CLICK HERE TO SIGN UP FOR THE FREE TAPINTO.NET NEWSLETTER

“We started building out an entire platform that would allow retailers to gain insights into shopper behavior, trends and preferences,” says Raman. “That's kind of how we got into the personalization side of things. And then we realized that you could do a lot of things in terms of giving people relevant offers and promotions at the right time to drive their purchase behavior. And what we of course learned in that whole exercise is, doing that makes people want to come back to your store more because they get the feeling that ‘this retailer really understands us.’”

Today, Birdzi is a customer intelligence and engagement platform for grocers that leverages AI to provide retail and consumer goods companies with offers personalized specifically to each individual customer.

To learn more, COMMERCE Magazine sat down with Raman to talk about Birdzi’s journey and newest offerings.

Jamie Grill-Goodman: Can you give me an example of something the Birdzi platform does?

DOWNLOAD THE FREE TAPINTO APP FOR MORE LOCAL NEWS. AVAILABLE IN THE APPLE STORE AND THE GOOGLE PLAY STORE .

Shekar Raman: One of our clients is Weis Markets, and they have several stores in New Jersey. So, for instance, we would get data from Weis markets — we license the first party data, data at the point of sale — and we're mining that data to understand a shopper's behavior and preferences. What is her next trip going to look like and how can I incentivize her to go back to the store and spend some more money?

[The shopper] would typically get an email with six or eight offers based on their specific history. So, let's say they buy Tide once in 30 days, and now it's been 40 days and they haven't come back to the store. We would send a coupon for Tide and some other products [the shopper] would like in the store. And then we try to grow spending in departments or categories that we think they’re not spending.

Let's say [the shopper] doesn’t buy paper from [that particular store]. We know they buy paper somewhere. The question is, can we get you as a buyer for paper goods in our store? This email essentially contains offers that drive purchase behavior based on [1] existing purchase history, and then tries to expand it by trying to get [the shopper] into new categories that they’re not shopping much in.

Grill-Goodman: Tell us how you're incorporating AI into Birdzi’s solutions?

Raman: There are many types of AI. What we use in our systems is a lot of what I would call data-centric models, as opposed to generative AI, which revolves around natural language models, which are really processing data and trying to come up with predictions and forecasting, and looking at your purchase history and understanding, ‘Hey, what kind of products are you likely to buy in your next trip to the store?’

What we use AI for is really for trying to understand and predict shopper behavior and the likelihood that if you buy bagels, what's the likelihood that you are also going to buy chips and salsa? Essentially, we're training our models to try to predict the relevance of a certain product to a shopper.

The second thing we're using AI for is to forecast sales for the retailer, where they're asking ‘I want to put this on the front page of my weekly ad. How do you think it's going to perform?’ We’re able to use these models to do some prediction of how their weekly ad they're going to put out four weeks from now is likely to perform, so they can make any changes based on forecast.

The third place where we're using AI, and growing fast, is to ask questions of your data in plain English. People are typically used to building dashboards and reports. We're also building out systems which allows users to just ask questions in plain English and get answers back.

Grill-Goodman: Can you tell us about VISPER Live, the new solution you just launched?

Raman: VISPER Live is basically a platform that generates unique offers for each shopper that are customized to their specific shopping behavior. Consumer packaged goods (CPGs) companies sell most of their products in the center store, not the perimeter of the store. And what's happened historically is they have decided what products to promote, and the retailers have taken those set of promotions and then tried to target it to shoppers that those promotions match up to. But what we've done with VISPER is, we flipped that equation and said, ‘let's start with the shopper to understand what the shopper wants. Forget about what the CPGs are willing to sell you.’

VISPER is a mechanism to create these bespoke promotions for each shopper that are tailored to grow their behavior and loyalty. But it does it at scale where, if you give it a million shoppers, it can potentially generate six million unique offers. So every shopper is getting unique offers based on their specific purchase behavior and history.

Grill-Goodman: The company also has an offering called ECHO on the horizon that will be able to connect Birdzi’s network of retailers and their customer data to translate customer behavior to CPG brands. Can you tell me more about that?

Raman: VISPER is purely for the retailer to grow shopper behavior. ECHO is a platform where I could essentially create an audience based on some rules. I could set up a journey which says, ‘anybody who buys my Lemonade for the first time, send them a $2 off coupon so they can buy another bottle within the next two weeks. But it's targeted to specific people based on specific behavior. It's not just blasting everybody. The difference with the Echo platform is you can think of it this morning and execute it before noon.

Grill-Goodman: There's historically been issues with retailers not wanting to share consumer data with CPG companies. How is ECHO helping to bridge that gap?

Raman: We're essentially becoming the intermediary, we're kind of like that in between layer. It masks the retailer's data directly, because all the CPGs really interested in is understanding behavior and targeting specific types of behavior. We're essentially becoming that data security layer between the retailer's first party data and the CPG.

Grill-Goodman: Before we part, what are your thoughts on running a technology company in New Jersey?

Raman: I think the ecosystem around here is pretty strong and exciting. I think the talent pool is great. You've got New York City, lots of people, lots of great companies based out of the city. I think there is a growing tech hub in the area. In our early days we certainly got a lot of access to resources through programs that were run by the NJ Economic Development Authority (NJEDA). I think the government is stepping up, trying to provide a lot of incentives for people to invest in New Jersey companies . It has worked out good for us.

For more local news, visit TAPinto.net