The world of unlimited choice is the consumer goods market, where every product on the shelf is a testament to risk calculation. Market research 2024 hit a global valuation of more than $90 billion, IBISWorld reports, as brands try to crack the code of customer desire and discover winning delivery routes. For retailers and consumer packaged goods (CPG) businesses, the optimal assortment choice, be it the most profitable toothpaste or snack, can dramatically affect millions of revenue. McKinsey estimates that a clever product mix can drive sales up to 10% higher; it is, however, the effective deconstruction of data, including sales trends, shopper behavior, and stock levels, that makes it possible. Get it wrong, and there are unsold products or disappointed shoppers. Firms like Kantar Consulting operate in this multifaceted landscape and offer a mix of analytics and strategic advice to industry leaders like Procter & Gamble and General Mills. But with data generation projected to explode into 181 zettabytes by 2025, Statista reports, the idea of wading through such enormous amounts of information by hand is like the exercise of attempting to hold water in a sieve.
Vibhu, who served as a Data Analyst at Kantar Consulting in Norwalk, Connecticut, from February 2021 to January 2022, contributed more than mere technical expertise to his role. Possessing a Master’s degree in Business Analytics from George Washington University and a passion for discovering solutions, he approached his responsibilities with a combination of determination and inquisitiveness. His primary objective was to elucidate data so that clients could take actionable steps rather than merely observing it. “It’s not only about crunching numbers,” he says. “It’s about handing someone a tool to succeed.” His most significant achievement? The development of a prototype for an automated product segmentation tool that fundamentally transformed operations for Kantar and its clients. In addition to that, another tool based on NLP and embeddings would help segment tougher private label items into the segments with human-like accuracy
Then Kantar’s team was using manual sorting, a labor-intensive and error-prone method that was frequently outsourced to engineers. The process was frustrating for engineers and slowed deliverables for clients who were keen to connect products to profits. Vibhu saw the bottleneck and took matters into his own hands. Using Python, he created a working prototype that used rules defined by the user for segmentation, and if no rules were available,” word/character embeddings and AutoML to sort products automatically into the available categories. “I couldn’t stop thinking about the hours we were wasting,” he says. “Why not let technology do what it does best?” His tool pulled information out of product labels, ingredients, names, and other applicable data, and sorted them systematically with astonishing accuracy. Kantar’s tech team saw what he was doing and built a full system, deploying it on their internal server to cater to client and team needs.
The payoff was real, and it hit fast. For clients like Keurig Dr Pepper, Vibhu’s tool cuts segmentation time by up to 40%, saving weeks in the process. It wasn’t just faster, it was more precise, cutting down on errors and delivering cleaner financial outcomes. “It was like giving them a new pair of eyes,” he says. “They could trust what they were seeing.” Kantar’s Perfect Category platform, which handles $600 billion in assortment planning every year for over 50 top brands, was the beneficiary. Clients using existing solutions revenue grows 5-10% per category, but with new precision built into the process using auto segmentation and faster processing, there is a deeper sense of trust from the clients, a direct result of the smarter strategies Vibhu’s work enabled.
This victory was more than a project; it was a change in the way Kantar worked. In this company, speed and accuracy are paramount, and Vibhu’s tool effortlessly converted a weakness into a strength. His supervisor recognized him as one of the most innovative recruits of the year, a silent nod that mattered a great deal. “You don’t see the whole picture at the start,” he says. “Then clients keep coming back, and it comes home.” The renewed contracts meant a great deal, making Kantar a go-to partner for 42% of the world’s leading consumer packaged goods companies.
Vibhu’s experience at Kantar was a dynamic and active one. He started with building blocks, creating ETL pipelines in SQL to pull data from FTP servers, making it all come together. But he did not stay stuck in this position. Soon, he was creating demand curves to change product mixes and creating Power BI dashboards that gave clients more visual learnings about their assortments and new information about their product market fits. This transition from a technical to a strategic thought process set him apart from his professional peers, who stayed stuck in data analysis. He was already working with top leaders and clients, basically translating numbers into reality-based plans.
What set him apart was his combination of skills, Python and SQL coding, of course, but also a vision for the larger picture. His early foray into NLP for private-label segmentation in 2020 was a risk, ahead of the clunky old methods. “I wanted to push what we could do,” he says. “Not just get by, but deliver.” That attitude paid off, Kantar’s workflows streamlined, and clients left in a better frame of mind.
Kantar itself is a powerhouse, steering assortment for brands across six continents. It’s Perfect Category solution mixes predictive analytics with virtual reality, letting clients test shelf ideas without touching a display. Vibhu’s tool fit right in, amplifying Kantar’s reputation for sharp, forward-thinking work. Word of his approach spread, industry talks started pointing to it as a model for blending automation into consulting, nudging others to rethink their methods. Now at Capital One, Vibhu’s working on new challenges, including reducing employee turnover. But his Kantar chapter remains relevant: 40% faster segmentation, tighter accuracy, and clients who stuck with him. His work not only took Kantar up a notch, it signaled where the field’s headed, showing data’s power when used with caution and creativity.
