Uncovering the Counterintuitive with L’Oreal’s Data Chief: How Data Challenges Our Assumptions
Ahead of his speaking engagement at CDAO Melbourne, L’Oreal ANZ’s Chief Data Analytics Officer, Varun Verma, speaks to Corinium’s Vanessa Jalleh about uncovering anomalies through data and how these discoveries can change an entire business approach.
Misconceptions and counterintuitive insights are constantly uncovered as we seek to unveil more truths from the vast amounts of information readily available.
For data leaders like Varun Verma, the Chief Data Analytics Officer at L’Oreal, uncovering anomalies in data and through data is part of the job.
Anomalies
While there are several examples Verma could bring up from his extensive career, there is one data anomaly example that stuck with him from his days at Visa. Noting that Visa does over 60% of card transactions globally, Verma highlighted that the amount of data he and his team were dealing with was certainly vast and could uncover significant findings.
“The hypothesis that banks and financial institutions have is that people make their travel bookings and hotel reservations primarily on their days off, which are Saturdays, Sundays, or any of the public holidays,” Verma says.
“Banks usually market their products to consumers on these specific days in order to attract consumers’ attention and help them book the right price points and holiday options they are actively seeking.”
However, after digesting and processing the data available, they found something quite different.
“When we did this analysis with Visa around the optimal times for financial institutions to trigger their marketing process, we found that days off were not necessarily the most popular times people booked their tickets,” Verma says.
“Counterintuitively, we found that there is a specific time on a weekday where a great number of tickets and reservations were being made, and that time was Wednesday between 2:00 to 4:00 PM. It was not on a weekend or holiday.”
Verma highlighted that even leisure and travel-related websites got more traction during these hours versus the times we believe people to be off work.
Making Sense of It
And what was the reason for the anomaly?
“What we found with the limited analysis we did, was consumers were inspired by midweek motivation, especially corporate employees,” Verma says.
“What they want to do is disconnect from work and engage with their personal life, which drove the midweek motivation.”
Verma noted fare fluctuation is the second factor that was relevant in this context. This is because many websites employ AI systems which trigger higher pricing points on Saturdays and Sundays, working on the assumption that consumers do the bulk of their bookings then, which in turn is believed to maximise revenue earnings.
“There is definitely a substantial amount of fare fluctuation over weekends versus weekdays, and if websites do not detect the bookings coming in on Wednesdays, they might maintain lower prices on Wednesdays to attract more customers,” he says.
The third factor is time.
Verma noted that a lot of these decisions required further consideration and were not to be made quickly.
“Whatever they are researching over the weekend, they want a few days to consider their choice and make a booking in a more rational way,” he says.
“So, Wednesday was right in between.”
A surprising insight
As Verma continued to discuss counterintuitive findings from data, he brought up a case study from Heineken International regarding the deployment of branded fridges for display and product storage at retail stores.
“Branded fridges are an important marketing strategy for beverages companies. Drinks are generally meant to be enjoyed cold, hence they are mostly stored in fridges. One way to make your drinks appeal differently against competitors in that situation is to use fridge stickers. Another way is through a branded fridge itself,” Verma says.
While branded fridges are not unusual, they cost more than fridge stickers and your brand must be able to make the most out of this marketing investment.
“There was a famous belief amongst beverage companies, that having branded fridges will yield more revenue and eventually higher ROIs, as it helps with brand recognition and coupled with the belief that consumers instantly buy what they see,” Verma says.
With such a hypothesis in mind, the simple strategy is to maximise the deployment of branded fridges at all retailers. Before Heineken decided to rollout this comprehensive strategy and execution, they decided to do a test and learn. Test and learn is an important process in analytics where a company test a hypothesis with small scale execution and later evaluate the success of the program.
“After installing thousands of fridges at various retailers during the test phase in Vietnam, we found that half of these shops were not profitable, and the fridges did not achieve incremental sales. This seemed to disprove the hypothesis that having more fridges will allow us to generate more revenue,” Verma says.
To get to the bottom of this, Verma and his team did an analysis to find out why they were getting this result.
“We further assessed the data we had about the retail shops with fridges that did not achieve incremental sales against the ones that were successful. What the data told us was it was not only the fridge, but other elements that were influencing the success of fridge installation. These elements were placement of the fridge, fridge design and clear usage of the fridge with product visibility that made a difference,” Verma says.
“With the placement analysis we found out that placing the fridge right next to the counter, where the shopkeeper is making the sale or where customers enter the shop was the most accessible, most visible point. It also created more direct brand recognition compared to fridges positioned at the back-end of the shop.”
Similarly, the usage of the fridge is very critical with the proper product display.
“Some retailers with mixed usage fridges placing the desired brand product alongside other branded products or storing other category goods may deteriorate the impact that this important marketing assets can create,” Verma says.
Contemporary Case
A more contemporary case is Verma’s current organisation, L’Oreal, which specialises in creating beauty products and services.
While there is no age limit to beauty and L’Oreal’s products are meant for all, there is primary focus on younger demographics, which are Gen Z, Gen Alpha, and Millennials. As with these consumer groups, beauty trends are very popular.
“While Loreal is capturing data from e-commerce websites, we could identify which types of consumers were engaging with L’Oreal, and what products they were buying. While most of our sales do come from that younger group, we also realised there was a significant opportunity from the age group between 50 to 65,” Verma says.
“What we found is that there is a significant proportion of the consumer group that falls under the 50 to 65 age category and their beauty needs are quite different from the younger groups. This consumer segment is growing fast and they have better ability to spend.”
Verma noted that there are a number of factors involved.
“I think the perception of ageing was quite different in the past, and it is possible that people believed there were limited products that could help them after the age of 50. But these days, the ageing perception has been redefined. It is more about embracing the right products which can give them the beauty that they desire,” he says.
The second factor is this consumer segment’s increasing spending power and disposable income.
“This particular age group has now become economically strong; therefore, they are able to fulfil their beauty needs at this age,” Verma says.
“They are willing to pay more and need the right product fitment. These insights have become available in the digital era where we have a far better ability to connect with consumers and understand their needs. “
Data and Secrets
While we know a lot about consumer behaviours from data, Verma believes it is important to continuously seek out more information. By validating and reaffirming these hypotheses, we can find the correct insights instead of acting on the wrong assumptions.
“Data will help you prove your hypothesis, or it will allow you to understand something that you did not know,” he says.
“It is a tool that provides reality, even if that reality is something counterintuitive. Data gives you the insights to propel decision-making and provides a greater understanding about our consumers, the population, and different institutions.”
“Gut feelings can only be 50% correct, whereas data can take those odds up to 90%.”
Varun Verma is a speaker at our highly successful CDAO Melbourne event. If you want to find out more about the evolution of analytics and how AI is transforming business, we have some amazing sessions coming up at CDAO Melbourne happening 2-4 September 2024, check out the agenda here.
Photo by Kelly Sikkema on Unsplash