Page 59 - Banking Finance June 2024

FEATURES

How AI & automation are making

retail come alive for the new gen

R etail customers are changing so dramatically that The path to AI

retailers say the only way to provide the experiences
Dealing with this level of speed, personalisation, accuracy,
customers expect is to adopt AI and automation.
Yashwanth Kumar, group president of customer loyalty & and complexity is humanly impossible. So, retailers globally
are digitally transforming themselves to be able to use tech-
analytics at Landmark Group, owners of retail brands like nologies like AI and automation. This is the reason why many
Lifestyle and Max, says customers have become very digital of the world's biggest retailers - including Walmart, Ama-
savvy and want the same experience in both online and
offline. zon, Lowe's, Target, Tesco, Lululemon - have big technol-
ogy centres in India. It's the reason why large Indian retail-
If they go to a website, they want recommendations, which ers are charting a similar course.
is what a good sales person would do in a store. And in a
Siddhartha Niyogi, CEO of o9 Solutions India, which provides
store, they want instant billing and self-service, like in online.
analytics and AI/ML solutions for faster planning and deci-
They are also the Blinkit generation - they want quick re- sion-making, says retailers are also using AI/ ML for improv-
sponses, instant rewards and resolutions. This generation, ing the assortment in stores, to optimise channels of distri-
Kumar says, is far more discerning too. "When I was bution, predict demand, to improve effectiveness of promo-
younger, we had this concept called Friday Dressing - you tions, to optimise discounts.
had workwear and then you had Friday dressing, and then
what you wear on weekends. Now, new-gen customers have Ganesan V P, the distribution sector leader at IBM Consult-
10-15 different shopping occasions - partywear, wedding ing, India/South Asia, says AI has become a boardroom topic
wear, engagement wear, hangout with family wear, and so because of the potential it has to solve the challenges in
on. There are those who want comfort in everyday wear, retail. But he says the implementation needs to be thought-
but in party wear, they are not particular about comfort or fully done to get the best outcomes. The first step, he says,
quality, they just want to look nice," he says. is to have a vision and intent - be clear what are the prob-
lems you want to address.
Kumar Rajagopalan, CEO of Retailers Association of India
(RAI), says the concept of loyalty is also changing for retail- Once that's done, you need to ensure you have the data
ers. "The traditional method of loyalty was that, if a cus- necessary - AI learns from data, and the more good data
tomer was not buying from us, then we would go and give you have, the better the AI solution will be. Data, Ganesan
them incentives to come to us. But if they were already our says, is often the biggest limiting challenge, because the
customers, then we would not give them any special price data is in various forms and shapes in different places, and
or gift. Those days are over. Now, we as retailers have to they need to be brought together. Niyogi says while e-com-
show our loyalty to the existing customers, to keep them merce tends to be rich in data, it's much harder to get data
shopping with us," he says. for brick and-mortar stores. But brick-and-mortar, he says,
is today ripe for interventions.
Customers, Rajagopalan says, also expect to be recognised
wherever they are - be it online, offline, whichever geo- The third step is to identify the best AI or Gen AI model.
graphic location they are in - and expect to be served quickly Today's large language models are built on top of a lot of
and accurately. "The attitude is, I'm not going to wait for data. But there are often questions around the quality of
you to give me some excuses, if you recognise me fine, if that data. So, Ganesan says, you may have to finetune that
you don't recognise me, I know somebody who will model, or maybe even build your own model, which can be
recognise me," he says. expensive. (The Times of India)

52 | 2024 | JUNE | BANKING FINANCE