Eight in 10 Southeast Asian shoppers turn to AI for product discovery: Bain

A trust gap remains around allowing AI to complete purchases.

Around 80% of shoppers across Southeast Asia are already using artificial intelligence (AI) in their shopping journey, with the technology mainly being used for recommendations, promotions and product comparisons.

Nina Kaspar, a partner at Bain & Company, said AI has rapidly become a shopping tool for consumers, after playing no role in the shopper journey just three years ago.

“Three years ago, AI was playing no part in the shopper journey. Today, majority of shoppers, they use it somehow,” she told the Retail Asia Summit 2026 in Singapore.

However, AI use remains concentrated at the top of the shopping funnel, Kaspar said. Consumers are using it for recommendations, finding promotions and comparing alternatives, but adoption remains limited when it comes to allowing AI to complete purchases.

“The real question mark, or the big question is, when will it move to the bottom? When will it do the actual purchase? When will the AI not just be the advisor, but really becoming the agent?” she said.

Kaspar said shoppers could eventually delegate purchases to AI, such as asking it to find and buy a birthday present below a certain price and have it delivered by a specific date.

“Today, there is still a trust gap,” she said. “But people are open to experiment, and as there will be new and more offers, the direction of travel is clear, and it will continuously to go there.”

For retailers, Kaspar said the shift requires more than adding individual AI applications. She said retailers need to rethink how their businesses operate with AI at the centre, including their channels, personalisation, commercial decisions and operations.

Kaspar cited Brazilian retailer Magalu’s virtual personality Lu, which has been developed into a conversational shopping channel on WhatsApp. Shoppers can use the channel to ask about products, compare options, check out and pay, as well as track purchases and access after-sales support.

Kaspar said the WhatsApp channel has achieved more than three times the conversion rate of the retailer’s other apps or channels, alongside a higher net promoter score.

“We have been optimising our websites and apps for the last 10 over years, but actually, it's now moving into having a conversation with our customers,” she said.

Zhu Hui, also a partner at Bain & Company, said retailers also need to establish clear priorities rather than pursue every available AI opportunity. “You can't go after every single opportunity,” she said. “You need to set a very clear objective so that we know what AI is for.”

Retailers also need to decide whether they want to be early market leaders or fast followers, as well as how quickly they expect returns from their AI investments, Hui said.

She added that retailers must also determine whether AI solutions should be developed internally, bought off the shelf or built with the help of external vendors.

Hui said these decisions need to be aligned across the leadership team before retailers begin pursuing individual AI solutions.
“If you don't have these things discussed, laid out, and aligned as a leadership team upfront, you're going to get confused in this journey because it's just too tempting to look at so many solutions,” she said.

For larger AI initiatives, Hui said retailers will need cross-functional teams involving business and technology leaders, rather than leaving the development of AI tools solely to technology teams. “Building the tool is just a very small part of that.”

Hui said retailers should determine whether they are seeking revenue growth, operational efficiency or innovation, and decide which areas offer the greatest value.

She also said proprietary data, workflows and organisational capabilities will become increasingly important as access to AI tools becomes more widespread. “Proprietary data is going to be your key asset,” she added.

Hui urged retailers to move beyond small-scale AI experiments and consider broader business transformation.

“Many of the companies have been trapped in this kind of micro productivity improvement phase last year, I would say, doing a couple of use cases and trying to get some ROI out of them. Many failed,” Hui said.

“And now it's time to take a different lens, to take on the transformational lens and think about the nine boxes that I just walk you through and think about reimagining your business processes, thinking about leveraging agents,” she said.