By Sanjana B

Copyright thehindubusinessline

As AI becomes embedded in daily routines, even casual conversations with digital assistants are no longer just exchanges of convenience. Increasingly, companies are exploring ways to repurpose these interactions into marketing data.

Meta Platforms announced on Wednesday that it will start using people’s interactions with its GenAI tools to personalise content and advertising across its apps — Facebook and Instagram, starting in December. Meta’s plan to mine exchanges with its AI assistant for ad delivery spotlights a growing industry trend — turning personal conversations into marketing data, while reigniting debate over where to draw the line between service improvement and monetisation.

Digital marketing

Conversation adds dimension to the persona, which traditional digital marketing does not. They also provide data on intent and their related sentiment, and can be used to qualify an opportunity, Imtiaz Bellary, MD & Co-Founder of Engati, explained.

As consumers leverage AI for their day-to-day tasks, they leave traces of their likes and wants. Be it conversations with the assistant or content they are engaging with or activities they intend to complete, the behaviour can be anonymised for better profiling.

However, not every piece of data collected is, or should be, reused for marketing. Tabrez Alam, Chief – Business Strategy & Data Alliances, Segumento, noted that often, data is first applied to make the service itself better.

For instance, when you watch a show or a movie on streaming apps like Hotstar or Netflix, the data is used to fine-tune recommendations. This, he said, is service improvement, which users expect. However, if the same viewing habit is shared with an advertiser who starts pushing ads for related products outside the app, the experience changes; People start asking: ‘Was my data used to help me, or to sell to me?’

Sales moment

“It’s tempting to think about every chat, click, and call as a chance to sell. But if companies treat every interaction as a sales moment, customers will tune out quickly. Nobody wants to feel like they’re being sold to all the time. What conversational AI does well is uncover intent. If a customer is asking a bank chatbot about home loan interest rates, that’s a signal. It doesn’t mean you immediately push a product, but it tells you the person is exploring, and that’s valuable context. Done right, these insights help brands be more relevant when the customer is ready, instead of spamming them when they’re not.”

The risk arises when insights are stretched too far — when a casual question is treated like a buying signal. This is when personalisation flips into manipulation.

Transparency boost

“The moment personalisation feels like surveillance, people get uncomfortable. Transparency goes a long way. If people know their data is being used in a safe, aggregated way, and that they have a say in opting in or out, trust builds naturally. Not every conversation should lead to a pitch,” Alam highlighted.

In India, the DPDP Act has nudged companies towards being more transparent so users know when their data is improving a service they are using, and when it is monetised for marketing.

Nitin Mahajan, Founder & CEO of BrandBooster.ai and QuickAds.ai, echoed this, adding that not every interaction needs to be monetised. Conversational AI can surface buying signals, which can convert naturally, simplifying it for the user who intended to buy anyway. But it also identifies support needs, frustrations or feedback of customers, all valuable to brand growth. The real opportunity is to use conversations to build long-term relationships, not just immediate transactions.

Published on October 2, 2025