Two weeks ago, I argued that being absent from an AI-generated answer is now more damaging than ranking second in a search result. The response was strong. The fair follow-up question was: where’s the proof this is already happening, not just theorised?
Two data points, from two different directions, answer it.
1. The Consumer Side Moved First
Across Southeast Asia, 78% of shoppers now say they use AI somewhere in their shopping journey — chatbots, visual search, comparison, translation — with product availability and price comparison the leading use cases. This is not early adoption. This is the majority behaviour.
The AI-mediated decision I described is not an emerging pattern. It is already the default path for most shoppers in our region.
2. Measurement Is Finally Catching Up
The harder problem I raised was different: attribution wasn’t built to see this. If a consumer’s decision shifts inside an AI answer, most dashboards show nothing.
MMA India, working with OptimizeGEO, has just launched the first study of its kind built to close that gap — the IPL 2026 AI Brand Visibility Tracker. Across 84 matches, 10 teams, and 12 industries, it’s tracking, match by match, whether India’s biggest brands are actually being cited and recommended by ChatGPT, Gemini, Perplexity, and Google AI Overview — not simply whether they bought the broadcast spend. The same model has since been extended to track brand visibility through the FIFA World Cup 2026.
This is the beginning of a measurement category that didn’t exist a year ago: not share of voice, but share of recommendation.
3. Trust Is the Piece Still Missing
None of this means the shift is complete. Reddit’s own research, released at Cannes Lions this year, found that nearly half of shoppers still verify an AI-generated recommendation with a real person before they buy. Consumers are asking AI. They are not yet fully trusting it alone.
That gap isn’t a reason to wait. It’s the actual opportunity. Brands that build genuine trust signals now — clear, structured, first-party proof of who they serve and why they’re credible — are building the foundation that both AI systems and sceptical humans will recognise later.
Where This Leaves Marketing Leaders
The prediction was directional. The proof is now regional, sourced from India rather than a US case study. And it points to a discipline shift, not a tooling shift: brands now need to build for two audiences simultaneously — the algorithm making the recommendation, and the human still asking a friend before they act on it.
The organisations already measuring this — rather than assuming their media weight is enough — are the ones that will still be found when the click is gone entirely.
What is your brand doing to earn a place in the answer, not just the algorithm?
