An Indian FMCG brand just overtook a multinational on AI search.
@ITCCorpCom #1 in Brand Mention (15.62%) and Share of Voice (31.62%) on ChatGPT, ahead of McCain and Godrej Yummiez.
The bet was placed 8 months ago. Story ๐
Rufus is now Alexa Shopping. The rename is the small news. The bigger one is that it does not return what the search bar returns. Brands spent ten years optimising the search bar. That work does not carry over.
Amazon already tells sellers what customers ask & which pain points to cover. That is the input side, and it does feel obvious. The output side is invisible. You cannot see what the assistant actually recommended, or why the other brand surfaced instead. That gap is the problem.
Rufus is now Alexa Shopping. The rename is the small news. The bigger one is that it does not return what the search bar returns. Brands spent ten years optimising the search bar. That work does not carry over.
Search a product in Amazon's search bar, then ask Alexa Shopping for the same thing. You get different results. Same catalogue, same brand, two ranking systems. Every brand has a decade of playbook for the first one. Almost nobody has one for the second.
ChatGPT referral share is now material and Amazon is losing ground. ChatGPT drives 20%+ of Walmart's and Etsy's referral traffic, ~15% for Target, ~10% for eBay while Amazon sits under 3% and is falling 18% month-over-month
@htTweets covered the shift to agentic commerce this week.
AI agents won't just recommend products. They'll discover, research, and buy for you.
What it means for brands, in my words: "Earlier, brands worried about ranking on top in a marketplace. Now they need to be within the purview of the AI agents."
That's the layer we're building at @TruCommerce
Got quoted in @FinancialXpress this week.
Our bet at @TruCommerce : all commerce is going to route through agents.
Buyer agents (ChatGPT, Google AI-mode, Claude) will talk to seller agents to discover and transact.
The shopper states intent. The agent discovers, researches, and buys.
That layer doesn't exist yet. We're building it.
1/ We just got published on ChatGPT's app store.
Asva AI is officially live. People find us, talk to us, and buy without ever leaving the chat.
Here's why this matters for every brand ๐
4/ 12 months from now, being on ChatGPT will be as basic as having a website.
The brands that move first own the shelf.
Free to start. Live in weeks, not quarters.
Hot take: Most D2C sites are invisible to AI agents in 2026.
@Shopify Plus + @Google UCP changes that overnight. Standard stores?
Still stuck in SEO/ browsing era.
D2C founders on Shopify: Is Plus worth โน2L+/mo for agentic scale?
Now everyone's racing to win orchestration. What's interesting is who actually plays there.
Mass market custom orchestration is still gated. Hermes, OpenClaw are solid but the barrier to implement is real.
Curious how HarnessAgent handles state across brain swaps. That's where most abstractions leak.
The agent economy is way more real than people think. Agents don't consume like humans.
They need tools, APIs, context, compute. All in granular nano-transactions.
The consumption model shifts from monthly SaaS seats to tokenized, on-demand, per-call. That's a fundamentally different internet being built right now.
@ashpreetbedi The interesting part is which slice of chat history becomes memory and how it gets reusable.
Humans handle this effortlessly but it's not in our control. If we want to give agents that same autonomy, what's the right framework? Curious how you handle memory pruning at scale.
@0xJeff Agentic commerce stops being a trend when you can trace a single transaction end to end: agent discovers merchant agent via A2A, negotiates, pays via scoped mandate, human just gets the package.
We're already enabling these transactions for brands. This isn't a prediction.