Flipkart moves beyond keyword search with AI-led product discovery
Flipkart is using conversational search, voice, visuals and adaptive feeds to personalise product discovery. CPTO Balaji Thiagarajan outlined longer-term plans for shopping agents that help select products, scout deals and manage replenishment, while keeping consumers in control.
The development
Flipkart serves more than 22,000 pincodes in India and is using AI to personalise product discovery through conversational search, voice, visuals and adaptive feeds. CPTO Balaji Thiagarajan outlined longer-term shopping agents for product selection, deal scouting and replenishment, with consumers retaining control.
The numbers
- more than 22,000 pincodes
Why it matters to operators and investors
Strengthen product metadata and visual assets for Flipkart’s conversational, voice and visual discovery rather than relying on keyword optimisation alone.
What to watch next
- Repeat usage of conversational, voice and visual discovery—not just feature launches or initial trials.
- Conversion, abandonment and return rates versus comparable keyword-search sessions.
- Changes to seller catalog requirements, ranking guidance or discovery analytics.
- Sponsored recommendations appearing in AI interfaces, including disclosure and advertiser reporting.
- Any agent pilot specifying purchasing permissions, confirmation steps, spending limits and cancellation controls.
The counter-case
This may be an interface upgrade rather than a discovery breakthrough. Conversational, voice and visual search could add cost and friction without improving purchase outcomes; longer-term shopping-agent plans are not evidence of deployed capability.