On this page
AI agents finish 80% of shopping tasks on Indian ecommerce sites, yet average readiness is only 43/100: FTA Global
Our read
The desk expects a split: the top-scoring sites keep their lead while the weakest sites change little, so the 43/100 average rises only modestly even as 80% task completion looks healthy.
For operators
Agents already complete about 80% of shopping tasks, but delivery charges (74%) and product-and-price discovery (73%) trail return-window answers (91%), so make those two machine-readable first, since 18 of 42 scored sites sit below 20/100 and a fast fix gets you ahead of them.
Watch
A follow-up FTA Global study showing the 43/100 average rising, or the count of sub-20 sites falling from 18
The report,
FTA Global's Agent Readiness Study 2026 found AI agents completed 358 of 450 shopping tasks (80%) across 50 Indian ecommerce sites. The average readiness score was 43 out of 100, with boAt at 95, Flipkart, Mamaearth and Moglix at 92.
- Agent completion: return-window queries91%
- Agent completion: all shopping tasks, 358 of 45080%
- Agent completion: delivery charges queries74%
- Agent completion: product-and-price discovery73%
Channel facts
From the report. Source details below
| Sites scoring 80 or above: | 12 of 42 |
|---|---|
| Sites scoring below 20: | 18 of 42 |
What it means for online and offline
With average readiness at just 43/100 and scores split between 12 sites at 80+ and 18 below 20, agent-driven shopping could favor the few prepared leaders like boAt, Flipkart, Mamaearth and Moglix, though this is a single study of 50 sites and should be treated as directional.
Signals to track
- Delivery-charge success moving above 74% or product-and-price discovery moving above 73% in a re-test
- A top-scoring site announcing agent-specific terms, rate limits or blocks
- Retailers or platform vendors publicly citing agent-readiness scores in launches or earnings commentary
- Any of the sites in the 80+ group slipping, which would suggest the lead is not durable