Ecosystm study: Indian firms scale artificial intelligence, but ROI proof is toughest

AI adoption in India is shifting from experimentation to execution. But while enthusiasm for agentic and generative AI is high, most organisations are still struggling to turn pilots into measurable business value.

A new report by Ecosystm, commissioned by Snowflake, examines how organisations across the Asia Pacific and Japan (APJ) region are applying AI in real business settings. Titled Making AI Work: Strategy, Data, and the Power of Ecosystems, the research draws on responses from over 700 business and IT leaders.

From proof-of-concept to deployment

In India, adoption is moving quickly from proof-of-concept to deployment. The study finds that organisations are piloting or rolling out agentic and generative AI across customer experience, marketing, operations and IT. Customer-facing use cases dominate:

• 69% are evaluating AI-driven customer interactions across channels
• 65% are using AI to generate marketing content
• 58% are improving chatbot responses

However, momentum slows when it comes to scale. According to the report, 77% of Indian organisations cite demonstrating clear return on investment as their biggest challenge. Regulatory and compliance concerns follow closely, flagged by 66% of respondents.

Lack of data quality security and accessibility

Data readiness remains a critical bottleneck. Indian organisations report persistent issues with data quality (60%), data security (54%) and data accessibility (50%). These weaknesses directly affect AI outcomes, with only 23% of Indian companies saying AI is fully integrated into their business strategy.

The report also highlights a technology gap. Just 38% of organisations across surveyed countries have invested in tools that allow them to analyse unstructured data, despite its growing importance for generative and agentic AI systems. Fragmented data platforms, poor metadata management and limited monitoring of model performance are cited as common reasons AI projects stall.

Commenting on the findings, Vijayant Rai, Managing Director- India at Snowflake, noted that enterprises are now focused on extracting tangible value rather than chasing novelty. He emphasises that trusted data foundations and structured roadmaps are essential before AI can be embedded deeply into business strategy.

Partnerships are emerging as a key enabler. The research finds that 83% of Indian organisations are already working with, or plan to work with, technology partners to address skills gaps, platform complexity and data challenges. According to Dhiraj Narang, Director and Head of Partnerships- India at Snowflake, ecosystems involving cloud providers, system integrators and domain specialists are becoming central to achieving AI-led returns.

Practises to address ROI challenge

To address the ROI challenge, Ecosystm outlines five practical practices. These include balancing quick efficiency wins with long-term capability building, measuring ROI across the entire AI lifecycle, integrating fragmented AI tools, investing in strong data and skills foundations, and recognising the long-term cost of delaying AI adoption.

The message from the Ecosystm report is clear: Indian organisations are serious about AI, but success now depends less on experimentation and more on execution discipline. Without reliable data, integrated platforms and clear ROI frameworks, AI risks remaining stuck at pilot scale. Those that invest early in foundations, governance and partnerships are more likely to turn AI into a durable competitive advantage.