As enterprises rush to adopt agentic AI, Salesforce’s latest report reveals that poor, siloed data — not algorithms — is the real bottleneck holding India back.
Bengaluru, NFAPost: India’s technology leaders are united on one crucial truth: artificial intelligence cannot outpace the data foundations beneath it. According to Salesforce’s latest State of Data and Analytics report, 89% of India’s data and analytics leaders say their organisations must modernise their data strategies if AI is to deliver real business impact.
The finding lands at a critical moment. Enterprises across sectors are racing to deploy AI — particularly agentic AI systems that can autonomously analyse data, make decisions, and act within business workflows. Yet the report exposes a widening gap between ambition and readiness. While 75% of business leaders feel increasing pressure to extract value from data, many remain constrained by incomplete, outdated, or poor-quality information — a structural weakness that AI only amplifies, not fixes.
“AI cannot fix what incomplete data creates,” said Deepu Chacko, VP – Solution Engineering, Salesforce India. “For India to truly unlock the promise of agentic AI, leaders must treat data as a strategic asset — unified, governed, and contextual. The companies that modernise their data foundations today will be the ones that scale AI responsibly and lead the economy tomorrow.”
The Illusion of Being ‘Data-Driven’
On the surface, Indian enterprises appear confident. Nearly two-thirds of business leaders describe their organisations as data-driven. But scratch deeper, and the picture becomes more fragile. Over half (52%) of data and analytics leaders admit their companies struggle to use data to drive core business priorities, while only 51% say they can reliably generate timely insights.
Even more concerning, 54% report that their organisations occasionally or frequently draw incorrect conclusions because data lacks proper business context. In an AI-powered environment, such errors can scale rapidly — turning flawed inputs into enterprise-wide misjudgements.
AI’s Fast Track, Data’s Weak Foundation
AI has now become the number one data priority for Indian organisations — a position it has held consistently since 2023. This urgency is palpable: 56% of data leaders feel pressure to deploy AI quickly, often before the data backbone is ready.
The consequences are already visible. Nearly 39% of leaders lack full confidence in the accuracy and relevance of AI outputs, while organisations estimate that 25% of their data is fundamentally untrustworthy. Predictably, 94% of companies with AI in production report experiencing inaccurate or misleading outputs, and half of those training their own models say they’ve wasted significant resources due to bad data.
In other words, the promise of AI is being undermined by the very foundations meant to support it.
When Valuable Data Is Locked Away
Data quality is only part of the problem. Accessibility is the other. Although 89% of leaders agree that unified data is essential to meeting customer expectations, organisations estimate that 26% of their data remains siloed or unusable— and alarmingly, 75% believe their most valuable insights sit within this inaccessible pool.
The fallout is extensive: diminished AI capabilities, fragmented customer views, weaker personalisation, and missed revenue opportunities.
To counter this, Indian enterprises are increasingly turning to zero copy data integration — a method that allows systems to access distributed data without physically moving or duplicating it. More than half (52%) of organisations have adopted zero copy architectures, and those that have are 40% more likely to fully connect customer data and 22% more likely to succeed with AI initiatives.
The Governance Gap in the Agentic Era
Technology alone will not solve the challenge. Governance is lagging. Only 52% of data leaders say their organisations have formal data governance frameworks, even as 90% agree that AI requires entirely new approaches to security and oversight.
At the same time, usability remains a hurdle. 69% say translating business questions into technical queries is error-prone, while 95% of business leaders believe they would perform better if they could query data using natural language — a gap agentic analytics tools are beginning to address.
A Defining Moment for Indian Enterprises
The Salesforce report makes one thing clear: India’s AI journey will not be defined by how fast companies deploy models, but by how deliberately they modernise data — making it trusted, connected, and governed.
As Chacko puts it, “Agentic AI isn’t the next technology — it’s the next revolution. AI agents handle routine tasks so humans can focus on creativity, relationships, and impact.” But that revolution, the report suggests, will only succeed if enterprises first fix the data reality beneath the AI dream.
