India’s tech leaders bet on data modernisation as AI ambitions hit reality check

India’s technology leaders are largely aligned on one message: artificial intelligence will not deliver real business impact without fixing the data underneath. According to Salesforce’s latest State of Data and Analytics report, 89% of data and analytics leaders in India believe their organisations must modernise data strategies to make AI work at scale.

The finding comes at a time when expectations from data teams are rising fast. Around 75% of business leaders say they face growing pressure to extract business value from data. Yet the report shows a clear disconnect between ambition and execution, driven by incomplete, outdated, and poor-quality data. As companies move into what Salesforce calls the “agentic AI era,” this gap is becoming harder to ignore.

Data driven in theory, strained in practice

Nearly two-thirds of business leaders, about 66%, describe their organisations as data driven. The reality appears less convincing. More than half, 52%, of data and analytics leaders say their companies struggle to use data effectively for business priorities.

Only 51% of business leaders say they can reliably generate timely insights. At the same time, 54% of data leaders admit their organisations occasionally or frequently draw incorrect conclusions due to missing business context. Poor data quality remains the single biggest reason companies fail to become truly data driven.

AI raises the stakes for weak data foundations

AI has become the top data priority for Indian organisations, a position it has held since Salesforce’s 2023 report. This focus is also creating pressure. About 56% of data and analytics leaders say they feel rushed to implement AI.

That speed comes with risk. Nearly 39% of leaders lack full confidence in the accuracy and relevance of AI outputs. Data leaders estimate that 25% of their organisation’s data cannot be trusted. The consequences are already visible. Among companies running AI in production, 94% report experiencing inaccurate or misleading outputs. Half of the organisations training or fine-tuning their own AI models say they have wasted significant resources due to poor data.

Valuable insights stuck in silos

Even when data quality improves, access remains a major hurdle. Around 89% of data leaders believe unified data is critical for meeting customer expectations. However, they estimate that 26% of enterprise data is siloed or unusable.

More worrying, 75% believe their most valuable insights are hidden within this inaccessible data. Over eight in ten leaders link trapped data to weaker AI capabilities, limited customer visibility, reduced personalisation, and missed revenue opportunities.

Zero copy and natural language gain attention

To address these challenges, organisations are rethinking how data is accessed and governed. About 52% of Indian companies are adopting zero copy data integration, which allows data to be accessed across systems without moving or duplicating it. Salesforce says companies using this approach are 40% more likely to have fully connected customer data and 22% more likely to succeed with AI initiatives.

There is also growing interest in natural language interfaces. Around 69% of data leaders say translating business questions into technical queries is error prone, while 95% of business leaders believe they would perform better if they could ask data questions in plain language.

Governance lags behind AI adoption

Despite the urgency, governance remains a weak spot. Only 52% of data leaders say their organisations have formal data governance frameworks. At the same time, 90% agree that AI requires entirely new approaches to governance and security.

Commenting on the findings, Deepu Chacko, VP of Solution Engineering at Salesforce India, said AI cannot compensate for weak data foundations. He added that organisations treating data as a strategic, governed asset today are more likely to scale AI responsibly and lead future growth.

The message from India’s tech leadership is clear. AI may be the headline, but data modernisation is the real work happening behind the scenes.