According to ZDNet, Salesforce’s State of Data and Analytics Report reveals that 63% of business leaders now call their organizations “very data-driven,” up from 53% just last year. But here’s the catch: nearly two-thirds of technical leaders admit their companies struggle to actually drive business priorities with data. Data and analytics leaders estimate that 26% of their organizations’ data is “untrustworthy,” and 42% of business leaders say their data strategies don’t fully align with business objectives. The research surveyed 3,800 data leaders and 3,852 business leaders worldwide, showing that 84% of CIOs believe AI will be as significant as the internet, yet 84% also agree that AI’s outputs are only as good as its data inputs. Companies are pouring money into data infrastructure, with CIOs spending four times more on data infrastructure than on AI itself.
<h2 id="the-unstructured-data-problem”>The unstructured data problem
Here’s where things get really messy. Between 80% and 90% of enterprise data is estimated to be unstructured, and 70% of data leaders believe their most valuable insights are trapped in that unstructured mess. Think about that for a second – we’re building these sophisticated AI agents to revolutionize business, but they’re trying to drink from a firehose of disorganized documents, emails, spreadsheets, and who knows what else. And data volumes are growing at 30% annually, up from 23% just last year. So the problem isn’t just getting worse – it’s accelerating.
The trust crisis
Business leaders are basically flying half-blind. More than half (54%) aren’t entirely confident that the data they need is even accessible in the first place. Data and analytics leaders estimate that 19% of their companies’ data is completely trapped. And get this – 49% of data leaders say their companies occasionally or frequently draw incorrect conclusions from data because they miss or misunderstand business context. We’re building AI systems on foundations that half the organization doesn’t trust. That’s like constructing a skyscraper on quicksand.
The AI pressure cooker
Now AI is turning up the heat. With 93% of organizations having at least one AI instance in their tech stacks, the pressure to perform is immense. Real-time data has suddenly become the top data challenge, surpassing even security threats and data quality issues. And 91% of business leaders believe the rise of AI makes it more important than ever to be data-driven. But here’s the irony: we’re using AI to solve problems that exist because our data is a mess, while simultaneously expecting AI to perform miracles with that same messy data. It’s a classic chicken-and-egg situation that’s leaving everyone frustrated.
The human factor
The human element might be the biggest bottleneck of all. 92% of data leaders cited lack of data fluency among staff as a major limitation. The average enterprise uses 897 applications, but only 29% are connected. Basically, we’ve created these digital kingdoms where nobody speaks the same language. That’s why 93% of business leaders say they’d perform better if they could just ask data questions in natural language. They want to have conversations with their data platforms, not fight with technical queries that 63% of data leaders admit are prone to error. The promise of agentic AI is compelling, but we’re trying to run before we can walk.
What comes next
So where does this leave us? Companies are caught between the exciting promise of AI transformation and the harsh reality of their data chaos. The State of Data and Analytics report shows that top priorities now include building AI capabilities, providing real-time data access, and improving company-wide data fluency. But these aren’t quick fixes. We’re talking about fundamental cultural and infrastructural changes that most organizations have been putting off for years. AI isn’t creating new data problems – it’s just exposing the old ones we’ve been ignoring. And now the clock is ticking louder than ever.
