Why Unstructured Data and AI Should Be Every CIO’s Top Priority in 2025

At the start of the year, I was asked what should be a CIO’s top priority for the year ahead. It’s pretty simple: CIOs must go beyond traditional structured data and unlock the full potential of unstructured data. The top priority should be to run at least one analytics project that takes unstructured data - documents, emails, images, text fields - and enriches a key performance indicator (KPI) in the data warehouse.

Why This Matters: Unstructured Data is an Untapped Goldmine

Data leaders know that 80-90% of enterprise data is unstructured, yet most organizations rely almost entirely on structured data for decision-making. Unstructured data is typically text-heavy and generally contains data such as stats, numbers, figures and other facts. So, there’s a veritable treasure trove of data being left untouched. The rise of generative AI (GenAI) has undoubtedly changed the game, making it easier than ever to extract insights from unstructured data and enrich traditional analytics.

Once CIOs understand the value of what’s hidden in their unstructured data and how GenAI can unlock it, there’s no turning back. This shift represents a fundamental change in how businesses approach data strategy, AI-driven decision-making and enterprise intelligence.

How to Execute: The Hackathon Approach

So how do you get to the point of realizing value? I’m always a proponent of a ‘code wins’ approach. Make it simple: run an internal hackathon. The goal? Identify ten critical business questions that could be answered using unstructured data already sitting within the company. The scope can include:

• Existing Word or PDF documents
• Form fills and surveys
• Email messages
• Email attachments
• Text fields in SaaS applications
• Images or scanned records
• Transcripts from meetings or audio recordings
• Web pages

If a large language model (LLM) can process it, it can add value. The key is to experiment with real-world business challenges that could be solved with better access to that previously untapped data.

The Minimal Resources Needed

One of the biggest advantages of this approach is that it doesn’t require an enormous team or budget. A small, focused team can drive results with a business leader who knows what critical questions need answering. They’ll also need a data engineer who understands how to use prompts and is able to create a retrieval-augmented generation (RAG) database to feed AI models with relevant company data.

The Role of AI-Powered Data Integration

The good news? This isn’t just theoretical. There are no-code solutions, like Matillion’s Data Productivity Cloud, that make this process accessible to everyone. By leveraging an AI-powered data integration platform, businesses can extract, transform, and load (ETL) unstructured data into analytics-ready formats without requiring deep machine learning expertise.

The Future of Data Strategy

CIOs who embrace this approach in 2025 will not only drive better business decisions but also future-proof their data strategies. As AI continues to advance, organizations that integrate unstructured data into their analytics pipeline will gain a significant competitive edge.

Want to know more about AI-powered data integration and Data Productivity Cloud? I joined Matthew Scullion (Matillion CEO), Sridhar Ramasway (Snowflake CEO) and a host of industry experts at Zero Data Friction, a virtual event to share the vision and roadmap of intelligent data integration and Data Productivity Cloud.

You can catch up on everything that took place at Zero Data Friction here.