A modern data strategy rests on three pillars: an enterprise data platform, practical data governance, and empowered business users. AI has made all three more urgent.
Asset managers hire top talent and want to get the most out of them. To do that, people need the right information, tools, and culture to succeed. That means freeing data and promoting a data culture where people base decisions on data-driven insights. Data is a product that drives business outcomes. Success comes from cultivating the best ideas and making the best decisions based on the best available data.
Data trapped in applications
A key problem is that applications often trap data, creating silos. Ask a simple question, such as the firm’s exposure to a single issuer across every portfolio. At many firms, the answer takes days. Someone has to pull pieces from several systems and assemble them by hand in a spreadsheet.
That problem is getting harder, not easier. More firms are adding private credit, private equity, real estate, and infrastructure alongside their public markets business. That brings in data unlike anything their systems handle today. Think quarterly valuations, capital calls and distributions, and manager reports that often arrive as PDFs. That data frequently lives on separate platforms from the public book. Meanwhile, clients, boards, and investment committees increasingly expect a total-portfolio view across public and private markets. No firm can provide one while its data stays locked inside the systems that produced it.
Becoming a data-driven organization requires access to high-quality information from across the enterprise. Firms need to free business users from application constraints and give them a holistic view of enterprise data.
Why data is the new application
Our point of view is that data is the new application. Architectures need to evolve to put data center stage.
AI makes this truer today than when we first wrote it. Increasingly, people will get answers by asking questions in plain language, not by navigating an application’s screens. As that happens, the interface matters less. The quality, accessibility, and meaning of the underlying data matter more. An AI assistant is only as good as the data it can reach, and its context for interpreting that data.
We see three pillars for a modern data strategy: an enterprise data platform, data governance, and empowered business users. Together they support what every firm is ultimately after: a data culture. Trusted, accessible, and understandable data empowers everyone to make smarter decisions and unlock the organization’s full potential.
We explain the pillars in reasoning order, starting with the business users whose needs drive everything else. Implementation runs the other way. The platform comes first, governance builds on it, and empowered users are the payoff.
Empower business users
Firms need to make the most of their subject matter expertise. That means empowering business users to work with enterprise data without IT bottlenecks. This begins with access to data and data tools. Business users have become far more tech-savvy, and AI is lowering the bar further. A portfolio manager who has never written a database query can now ask the data a question and get an answer.
IT departments have traditionally tried to restrict user-developed solutions, aka “shadow IT.” They need to pivot to prioritize self-service and allow the freedom to innovate. Restriction has always had a cost, and AI adds a new one. When people lack sanctioned access to the data and AI tools they need, some will find their own. Some will even paste firm or client data into public AI tools. The best defense against shadow AI is giving people a safe, official path that’s easier to use.
Provide data governance
Data governance provides a critical foundation for managing and supporting enterprise data. It keeps data accessible, available, secure, and of high quality. It provides self-service resources such as a data catalog, a business glossary, and transparent data lineage. And it gives each data product an owner who answers questions, investigates issues, and evolves the product over time. We summarize the outcomes governance should deliver as FASTIR: findable, accessible, supported, trusted, interoperable, and reusable data.
AI raises the stakes. AI doesn’t just use bad data; it amplifies it. A fluent, confident answer built on a bad definition is more dangerous than a wrong spreadsheet number, because it’s more convincing. The business glossary and catalog that governance produces give AI the context it needs. They tell it what “AUM” means at your firm, or which performance figure is official.
Organizations must still resist the urge to overregulate. People use a lightweight framework that prioritizes agility and clarity. They work around a heavy one.
Architect an enterprise data platform
The enterprise data platform is the starting point. Its job is to provide curated data: data gathered from across the enterprise and organized for confident use. We advocate evolving from legacy application-centric architectures to a data-centric architecture that makes data a first-class citizen. Over 1,000 technology thought leaders have signed the Data-Centric Manifesto. It states, “Data is the center of the universe; applications are ephemeral.” Applications are valuable, but the underlying data is often more valuable and will usually outlive the application. We make the full case in Applications Are Ephemeral: The Case for a Data-Centric Architecture. It shows how a data-centric architecture makes even core systems like accounting far easier to replace.
The technology has matured considerably since we first wrote this article. Cloud data warehouses and lakehouses are now mainstream, well-proven foundations for investment managers. A lakehouse combines a data warehouse’s structure with the ability to store almost any data, including documents. That flexibility matters more as private markets and AI both bring documents and other unstructured data into scope. Data providers increasingly deliver data directly into these platforms, which removes a whole category of integration work. The Investment Data Hub we built for one client shows what this looks like in practice.
Build the platform around a real question
The question for asset managers is no longer whether these technologies work. It is how to get business value from them quickly, and without a multi-year program before anyone sees a result.
Implementation starts with the platform, but don’t build it in a vacuum. Anchor it to a question the business can’t easily answer today, like single-issuer exposure across public and private portfolios. Then deliver the answer incrementally. Each delivery extends the platform, then the governance around it, and finally the users’ confidence. Positive business impact arrives regularly, not all at once at the end. The firm’s data foundation grows in step with the value it provides.
If you want your data at the center of your strategy and ready for AI, contact us. We’d welcome the conversation.