Five layers of value
One way to examine a data company's proposition is through five archetypal layers of value.¹
At the foundation are data proprietors, whose assets are essentially proprietary information valued by customers (particularly in opaque markets like commodity flows). Data refiners collect, clean, and combine information into useful data products. Access enablers make information usefully available through delivery mechanisms like data feeds and APIs.
Further along the chain, workflow integrators provide tools that help customers analyze information and embed it within their workflows, while insight interpreters turn data into expert opinions, forecasts, and qualitative analysis, increasingly supported by technology.
AI disrupting value across layers
Most data businesses’ value proposition is a combination of several of these layers. For many of them, AI is now changing that balance.
Data refinement is a prominent example area experiencing significant change. Collecting, classifying, and structuring information has long demanded significant investment in specialist teams and labor-intensive data ops. AI can now very effectively perform a lot of this work faster and at lower cost. Their reliability varies by data domain (depending on tolerance for error), and domain expertise remains important in many markets.
AI is also prompting information providers to rethink how customers access, analyze, and use their data.
Technologies like MCP and agentic are increasingly changing how customers expect to consume data, allowing them to co-mingle datasets to find new insights across previously unrelated areas.
Despite these changes, trust remains a durable moat in this sector. Data providers that have a reputation for robust domain expertise can continue to thrive as long as they learn to evolve their delivery methods with AI.
Finding the next source of advantage
For leadership teams, the five layers provide a way to examine where future differentiation might emerge.
For example, a provider with valuable proprietary datasets could strengthen its position by leveraging AI technologies to unlock new ways customers can productively use their data.
Another business might find greater opportunity in embedding its data more deeply into customer workflows, making them more operationally ‘sticky’.
Commercial models will also need attention.
As customers use information through APIs, MCP, automated systems, and agentic workflows, providers may need to reconsider how access, consumption, and value are reflected in their pricing.
A telling case in point here is how seat-based pricing will become increasingly sub-optimal as the link between number of direct human users accessing the data and economic value delivered becomes weaker for many data companies.
As AI allows customers to find new use cases for a provider’s data, finding ways to systematically monetize these use cases, for example through rights licensing, will also be crucial.
Which layer are you standing on?
AI investment is becoming a familiar part of product development across the data and information industry. The more consequential decisions concern which capabilities businesses should develop, protect, or rethink.
The five layers offer a useful starting point for examining those choices. They also encourage a question that deserves regular attention as technology and customer expectations evolve:
Which parts of your proposition will customers value most in five years, and are you investing in them today?