Over the past three decades, insurance organizations and other enterprises have invested billions in agency management systems, CRM platforms, policy administration, claims applications, document repositories, analytics, cloud computing and collaboration tools. By every traditional measure, that investment worked. Operations are faster, processes are digital and information is easier to reach than ever.

Yet executives keep describing the same frustrations. Professionals spend too much time searching and too little time advising. Knowledge sits in separate departments. Important decisions require pulling information from several systems before any analysis can start. Experienced employees retire with decades of expertise that the firm never recovers.

Systems were built around functions

Enterprise software was designed to manage business functions. Agency management systems manage policies. CRM platforms manage relationships. Claims systems manage claims. Financial systems manage accounting. Each one does its job well. The limitation sits between them: information moves from system to system, and understanding stays behind.

Every one of those purchases reflected sound judgment at the time. Together they produced an unintended result. Departments became very efficient while the enterprise as a whole stayed fragmented. Client information lives in one application, claims history in another, financial data somewhere else, regulatory guidance in a separate repository, and the conversations that explain all of it are scattered across email, carrier portals and document folders.

What human middleware looks like

One of the most expensive technologies in every enterprise never appears on an architecture diagram: the employee. Every morning, experienced professionals do the same invisible task. They open several systems, read email, search documents, compare client records, interpret old notes, verify policy details, ask colleagues and piece together past conversations. Only after all of that can they start creating value.

Erik Davison gives this a name: human middleware. Traditional middleware moves information between systems. People move understanding. That difference matters, because highly skilled professionals spend their best thinking on assembling context. Producers spend less time advising clients. Claims professionals spend less time on complex cases. Executives spend less time on strategy.

Over time the organization accepts this as normal, even though it reflects how the enterprise was designed and says little about the people doing the work.

Integration moves data. Meaning stays behind.

Firms have spent heavily on integration: APIs connect applications, middleware synchronizes records, single sign on simplifies log in, and data warehouses consolidate reporting. Each of these creates measurable value, and people still have to build the shared understanding themselves. A policy record can move between systems, while its relationship to underwriting history, client strategy, market conditions and past executive decisions stays behind.

Enterprise Intelligence addresses that gap. Its aim is to connect institutional understanding across the systems a firm already owns. Information becomes enterprise knowledge, which strengthens enterprise memory, enriches context and improves reasoning. Better reasoning leads to better decisions, and learning from those decisions expands what the institution can do next time.

Information → Enterprise knowledge → Memory → Context → Reasoning → Decisions → Learning

Keep the systems you have

This approach keeps the agency management system, the CRM, the financial platform, the claims application and the collaboration tools in place. It lets them contribute to one shared capability. Firms keep their previous technology investments and finally get those investments working together.

In Erik's words, technology rarely limits organizational performance. Architecture does. Every insurance organization he has worked with has talented professionals, capable systems and valuable information, and many still ask those professionals to assemble the full picture by hand before they can apply their expertise.

Questions to ask inside your own firm

  • How many systems does a producer or account manager open before a renewal meeting?
  • When a senior person leaves, which client context leaves with them?
  • How often do teams rebuild an analysis that someone else already did?
  • Can anyone explain why a major decision was made three years ago, with the reasoning written down?

If the answers are uncomfortable, the problem is in the design of the enterprise. That is good news, because design can be changed.

Adapted from Chapter Three, "The Hidden Cost of Fragmentation," in Enterprise Intelligence: A New Theory of Management for the Intelligence Age by Erik Davison.