• Middleware solved the problem of integrating multiple systems in the 90s
  • The proliferation of technologies and data can create a new integration challenge
  • Updated versions of middleware have returned as a key part of the technology stack

In the 1990s, as the world was completing the transition from the “big iron” age of mainframe computing to the world of PC-based clients and servers, a form of software called middleware became a critical component of the computing landscape. Composed of a mix of business logic and data, middleware connected the 1970s and 1980s world of mainframes – and the often mission-critical applications and data they housed – to the new thin PC clients used at the edge of the network. Without middleware, much of the business applications that ushered in the internet age – from direct partner to partner ordering systems to the first online airline reservation sites – would not have been possible. As mainframes slowly phased out and mission-critical apps and data migrated to the PC architecture, middleware became less common and largely limited to esoteric and specialized uses.

However, just as 90s music and fashion seems to be returning, the concept of providing a bridge between systems to provide a clearinghouse of data and business logic has returned to importance. This trend is driven by the rapid proliferation of systems and data required to support a modern digital marketing effort. For example, personalizing content for an individual requires customer data from the organization’s sales force automation system, analytics and tracking data from the Web site, offsite intent data from data brokers, and organizational data from third-party sources. Connecting this wide array of technology and data requires a new and updated version of middleware.

The modern version of middleware collects and provides data to a wide variety of systems, and then rationalizes it into a structured, consistent set of data ideally tagged and tied to individual customers and accounts. To accomplish this task, systems must be capable of the following:

  • Storing complex and varied data (e.g. company master data, proprietary databases)
  • Executing operations to aggregate and align the data into useful information
  • Using APIs and other integration technologies to connect bi-directionally to other systems
  • Utilizing governance mechanisms to ensure the integrity of the data

Currently these capabilities can be sourced in a variety of ways:

  • Through custom development using off-the-shelf database technologies – like those provided by Microsoft or Oracle
  • As part of a broader personalization or data management system like Evergage or Get Smart Content
  • As a standalone system specifically designed to serve the data integration role of a data management platform (e.g. Oracle’s BlueKai, Adobe’s Audience Manager) or a data analytics platform (e.g. SAP’s Hana)

To determine if you’re a candidate for the modern reincarnation of middleware, ask yourself the following questions:

  • Do you have a large and varied set of data to integrate on an ongoing, real-time basis?
  • Are these data sources changing regularly with new ones added and existing ones updated?
  • Do you have a variety of systems that provide and access to the data?
  • Must the data you collect be rationalized against an existing set of data (e.g. individual or account-level customer information)?

If your answers to two or more of these questions is “yes,” then it may be time to take a trip back to the 90s and consider a middleware solution for your organization.