"Complete integration" means all systems that collect and store information can share data with other systems that collect and store information. This concept has been around since we began playing the data integration game back in the 1990s. It's such an important concept that I wrote three books about it.
However, we're nowhere near where we should be. The fast pace of technology iteration has left integration out in the cold since many enterprises are moving too fast to deal with integration issues. Now there is the cloud and greater distribution of information that may be used by other systems.
If complete integration is to be achieved, we need to think differently about how we build and deploy systems. Integration needs to be a priority, commensurate with its strategic value.
The ability to automate processes and decisions is the best example. If you have access to all the data in all the systems, you can drive processes around that near-perfect information. "All the data in all the systems" means, for example, the ability to see inventory levels, price trends, logistics, and even the weather, then define the best decisions that can be automatically made during the normal course of business.
The actual value depends on how much data drives decisions in your enterprise. About 90 percent of businesses make massive amounts of data-driven decisions each day, and they do so without the data visibility that they need. As a result of that imperfect data-driven decision-making, millions of dollars are lost. Most businesses accept the inefficiencies, but that's unacceptable to me.
Enterprises have of course already done some thinking about complete integration. You may even have a data-integration group and some underlying technology to support the notion.
However, much of that thinking and technology may need to be rebooted because of the major technology advances in the past few years. You might even have to start from scratch.
As we move to cloud and the data silos get more distributed, let's focus on achieving complete integration. The work and money to build the needed data integration should be a no-brainer for most enterprises, even if they've invested some work previously.