Opinion: In real-time enterprises, keep it simple

More and more, I see references to the "real-time enterprise" or the "agile, sense-and-respond enterprise." It might sound like just another IT fad, but I think it's actually something big. It's where IT can really show its value to business.

So, what is the real-time enterprise? At its core, it means continually improving key business processes and adapting them to constant change. This way, we can deliver better value to customers and more profit to the company. In other words, the real-time enterprise is not so much about any specific technology or system as it is about how people use technology to continually deliver business improvements and adapt to change.

I find that the most effective way to employ real-time technology is to automate routine transactions and empower people to deal with all the exceptions. In my company, I'm keeping the automated systems relatively simple. They handle only the 70 percent to 90 percent of business transactions that are standard and well defined. I don't try to have automated systems handle everything. Unnecessary cost and complexity are avoided by keeping systems simple. And keeping systems as simple as possible also reduces the high failure rate that plagues IT projects.

"But wait!" you say. "The world is a complex place. How will simple systems help me deal with it?" The answer, my friend, is that computer systems are better at handling the simple, rote, routine work, and people (not computers) are better at handling the complexity. I don't need lots of fancy artificial intelligence (and complexity and expense) in my systems if I use the natural intelligence of people instead.

I define complexity as any transaction that's an exception to the standard business rules built into my automated systems. When exceptions occur, the automated systems just flag them, trap the related data and send alerts to the appropriate people to examine this data. The systems then return to handling the vast majority of routine business transactions.

There is a lot of power in handling complexity this way. An exception to a standard set of rules will either be because of an error in the data or because the data reflects something out of the routine. The people who are alerted will decide whether the nonstandard data is the result of an input or process error or if it indicates something new in the environment. Either way, it's best to have people, not complicated and hard-to-understand computer systems, making these decisions.

If the data or the business process shows an error, people do root-cause analysis (like Six Sigma) and fix errors at the source. This makes the standard business processes even more efficient, and value is gained from that. If the data shows a new development, then early detection allows more time to create business processes to respond effectively. Big wins can come from this.

I've begun deploying real-time systems like this in my own company. The systems are relatively cheap and quick to build, since they're designed to handle only the standard, routine transactions. And people now have time to do more valuable work. They aren't worn down doing the boring, routine stuff, and they have time to focus on the exceptions, learn from experience and find ways to increase their efficiency. A process of continuous improvement is starting to take hold.

Could it be this simple? Am I missing something? I'll keep you posted.

Join the newsletter!

Error: Please check your email address.
Show Comments

Market Place