A few years ago, I had a meeting with IT staffers at a large insurance company. The new CEO had asked the IT department to come up with new ideas that would help the company save money and demonstrate it was really innovative. Strangely enough, the mainframe people I was meeting with had the attitude that this was something for “all other IT.” How could they, with their mainframe, be part of something new?
I had my health insurance with this company, so I was familiar with their customer service practices. While we were brainstorming, I asked why they were still sending out monthly letters to all their customers to inform them what they paid out to various hospitals, doctors, specialists, etc. Why not change these communications to emails with a PDF attachment? Surely, this could save money, improve service, and demonstrate “green” thinking. The cost of implementing this change was quite low and the savings turned out to be about 30 million euros (more than $40 million USD) annually.
Not all innovative ideas can be accomplished in such an easy way, but since we mainframers often take pride in the fact that so many of the company’s mission-critical business applications still run on the mainframe, we should be in the frontlines when it comes to optimizing the business services. The process often starts with simple and pragmatic things—take a step back, clear your mind, talk to the users of your systems and applications and, most important, look around at what has changed in the world in the past three years. Here are four simple steps to help you come up with practical, business service innovations:
1. Model. Before any real investments are made, start modeling. This can be a simple model, where you identify what you already have and what can be reused from other IT departments (yes, this means you actually have to talk to the distributed folks), and determine if there’s anything in the cloud you could use. I know it sounds simple, but often companies have almost all the components in-house already but they just aren’t connected.
2. Assemble. Once you’ve identified the various components, make sure they connect. This can be as simple as turning some CICS applications into a Web service, or implementing a simple but intelligent WebSphere MQ protocol. Remember, your data is vital for the company; the easier you can make it available, the more others can do with it! And if you can make that data available in an intelligent way, by taking care of the underlying logic on the mainframe itself, you make it even easier to connect to it. Better yet, you stay in control of the performance and, more important, security.
3. Automate. Before you make your services available for deployment, monitor them as a service. Make the monitoring data (response times, errors, events, and alerts) available to the outside world as a bundle. In other words, if external programs access your data, you also enable other programs to monitor access to that data. This helps provide a single, end-to-end view of the performance of the composite business service. This can be as simple as sending the various alerts (DB2, WebSphere MQ, CICS transaction times) that are part of this service to a central point with a unique identifier.
4. Assure. Monitor your services constantly. Implement trending to ensure you spot the slightest change in behavior. After all, things change over time and it’s cheaper to fix a problem now than it is when alarm bells go off and users start to complain. Don’t say you’re monitoring the database and CICS and WebSphere MQ already; that’s simply not enough. If you deliver a service, you have to monitor and manage it as a service.
As mainframers, we haven’t always thought like this. We managed the various aspects of IT, including security, storage, the application, the operating system, and the TP monitor. Our employers are asking for more innovation and flexibility, and with a little innovative thinking, we can offer that better than anybody else. But it means we must change the way we think about the consumers of our services.
Business service innovation is nothing new, but calling it that and describing how to accomplish it might help us be part of requisite change instead of standing on the sidelines.