Innovation doesn't just happen... it evolves out of business desires and it frequently involves changing your own view of the IT value statement. This article focuses on one of the key components needed to make the shift from maintenance to an innovation IT philosophy, Business Service Reliability. As a progressive IT department you must understand exactly what it is, how you can manufacture Business Service Reliability in real time so that you can measure and report your contribution in business language, and how adopting the framework shifts your organization into an innovative business partner.

What Is Business Service Reliability?
Because of our long-lived traditional IT department philosophies, most IT organizations, whether or not they realize it, focus almost exclusively on component availability rather than the overall reliability of their business services and the related customer interactions. The problem, of course, is that you can achieve decent availability metrics for individual tiers of IT services - but still wind up with customer interactions (the business services) that are unpleasant. And frequent painful customer experiences drive customers and their money away from your company. They cause your customers to say bad things about you... in public and online. They undermine your credibility, so that the business sees you as a maintainer of poor systems and an overall liability, not an innovative business partner. The shift to innovation starts with your customer, and understanding what gives them a good experience, and proactively transforming your operating model to manufacture what the customer wants.

Given this, it's obviously critical to define exactly what a good customer experience is and determine how to measure it. Good customer experiences occur when every link in the business service delivery chain performs properly. These links, by the way, are not just infrastructure components. They include calls, methods and other activities of software code that are not monitored on the typical enterprise management console. And it's not enough for these links to simply occur or be available. They also have to perform to the required level while maintaining data accuracy. That is, they have to retrieve or display the correct data in the correct way. In other words, a successful customer interaction occurs when every link in the chain is available, performs adequately, and actually does what it's supposed to do.

Manufacturing Reliable Business Service Delivery
"Reliability," therefore, can in one way be defined as how often you get the customer experience right; it is clearly not component availability. It's how often something - anything - goes wrong in the end-to-end service delivery chain. The exciting thing is that this shift to an innovative mindset can actually be mathematically expressed and measured in the real time. Examine, if you will, the following real-world formula and measurement for Business Service Reliability:

Where m = a link in the chain of delivering a customer interaction is doing exactly what it should, then:

Business Service Reliability = customer interaction link 1,2,3...(availability)^m + (performance)^m + (accuracy)^m

And the Business Service Reliability measurement for any period (usually daily or hourly) would then be:

1-(defective interactions/total interactions)

Again, we aren't just looking at problems with component availability here - although an unavailable component will obviously result in us having a broken link in our service delivery chain. But now we're really looking at all the links in that chain and discovering where they are failing, where they are under-performing, and/or where they are not doing exactly what they're supposed to do. We get two benefits from this experience-centric approach. One is that we can now actually measure the real reliability of our service delivery. Now instead of patting ourselves on the back because we're doing such a great job of keeping our networks and servers up and running, we can take a look at a more meaningful measurement - where is our perfection rate when it comes to customer interactions?

But the second benefit is something even better. This Business Service Reliability approach doesn't just tell us how bad we're doing, it also tells us how to fix the defects in our ‘customer interaction manufacturing production line.' When we look at our defective interactions, we can begin to recognize patterns in our business service delivery. For instance, a critical Java method to get customer account information from the mainframe through the SOA layer has been randomly slow all day. In real time and related strictly to how this defect is impacting the customer interaction we provide, we can investigate the problem with only the resources related to that area of delivery, and then not only alleviate the problem, but eliminate it in the future.

How to Create the Shift
Business Service Reliability proves to the business that your IT philosophy is that of first understanding what the customer is expecting from you, and then creating that with supporting IT components only providing the means to the end. That ‘end' is what enables you and your department to develop ideas that actually mean something to your business and the framework allows you to actually deliver against the vision.