IT Credibility Challenge #2: Inconsistent View of IT's Performance - Page 1

When it’s all working, no one notices. When big crises have been thoughtfully and heroically avoided, no one notices. When there is an executive email outage for 30 minutes, now that’s newsworthy! When an executive’s call to the help desk got the wrong routing, that can trigger an outright campaign against IT.

These random incidents have a way of casting a dark shadow over the reputation of the IT department. Because you can’t really prevent all such incidents from occurring, the only way to counteract their damage to your credibility is to make sure you also have a steady stream of clear, consumable, positive news to set the right context for the periodic “disaster”.

The Magic Bullet - Reporting

Here’s a very common situation. Users are complaining about the performance and availability of a particular IT Service. IT is frustrated because the service is performing to the agreed service levels. Telling users this only irritates them more.

But surprisingly, when you tell them this on a regular and consistent basis with graphs and charts, they get happy. The best part is, you usually don’t even need to improve the service! When IT organizations begin to regularly measure and publish their service level achievements, the perception of the business users changes to: “Wow, that IT service has really improved”—even when it’s unchanged. I’ve seen it over and over again. It’s a good deal.

Pitfall: “We don’t use SLA’s”

Yikes! There is a definite dynamic that exists in some businesses where SLA = legal contract = “I don’t want to do that”, i.e., “If I miss a performance metric, then lawyers are going to get involved and it’s going to be even more annoying.”

This is not a typical outcome, and it’s not a reason to avoid clarifying what you are delivering and how you intend to measure it. Service level agreements (SLA) can be informal in the sense of their “binding consequences”, but they need to be formal and specific about articulating expected service levels. This has many benefits for you.

Having a specific understanding about objectives and expectations, and equally clear guidelines to measure your team’s performance will help you deliver better, and also create reports that are more meaningful, namely, “We did what we said we would do.”

The takeaway? If you haven’t committed to anything specific, you never get the chance to show you have done it.

Measure things people care about. Having a long list of network availability, server capacity, or disk utilization metrics is meaningless to a business person. You need invest time and energy to find out what is meaningful, and start measuring those things.

Talking to your business counterparts about business objectives will give you a much clearer understanding of two things: peaks and disasters. Funny how those go hand in hand when you haven’t talked about it. Ask them their view of a business disaster. Understand their view of an IT disaster. Learn about special situations and business scenarios that will drive peak usage.

If you start measuring things like the performance of the cash registers from 7 am-to-9 am in the coffee house franchise’s key locations, or the ability to re-price all the retail merchandise on the day after Christmas, or the performance and availability of the trading system 30 minutes before market close, now you’re getting somewhere.