Conversations from the Mirror42 Team

Although this interesting external post (”10 best ways to lie with metrics) focusses on business metrics, I think it also applies to IT management. Probably most relevant to Information Technology is number 3 — “use many metrics”.

Why the need for integrated IT management dashboards? Take as an example the application life cycle and its management. Software applications are normally introduced in organizations by a business case. After approval of the business case, a project is started, either to make or buy, and to design and purchase infrastructure. If a make-decision has been approved, a development project is started. A developed application is tested, for bugs and performance. And after delivery it is handed over to the IT operations organization to run it as a service to the business. After a while, business will have new requirements (demand) for the application and meanwhile changes to the infrastructure must be made to ensure quality of service. New requirements are implemented and a new version of the software is release and deployed. Etc.

Where is all the management information on this application life cycle residing? Indeed, in a range of IT management applications, such as a project portfolio management tool (e.g. HP/Mercury PPM - formerly Kintana, or CA Clarity - formerly Niku), an IT service management tool (e.g. BMC Remedy, HP ServiceCenter/ServiceDesk, or CA Servicedesk), an application test tool and a whole range of operational management tools for applications and infrastructure.

If, management is now interested in a single consolidated view of the application life cycle, it will need to consolidate reports from all the underlying sources. In pratice this is time consuming, error prone, and in most cases the lag in information delivery may hurt the management decision process. Therefore, amongst others, the need for integrated IT management dashboards.

Information technology enablers tend to gather a lot of data from the IT environment, mostly operational statistics. Is that because they can, or because they actually need it? More importantly, we tend to push this data out to the business. Is this because they can, or because the business actually needs it?

I have seen companies forwarding the temperature statistiscs of server rooms to the business. Why would they do this? Do they actually expect that the business is critically interested in these temperatures? On internal operational level tracking these kinds of measurements might be very valid, especially when availability of service is dependend on it. But, is a business manager interested…

I believe one of the reasons is that IT enablers tend to push operational management data to the business, is that they do not know what the business is actually interested in (and, in all honesty, often the business does not know either). And if it is known, in most cases they cannot measure that kind of information easily. Tracking the availability of an application is just much harder than tracking room temperature.

So instead of fixing this problem, IT enablers tend to forward all the data they have, and they call this “openness” or “transparency”. However, all this data does not make sense for the business. IT enablers need to understand what kind of information the business needs before “communicating”. Or even better: I would suggest that the business understands what kind of information it requires when purchasing IT services, and make the delivery of the information part of the purchasing agreement.

I believe, from a strategic perspective, IT stakeholders are interested in trends and aggregation of information, and even better in a combination of both. Trending allows focussing on improvement initiatives and tracking quality issues. Aggregation of information allows viewing the performance information from “top to bottom” with the ability to drill-down into problem areas. To help with strategic decision-making, key performance indicators (KPIs) are laid down to assess the present state of the business and to prescribe a course of action.

In my experience, the information needs from a strategic perspective differs from operational information needs. On operational level the information needs circle around lists of “things-to-do”, statusses of routers, servers or services, or re-active email alerts. The operation is basically focused on keeping the IT systems running, and its information needs are focused on doing that as efficiently and effectively as possible.

In between sits a somewhat ‘gray’ area: the tactical level. This level has its focus on short term, smaller objectives. The tactical level probably needs a combination of tracking trends in metrics (for short-term measurement of objectives), and access to operational information for decision-making.

In most IT organizations a number of reporting tools are used to serve each of these layers. However, at Mirror42 we notice that the information needs on strategic level are in most companies fulfilled by ‘cut & paste’ activities in Excel or Powerpoint. That does seem very inefficient, but foremost it leaves room for error (and in some case, gaming). So the question is why strategic management is investing in (in most cases not-up-to-date) reports, and is most likely making decisions based on error prone information.