CIO Dashboard

Executive dashboards have become very popular mechanisms to present executives with at-a-glance information on the organisation’s strategic KPIs. The CIO is IT’s executive sponsor for all the other executives’ dashboards. So then, what should the CIO have on his dashboard?

A dashboard is a visual display of the most important information needed to achieve one’s objectives, consolidated and arranged on a single screen so that the information can be monitored at a glance. Without having to page through screens full of data, an executive can instantaneously gauge the performance of the company or his business function. This is especially useful if the appropriate targets, thresholds and pre-warning indicators have been set and are displayed to provide context to the measured KPIs.

In this post I discuss three of the aspects that the CIO should be tracking, from the point of view of being the executive ultimately responsible for data management and information dissemination to the whole organisation.

Proliferation vs Penetration

Proliferation measures indicate to the CIO how many dashboards are out there in the organisation, and how many measures are reported on each. But that in itself is not such useful information… A really interested CIO wants to know whether they are actually using the extensive (and expensive) per-field drill-down capabilities his team developed with great effort. Detailed dashboard usage statistics would give the CIO a much better idea how far the dashboards have penetrated into the organisation and how they have affected the management styles of the executives and other key decision makers.

The same applies to the penetration and usage of analytical results and production reports into the organisation. A good quantitative measure is the penetration of BI-delivered reports vs the number of “spreadmarts” that are manually populated and used as reporting mechanisms. A tuned-in CIO needs to know whether his BI-CC is gaining an upper hand in the anti-spreadsheet war. Man hours saved vs man hours wasted is another way to express that outcome.

Users vs Usage

It is good for a CIO to know the degree to which the organisation is “informed”. The number of licensed business intelligence tool users may sound like a useful measure, but it is actually their style of usage that provides the really useful information. Do they simply run canned pre-developed reports or do they regularly do their own reports and analyses in power-user style? And how many of each do they actually use, and how frequently?

Another extremely valuable usage-related statistic is the amount of data accessed. It gives an indication of the ROI of the underlying data warehouse. The data warehouse may occupy terabyte upon terabyte of storage space, but if only 5% of the historic data is used for long-term trend analysis, i.e. only that related to the clients’ buying patterns, does it make sense to keep all of it on expensive super-fast on-line media? Now factor in and report the DBA’s stress-levels to make all this work, together with the rapidly-shrinking durations of the batch windows, then the CIO has real cross-functional information at his fingertips.

Data Quality Metrics

As the ultimate custodian of the organisation’s informational resource, the CIO is squarely responsible for data quality. There are a myriad of data quality measures that can be reported.

Examples of inherent data quality measures include:

• Completeness or coverage (meaning that there are required values for all required fields)
• Validity (the data values conform to domain and business rules)
• Accuracy to source (the reported data agree with the data values in the original source systems)
• Precision (that the values are correct to the correct degree of granularity).

Examples of pragmatic data quality measures include:

• Accessibility (the ability for users to access the data when required)
• Timeliness (the duration between the real world event and when the information about the event becomes available)
• Contextual clarity (that the data presentation enables an exact understanding of its meaning).

There are literally pages and pages of data quality metrics that can be measured and reported. The big thing to watch out for is that the CIO’s data quality dashboard does not start looking like a disco-fied encyclopedia with screens and screens of details which do not highlight any important or relevant aspects. One of the key characteristics of a dashboard is that it must highlight exceptions instead of merely reporting details on the norm.

Summary

The CFO is the custodian of the financial resources of the organisation. He has a dashboard reflecting the changes in financial status at his fingertips, with cash flow and debtor days updated as they change. Similarly, the sales executive has stock movements, sales campaign effectiveness and related sales figures flashing on his activity monitoring dashboard as they come in.

So, it follows then that our CIO, as the custodian of the information resource of the organisation, should also measure and monitor what he is trying to manage. He, likewise, should have meaningful graphs and dials indicating the penetration, usage and quality of one of the most undervalued resources of the organisation.