There are a few reasons why participation in data governance efforts is greeted with only moderate enthusiasm by business leadership, LOB’s, and business departments.
Articulating the potential impact of governed data on business value is not an easy task, and the ROI is hard to quantify.
There is an ever-present resistance to change and a shortage of resources for engagement, since employees in business functions are already overloaded with other tasks. But of course, ‘too busy’ has always been just a polite substitute to ‘not a priority.’
But there is one other reason. It may sound a bit harsh, but it leans closer to reality.
Data governance is, naturally, about data. But business has limited interest in data, if any at all.
This may seem like a reckless statement, but it comes from my 10+ years of experience on both sides of the fence: consulting and working with the businesses on the client-side.
Two remarks are needed at this point though.
First, there is a category of business specialists that possesses an immense interest in data: my former comrades-in-arms—functional business analysts—are always in search for a fresh subset of data to feed their analytical models.
Bad news—they show absolutely no interest in governance.
My second remark relates to data quality— that has been declared the key business deliverable in most data governance business cases I have seen so far.
There is no doubt that businesses are interested in the quality of data because it has a direct impact on the quality of managerial decisions. But in everyday life, data quality is clearly considered a subject residing in its entirety within the IT domain.
In the same way the conductor of a philharmonic orchestra is not concerned with the technical mastery of her oboe player, she just takes it for granted.
So, if my hypothesis has an element of truth in it, and business folk are not really interested in data, then we need to readjust the focus of a data governance program to gain their buy-in.
How exactly? Let me share a story first.
Call Setup Rate Falling
The project at this mobile telecom company was clearly heading nowhere. It all started with an unlucky title themed around ‘metadata management’ and after the initial outburst of enthusiasm, everything came to a halt. IT was juggling with ‘technical metadata’ and ‘business metadata’ clubs, while marketing (nominated to represent the business) showed no signs of interest.
It was at this point that we met with their Chief Technology Officer (CTO) for the first time, on a mission to gain a wider support from the C-Level. About five minutes into the discussion, he interrupted us with a fairly resolute ‘I need this.’
His primary responsibility, he explained, was maintaining the quality of their wireless network at the highest level, as it had an immediate effect on subscribers’ experience. To achieve that goal, his team had to monitor a set of around 300 metrics, based on data generated by network equipment on a regular basis.
Keeping each metric in this set accurate and actual was a critical issue, but almost impossible to resolve, as the only governance tool in their possession was a mammoth MS Word file. As a result, quite often a miscalculated or misinterpreted metric would lead to a wrong technical decision followed by failure of a certain network node. Which, of course, did not make their customer service managers any happier and their subscription rates any lower.
Three months after the first meeting with the CTO, we delivered results of this initially unplanned project track: full inventory of reports and metrics completed, business glossary filled with content, and data sources defined and linked to content. And, of course, stewards assigned and functioning, and (simple) governance processes running smoothly.
‘Good example highlighting the importance of C-Level sponsorship,’ one may say.
No doubt—but not only that.
The most important lesson we learned from this experience: business involvement is almost guaranteed where there is a highly visible (almost palpable) connection between the data, the context in which the data is presented, the outcome of a data-driven decision and, finally, the business performance of a certain function (or company).
This is exactly how, in my opinion, we should readjust the focus of data governance to make it more appealing to business—by replacing the vague promise of ‘better data’ with a clear path to improved business performance.
Business Glossary or Glossary for Business?
I must apologize here for the ‘data-driven decision’ cliché used in one of the sentences above. Business decisions are indeed driven by information, not data. And in most cases by a very specific type of information: metric.
Therefore, the metric (or ‘indicator’), in my opinion, is destined to be used as a key argument in our bid for business engagement in data governance.
Quality of data is just too abstract of a concept and definitely someone else’s problem. Quality of metric, in contrast, can be comfortably grasped and irrefutably related to business performance, as illustrated in the telco story above. And our pitch can get even more convincing when we switch the emphasis from a single metric to a system of metrics.
Now we are talking business.
Let us just clarify the terms first, as ‘system of metrics’ may sound a bit alien to our business colleagues. Within their borders they call it a ‘performance measurement system.’ This is something they had a go at in the Balanced Scorecard era and probably ended up failing. And, of course, performance measurement system is the cornerstone of business performance management. (I probably have to remind here the old mantra ‘you can’t manage what you can’t measure’, but that would be another cliché, would it not?)
But what if this whole idea of attaching ourselves to a clearly business practice of performance measurement takes us a step too far into our neighbors’ territory? Shall we not leave for Cesar what is Cesar’s?
Well, we shall not.
First of all, I cannot think of a stronger driver for a business-focused data governance initiative than the contribution to business performance.
And then, we are doing it already, are we not? Collecting definitions, terms, and (probably) metrics to charge our Business Glossary?
Let us just change our positioning and repurpose the Business Glossary as a Glossary for Business. Not a place to store some business content intended to guide us in our quest for better data, but as a business tool playing the three-part role of a guarantor, mediator, and navigator for one of the most critical information assets owned by business.
To perform this role effectively, the Glossary will need to comply, at minimum, to the following standards: be centered around the metric (performance measure), be structured with a business user in mind, and be complete.
Building such a business tool is a very challenging mission indeed. It is challenging, but not impossible if approached with a good plan at hand.
