I unapologetically admit to being a die-hard data geek, which makes my next statements a bit unnerving (downright disturbing, in fact). I am being haunted by statements made during a recent conference keynote – even to the point where they have taken over my thought processes. I suppose that should be a huge complement to the speaker. Who wouldn’t want attendees to be contemplating their words months later? Unfortunately for me, those words shake the foundation that I have had for many years.
At the 2011 Enterprise Data World conference, keynote speaker Rob Karel, from Forrester Research, spoke about aligning data management and business process initiatives. In my notes, I had a few statements highlighted and in red text - mostly to use as ammunition later. Among them were “You fail if you say that you build data governance for clean, trustable data” and “Don’t talk about duplicate data or lack of metadata. The business doesn’t care. Talk about operational efficiencies.” I remember that there were some grumbles around me… “Clean data is not important? Process, not data, is important? Blasphemy!” Okay, I admit it. Those thoughts may have possibly, momentarily run through my head also. How could data not be valuable in and of itself? It wasn’t until I had some time to really process the statements that I began to understand the symbiotic relationship between process and data.
Over the years, organizations have profiled data and presented the results, only to be told that the data were “good enough” to fulfill the business’ needs. We scratched our heads and wondered why the business didn’t want better data. It appears that the existence of “bad” data is not enough to trigger change. The data must possess enough quality issues to severely impact the business process in a way that the business cannot easily find a suitable workaround. This would explain the resistance to MDM in business areas that don’t rely heavily on clean Customer or Product for their success – two areas where huge gains have been seen by implementing MDM.
When I think about solutions that bring real value to the business, they are a combination of information married to business process. Business processes shouldn’t be designed in a vacuum any more than data management solutions. Not having a good understanding of how information influences a process or viewing information as merely a byproduct of a process can be detrimental and not truly leverage the organization’s information assets. But, by the same token, data professionals can lose the process while focusing only on the data. Are organizations structured in such a way as to bridge the two worlds and leverage them most effectively? A quote from my organizational change management course sums it up this way: “Organizations are perfectly designed to get the results that they are getting.”
In larger organizations, I have seen process modelers and data modelers in separate, distinct roles that one dare not cross. But is that truly the best for the organization? In the future, we should embrace roles that cross the process/information divide. Neither process nor data exist on their own, so why split the roles as if process and data were independent? For years, we have been separately developing process models and data models, and then attempting to map the two worlds via matrices and other tools. In essence, we treated them like independent things that have a relationship we can define in four letters – CRUD.
As organizations look at being leaner and more agile, it seems to make sense that these two parts should become one and merge into a more holistic role. Perhaps this applies to bridging across business and I.T. roles as well. I am sure that I will continue to contemplate Rob’s words and search for new insights into the data/process question and the role of the data professional of the future.
NOTE: Thoughts expressed in this blog are the author’s and not her employer’s.