Information – or the possession of information – is seen as a kind of power, and some people are loath to give it up. While some individuals hoard information for venal or nefarious purposes, in most cases this behavior is often tied to legitimate concerns about security, misinterpretation, regulatory compliance, and other, occasionally conflicting, data practices.
Whatever the reason, this behavior often results in data being managed in silos. Information then operating in something like a fiefdom, where a person or group controls information in the manner of a feudal lord, and in order to gain access to information, a seeker must come bearing gifts, or perform heroic endeavors. We've seen the forms in use at colleges and universities, and it's not that much of an exaggeration to say filling them out completely is like completing a heroic quest!
When data is hard to access, people come up with workarounds. So data ends up being collected in duplicate, and almost certainly according to varying principles; information obtained or managed in one area will not be linked consistently or productively to information held by another area, and in fact may well be oppositional. (Feud and feudal, after all, have a common etymology!)
Disagreements can arise: who “owns” the data; whose version of the data is correct; whose interpretation of the data is best; what is the proper way to talk about and understand data; and so on. In many cases, these disagreements can be handled through reasonable conversation, although this is time that could be better spent discussing what the data means and what, if anything, should be done about it. (We will return to this topic in a future post.) Sometimes, sadly, the disagreement escalates to conflict, and some data conflicts jeopardize an entire organization's ability to use information effectively.
When data is siloed – and one could argue that siloed data includes the proliferation of shadow databases and analytics data sets, including ones sanctioned by the organization – we might say that organizationally there is no accepted, animating vision for managing and utilizing data.
Plenty of organizations are struggling to develop a vision for data, but most are at least paying lip service to the idea of using data to improve operations and make better decisions. Still, it’s one thing to have a vision, and another thing to have a strategy. Data-related issues at the strategic level include a lack of knowledge and/or excessive certitude, an overemphasis on planning and a lack of emphasis on doing (or the reverse), too narrow a focus on the cost or instrumental value associated with data management, and an often unintentional separation of business strategy and data strategy.
A formal data strategy would be a written document or statement that articulates how at a strategic level data will support/reinforce the organization’s business strategy. (Note that we use business as a catchall to describe decisions and actions that stem from mission—the success of this kind of business strategy is not measured solely financially.) Must an organization have a written data strategy in order to make better use of its data? No, but an organization that doesn’t strategically prioritize data isn’t likely to be one that governs it well, either.
Hallmarks of immature data governance include missing or incoherent policies on the handling of data (includes access, use, shareability, even analysis), poor enforcement of policies that do exist (or rigorous enforcement of policies that don’t exist!), ill-defined or inconsistently defined data terminology, satellite data systems (information sprawl?), no data stewards (or data stewards without authority/mandate/portfolio), a fixation on control, a mentality that IT is responsible for data, and, of course, silos.
An organization with an incomplete vision for and a nonexistent or partial strategy about data is in most cases not going to have the executive sponsorship and top-to-bottom advocacy required for a successful data governance framework.
Would we love to see all organizations hire a Chief Data Officer, stand up a data governance council, staff it with data stewards, data trustees, and other caretakers and stakeholders from across the organization, and begin to systematically address the data management issues facing them? You bet we would. And as more and more organizations of all shapes and sizes evolve to value data as an enterprise asset, as they recognize the critical need for usable information in crafting and meeting strategic objectives, as they tire of the costs, direct and indirect, of data dueling, of time-consuming and labor-intensive data wrangling, we expect most institutions to set up a formal data governance framework (also referred to as data intelligence) with specific responsibilities and performance expectations.
But we also recognize that for many, perhaps most, organizations, the time is not right for this endeavor. The work of data governance should solve problems and improve the abilities of employees to do their jobs. If instead it generates meetings without measurable action, or additional policies without meaningful enforcement, or extra data artifacts without real utility, then it's more trouble than it's worth.
Some organizations are ready for a Data Governance Council, with all the bureaucracy and oversight that entails. Others are more suited to take on a Data Governance roadmap that identifies specific areas needing improvement - data quality, say, or more agile reporting, or even simply a modernized data RACI matrix - and appoints people to enact those improvements. Others may only be able to commit to deploying tools such as our Data Cookbook as they seek to form a common data vocabulary and account for their inventory of data assets (although the Data Cookbook can be indispensable for institutions at all levels of data governance maturity!).
But you've got to do something. If you're reading this, then you're already interested in using data to solve problems and get answers, and you know that consistent data management is critical in that effort. Wherever you are in the organization, there are steps you can take to participate in, support, or lay the groundwork for robust data governance. That discussion you're having about terminology? Write it down. That data request you're filling out? Find out what problems it's going to address, and see if your colleagues are also taking steps to address those issues. That data you don't think exists? Sure, see if there's a product out there that will help you collect it, but also ask around to see if anyone else in your organization needs the same information and perhaps has already started to gather it. Maybe you can't knock down a silo, but you can climb down from yours and knock on the door of another one; somewhere in your organization, probably many places, you'll find others who are anxious to get out of their silo, who also know that sharing their "grain" makes everyone healthier.
IData has a solution, the Data Cookbook, that can aid the employees and the organization in its data governance, data intelligence, reporting, data-driven decision making and data quality initiatives. IData also has experts that can assist with data governance, reporting, integration and other technology services on an as needed basis. Feel free to contact us and let us know how we can assist.
(Image Credit: StockSnap_3HGXPSXH2B_whatdoingwithdata_BP #1006)