A few months ago we discussed something we called the "reluctant data steward," and although we liked the way the phrase sounded, we're not sure it was the best description for what we observed. There may well be people pressed into service as data stewards who don't actually want to steward data, and that could be an interesting topic for further investigation. But what we were trying to describe strikes us more as a reluctance to change habits or tools or traditional ways of doing things, even when those ways are inefficient or cumbersome.
Most data stewardship is invisible. By that we mean that when data stewardship is performed adequately, it ensures a number of things. When those things are ensured, data functions pretty smoothly, if not perfectly. And when data functions pretty smoothly, not too many people complain.
What does data stewardship ensure? We'd include things like users getting access to the data they need, but not getting unauthorized access to other data. We'd include things like when people don't fully understand the data, they can get easy clarification from data stewards. We'd also include things like people basically trust the data, in part because they know that data stewards maintain rigorous enough data quality standards.
It can be comfortable toiling in relative obscurity. If people aren't that aware of the data you collect, then requests to see it, and to "do stuff" with it are less frequent. If you don't mind patiently explaining for the sixth time this quarter to the same person how a given metric is calculated, and/or why it's important, then there's no need to write that explanation down, post it online somewhere, and direct questioners to that page. If you and your teammates are willing to do that one weird trick to scrub data for weekly TPS reports, or to perform manual edits to the machine-generated dashboards that can't account for the idiosyncrasies only your unit understands, then why take the extra steps involved in curating data sets and certifying the products generated from them?
Those tasks all look like extra work, and while they might save time or make us more productive down the road, we've got enough work to do right now. And what if investing in a new process or application doesn't result in those time savings or productivity gains? We've certainly all seen that with other corporate initiatives, right?
This fear, often justified, of ultimately unproductive extra work, is where we try to situate our pragmatic approach to data governance. There are data-related tasks being performed every day. For instance, organizations are constantly developing and publishing new dashboards, or setting up an integration between systems. There's already a ton of work involved in doing that! Adding a minor requirement like, each dashboard must have a descriptive name and purpose statement in our report / dashboard catalog, means very little additional work. Many of our clients have a business intelligence initiative around setting up tiers of data products--gold layer, silver layer, bronze layer, what have you. Qualifying for any of these layers involves some kind of validation, which is a manifestation of data governance. Validating a dashboard, or proofing an integration, means agreeing on terminology and calculation methods! Don't let that work simply live in the heads of data stewards and analysts. Users should know what that validation involves, and they should be able to find that out from the BI tool or the report / dashboard catalog.
Data stewardship is a full-time responsibility, but it's not a full-time job. Although we do know some people with job titles like Data Governance Manager, we have yet to meet anyone whose title is, for example, Data Steward for Finance. In fact many of the data stewards we work with don't even have data stewardship called out by name in their job descriptions, although to be fair many key data management duties are mentioned. So that's another key aspect of actually performing many data stewardship tasks: the time you spend doing them is time you don't spend managing payroll, or analyzing cash flow, or scheduling and attending those meetings, etc. Now, it's not as if data stewardship is wholly unrelated to a person's "real" job. To pick an obvious and at this point somewhat dated example, it's hard to call that client or potential client if you didn't do a good job capturing and recording their phone number.
Excelling at data stewardship probably helps data stewards and their teams function a little better, but ultimately that is probably not the stuff of performance evaluations or end-of-year bonuses. So it's probably a rational choice to focus energy and attention in ways that respond to incentives. We can't tell you how to write your job descriptions, or evaluate employee performance, or structure your raise/bonus pool. But we believe there are many ways to incentivize superior and somewhat more visible data stewardship.
Of course, when data stewardship is visible, a different business corollary often applies: no good deed goes unpunished. Some data stewards are firefighters, helping put out data fires when they inevitably arise. Do that often enough, or publicly enough, and the fires start looking for you, so to speak. Other data stewards practice what we have seen referred to as heroic data stewardship, going above and beyond the call. Sometimes this looks like a procedures manual to help new employees learn tools and systems, other times it's working with analysts and developers to certify and annotate reports and publications, and other times it might be something like taking on the task of cleaning up a data set. While that work is laudable, it's often unsustainable, since eventually your "real job" rears its ugly head and makes its demands, and it's also not scalable: just because one or two data stewards are willing and able to do this special work doesn't mean all of them are.
We happen to believe that our Data Cookbook solution raises the visibility of background data management and stewardship activities, and that it also eases the burdens around fighting data fires or performing heroic data stewardship. The Data Cookbook offers multiple paths to document curated data sets and the business intelligence outputs based on them, and it allows for both formal certification and informal inventorying. Key data terminology and concepts are elaborated in the business glossary, and used to further contextualize and explain data in its various manifestations across your organization. Pragmatic data governance addresses data issues and exigencies when they arise, and it doesn't consider the issue closed until there is written agreement on terminology, methodology, responsibility / ownership, etc. The Data Cookbook notifies data stakeholders when there is governance need that affects them, and it records requests, suggestions, approvals, signoffs, and similar interactions.
A central data intelligence repository is going to be a new tool, whether it's the Data Cookbook, some other cataloging product, or something you cobble together across existing applications. Formal records of data governance decisions, large and small, are likely to represent a new business process for most organizations. There's going to be some reluctance to use new tools and adopt new practices, no matter what they are or what the situation is. But you don't have to write every business glossary entry on day one! Similarly, you don't have to certify every item in your report / dashboard catalog, you don't have to curate every data set in your lake house. And even those are your goals, you're not doing all of those as the first steps in your data governance journey! (Unless you want a short and potentially quite painful journey to exactly where you already are, about which experience the less said, the better.)
Data stewards are essential for data governance to succeed. Slightly increasing the visibility of data stewardship and incrementally incentivizing the improved performance of that stewardship are ways to better engage data stewards, to draw out their expertise, to make data a more valuable asset. We'd love to talk with you, or share with you our webinars, or otherwise demonstrate how our products and services engage your stakeholders and increase the value of your data assets. In the meantime: be pragmatic about data governance. And: take care of your data stewards, who are taking care of your data!
*We apologize for the bad pun in the title, but we don't regret it enough to change it!
We hope you found this blog post useful. Also check our our data governance spotlight resources including one on data stewardship located at https://www.datacookbook.com/spotlights. IData has a solution, the Data Cookbook, that can aid the employees and the organization in its data governance, data intelligence, data stewardship 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.
Photo Credit: StockSnap_CNXUTKY06W_sailboat_datasstewardship_BP #1308
