Five Whys as an Approach to Data Governance

Five Whys as an Approach to Data Governance

StockSnap_TJL5H02YIP_WhySign_BPWe were recently reintroduced to the "5 Whys" concept, which we first learned of many years ago as part of Lean Six Sigma training. When it's appropriate, it can be an elegant and useful tool to identify the root causes of problems, and to suggest directions and approaches to address those problems. In our work with clients, root cause analysis using 5 Whys can get a little circular, or it can widen rather than narrowing a focus. But in this post we thought we'd try it at a more generic level.  We commonly hear from clients, and potential clients, some fairly common refrains: we aren't data driven enough, we need better analytics maturity, we don't have a culture of data, etc. For the most part we hear these as variations on a theme, which might also be expressed as that they are not extracting strategic value from their data.

1. Why can't data be used more strategically? Well, what does data look like when used strategically? For other assets there are generally standard practices for acquiring, preserving, growing, and extracting value from them; some of those practices are required by law or regulation, some are specific to an industry, still others have been developed through experience and experimentation. Do those practices exist at your organization when it comes to data? 

2. If data is not valued to the same extent, or with the same rigor, as other assets, why might that be? Perhaps we think of assets in an accounting framework.  It might be difficult to conceptualize of data as an asset. Financial assets get turned into goods or services, physical assets such as buildings get turned into offices, meeting rooms, shop floors, etc., where we can see work performed, human capital assets make products, propose ideas, and so on. What kind of work do we see our data performing? What value is visible in it?

Let's unpack this idea in a little more detail. Assuming that you're not just looking to sell data (in which case we expect you're clearly valuing it), how does data result in a product, or a good or service, or some kind of measurable work or labor? Commonly, organizations are looking to find insight from their data and apply it, either by improving something they're already doing, or by doing something new, or by discontinuing a practice or line of business, or what have you.

3. Why can't they find insight in their data in order to be more profitable, or operate more efficiently, or generate greater effectiveness? One reason might be that there's no one at the organization who can analyze data, or who can put it into a form suitable for analysis. Another reason might be that leaders or decision makers aren't equipped to understand or act on insights generated from data. A third reason might be that the data is just too messy to lend itself to analysis, at least in the timeframe that would benefit the organization. 

In our experience, these first two explanations, while not vanishingly rare, are far less common than they used to be. Are data teams as skilled as they'd like to be? Do they have the freedom and the tools to dig as deep into data as they think would be best? Are leaders as data-savvy as they'd like to be? Do managers at all levels understand how integral data is to the organization? 

4. Messy data, however, seems more of an issue than ever. And we use messy as a deliberately vague descriptor. Why is data messy? 
  • Data silos occur for all sorts of reasons, and they are indicative of all kinds of issues, but we're more interested in this post in their effect, which is generally the exclusion of data that could generate or contribute to actionable insight.

  • Data sprawl is not unrelated to silos, or course, and in some cases is a direct consequence, and while one of its effects can be that distant data doesn't get into the analysis data set.  It seems equally common to manifest as disruptively structured data, by which we mean the more data systems you have in which you collect and store data, the harder it is to wrangle data into one location for study. 

  • Poor data quality can be caused by many factors--you don't need silos and sprawl to get poor data quality, although they're more than happy to contribute--but in our experience their causes are cultural and procedural more than they are technical: data entry and capture practices vary over time and across units; there is no strong methodology for verifying data integrity and for identifying and resolving data inconsistencies; and, probably most often, data quality suffers from the general bloat and benign neglect that comes from not making data management a priority.

Messy data, or a messy data stack, seems to be one pretty compelling root cause for why organizations struggle to establish a culture of data, or to use data regularly and consistently to support decisions. There's probably room for discussion or even disagreement whether your messy state of data makes it hard to value as an asset, or whether failing to value data as an asset is keenly contributory to its messiness.

5.  Most organizations have invested a lot of time and money in data tools and applications, so why haven't these investments made data less messy? We've seen--and continue to see--a lot of technology solutions proposed to address this issue, which in fairness is probably better understood as a loose agglomeration of uneven standards, inconsistent practices, brittle legacy architectures, and the substitution of tribal knowledge for structured information. 

Does that new HR system solve the problem that HR defines employee one way and finance defines employee another, and that the prevailing understanding across the company might be something else entirely? Does the new data warehouse span and select from all your sprawling data sources? Maybe you've got a new enterprise BI tool and have committed to a data mesh - is every business unit delivering data sets and products to the same standard? And where does that standard come from, anyway?

We're not opposed to technology, but as we've written in this space before, technology tends to recapitulate and amplify culture. Ultimately, our position is that the way you make your data less messy is to govern it, and to manage it in accordance with clearly articulate and highly publicized data governance principles.

It's an oversimplification, to be sure, but one way to take data governance seriously is to say it's what you do and how you act when you recognize that data is an asset, and when you treat it like one. 

Working backwards through our whys, if you govern data such that it isn't messy, then it's easier to derive insight from data. Useful data is easier to access, it's easier to work with it, it's easier to understand, it's easier to build a course of action around. Once you're in the habit of gaining insight from data, then it's easier for everyone to see how those insights affect your organization. That is to say, the value of managed data and data analysis becomes more apparent to more people. Finally, then, once data is used to solve problems or improve operations, it's likely that everyone will want to use it! That doesn't mean you'll have a culture of data overnight, but it might mean that aspects become more visible and more common: data training opportunities might open up, the hiring process might begin to include screening for data fluency, organizational strategy might start to explicitly incorporate data. Who knows? Maybe you'll even create a full-fledged data strategy!

At IData, our products and services focus on data governance, data intelligence, data catalog, and data integration. They're designed to help organize messy data, to establish guidelines for its management throughout its lifecycle, and to engage people in understanding, data stewarding, and explaining data as it finally takes its rightful place among your organization's critical assets.  Hope you enjoyed this blog post.

IData has a solution, the Data Cookbook, that can aid the employees and the organization in its data governance and data intelligence efforts. 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.

 Contact Us

(Image Credit: StockSnap_TJL5H02YIP_WhySign_BP #1235)


Aaron Walker
About the Author

Aaron joined IData in 2014 after over 20 years in higher education, including more than 15 years providing analytics and decision support services. Aaron’s role at IData includes establishing data governance, training data stewards, and improving business intelligence solutions.

Subscribe to Email Updates

Recent Posts