Should organizations separate analytics governance and data governance?


"Your scientists were so preoccupied with whether or not they could that they didn't stop to think if they should."

—Jeff Goldblum as Dr. Ian Malcolm, Jurassic Park

There's no shortage of content on the web about data governance. A quick Google search of the term reveals more than 450,000 results. On this blog alone you'll find more than 100 posts on this important topic.

But should we think of analytics governance as separate from data governance?

In this post, I'll address this question.

Alison Bolen makes the case for a separate, analytics-based form of governance here:

How can you be sure that all [organizational data]efforts are accurate, aligned and worthwhile? How do you know the analytics processes are being deployed, repurposed and secured as efficiently as possible?

In the same piece, Bolen continues:

Without analytics governance, you are opening your organization up to a laundry list of risks, ranging from competitive sabotage of your models to the common problem of data models that are built but never used.

You'll get no argument from me here, but I still wonder: Do organizations really need to formally separate their governance efforts into two buckets: data and analytics? How many enterprises actually do this? And, not incidentally, how many do this well? I'd love to see some data on these questions but I suspect that there's not much out there.

Feasible vs. desirable

should we separate governance for data and analytics?Even if separation of analytics duties is feasible, is it desirable? (Cue Jurassic Park quote above.) I have my doubts. Consider that many if not most data governance projects fail. Inasmuch as analytics stem from data, can an organization really – and effectively – practice analytics governance when it doesn't govern the use of its own data very well? Put differently, Company X's data governance efforts are slipshod. What are the odds that it handles analytics in a responsible, ethical and legal manner? Not great.

Beyond that, I can think of logistical concerns around formal separation of governance responsibilities. "No, that's not a data governance issue. That's an analytics governance issue." "We're only responsible for the data, not the analytics." And how does the role of a potential chief data officer or chief analytics officer muddy the waters? I can quickly see things becoming overly bureaucratic, inhibiting the very speed at which organizations need to operate.

Simon says: Embrace a more expanded definition of data governance.

Perhaps it's best to think of a more expanded definition of data governance – one that includes powerful analytics. Rather than trying to tackle a new – but admittedly related – type of governance, wouldn't your organization do better to shore up its existing data governance practices?


What say you?

Download a paper: Ten Mistakes to Avoid When Launching Your Data Governance Program


About Author

Phil Simon

Author, Speaker, and Professor

Phil Simon is a keynote speaker and recognized technology expert. He is the award-winning author of eight management books, most recently Analytics: The Agile Way. His ninth will be Slack For Dummies (April, 2020, Wiley) He consults organizations on matters related to strategy, data, analytics, and technology. His contributions have appeared in The Harvard Business Review, CNN, Wired, The New York Times, and many other sites. He teaches information systems and analytics at Arizona State University's W. P. Carey School of Business.

Related Posts

1 Comment

  1. Data Governance should extend to all data that are created, read, updated and/or deleted by any enterprise process, regardless of whether being an operational or an analytical process.

    Since not all data are equally important, measures of data governance should be prioritized based on legal and regulatory requirements as well as on their expected impact on revenue, cost, risks and opportunities. Also, it needs to be taken into consideration that analytics read, interpret and potentially transform data that have been created, updated and/or deleted by operational processes in the first place, i.e. analytics governance will be only meaningful, if appropriate measures of governance have already been applied upstream to the processing of the related operational data.

Leave A Reply

Back to Top