I read with great interest this article (Business Should Work with IT to get Self-service BI Governance Right) since it relates to a hot interaction that we are researching right now: that being the overlap and intersection of information government, always and every time a business concern, and how analytics are governed (in a BI/analytics environment).
As the article shows there is great confusion in the market about, ‘Self-service BI governance’. The article then mentions ‘BI data governance’. I don’t even know what that is or if it’s the same as self-service BI governance. I can’t tell if this is about data in the warehouse; how analytics being developed should be governed, or the reporting report/analytic and who uses it. I believe that no such thing exists anyway. Let me tell you why.
About two years as self-service data discovery got going, I was inundated by vendors from the traditional analytical side of the house wanting to brief me and tell me if their support for ‘governance’. Virtually and universally their capability revolved around the notion that someone in the business sets the (governance) polices concerning data used by the analytic app (represented by the vendor), and also the use of the resulting analytic. As such, the work related to data or analytic governance that takes place in such analytic apps is nothing more than executing and respecting a rule defined elsewhere and outside the scope of the analytic app.
But did those analytic apps have a rules engine? Did they have a workflow to support the setting of the rule? Did they have analytics related to the performance of the polices and supporting workflows? No, none did and frankly they don’t need to. It turns out that MDM and more recently information stewardship solutions emerged that were designed to help here. Such work is not specific to self-service data discovery or analytics or BI. It is the work of stewardship (policy enforcement) and governance (policy setting).
So at best self-service data discovery solutions need to focus on what they do best (discover data) and just respect and follow rules maintained elsewhere. At the data level (inbound to the analytic app) the rules set are called information governance and at the analytic level (outbound from the analytic app) the rules set are called analytic governance.
The former is well known and documented and still hard to do. The latter is relatively new, needs to follow the same model as does information, and remains equally harder to do. Just ask yourself this questions:
Which organization or body sets information governance policy? Answer: Information governance board
Which organization or body sets analytic governance policy? Answer: Analytic center of excellence or BICC
The next question becomes:
Why two bodies doing the exact same thing for all data and all analytics for all policy types (such as access, privacy, security, quality, ethics etc?) Why not have one body responsible for data and analytics governance that sets policy for all data types? Basically the same business people need to be involved!
The article then goes on about ‘self-service analytics adoption’ which is a very different topic. This has little to do with governance per se. To be fair the article has the right premise, and I quote:
“There’s no question we’re living in a self-service analytics world today, but that can’t mean anything goes.”
But maybe the article should have been a lot shorter:
All analytic apps, of all kinds, are just apps. They should respect the rules and policies and targets set by the act/work of information governance.
Information governance includes any data that needs government including structured data in operations systems, copies of same in warehouses and lake, unstructured data in lakes or documents, content, images, records as well as analytics and algorithms. Information governance is not just about security and access and compliance and regulation. It now includes business value.
So bottom line: there is no such distinct thing as self-service BI governance nor data discovery that includes governance. All such applications need to respect information governance (that embeds governance of analytics) and just follow along. We don’t need to over complicate this stuff.
I will add a final point: How to sustain information governance around data-lake versus a data warehouse versus an operational business application is quite different. That is a very different topic as to who should set policy, and who should enforce policy.
(About the author: Andrew White is research vice president and distinguished analyst at Gartner Group. this post originally appeared on his Gartner blog, which can be viewed here)
Register or login for access to this item and much more
All Information Management content is archived after seven days.
Community members receive:
- All recent and archived articles
- Conference offers and updates
- A full menu of enewsletter options
- Web seminars, white papers, ebooks
Already have an account? Log In
Don't have an account? Register for Free Unlimited Access