Why Data Governance before BI?

In the world of BI, there are key challenges that affect all projects; (1) data governance, (2) partner support and (3) user adoption. Looking at each in turn, this is the first discussion paper on data governance.

Active Data Governance

First and foremost, let’s be practical, data governance is not passive; it is not a review at the end of an IT deployment project – only to see what’s gone wrong! Data Governance is an active system of checks and balances and processes that ensure quality information is delivered.

Governance needs to be considered from solution architecting through implementation to the results, reports and dashboards. It should be an integral part of the data management from input, handling, manipulation and utilization and there should be a clear chain of command that takes responsibility for the technology element of governance.

But is that what usually happens? You and I both know that’s not the case!

And the end result of system deployment; users need to be able to “trust the numbers” and be empowered to access and use them securely. And that won’t be the case if governance has not been applied at every step of the way.

Who, what, where and when

Data governance provides the business practices and IT processes that define and manage the way valuable data is formally protected and used through the organisation, so decisions need to be made prior to any BI project commencing. Detailed thought is required at every point through which the data passes as to;

• Who has the right to access

• Who has the right to take action

• Who has responsibility and accountability and finally

• Who has authority to execute on the data.

The impact on the business of inadequate data governance can result in data breaches, loss of revenue and untold damage to the business. The single most prevalent point of attack is personal credentials which underpin this need for responsibility tracking through governance with policies that dictate how to enter data and who has user rights.

It is worth considering hiring a 3rd party to help build a set of policies that govern and dictate who does what, where and when.

Said Ed Paice of Harness IT (3) who acts as Data Governance and Project Assurance for large ERP conversions; “…I would argue that a BI project will start to ask fundamental questions about a company’s data sources, longevity and accuracy and it’s better to ensure you have those building blocks established before launching into designing of cubes or reporting metrics. We all know that a lot of BI projects come on the back of replacement of an organisation’s operational or ERP system and therein lies the challenge as reports will likely require year against year comparison, however you might not be taking that historical data into the new ERP system. So how to populate your BI cubes?”

This is where a good data warehouse can be the lifeline to legacy data.

After the fact is too late

As a business intelligence project unfolds, it often becomes clear there was not sufficient forethought and planning around the strategy, design and governance of the solution. One significant KPI is the level of customer satisfaction. Recent results emerged from the MSDynamicsWorld (1) survey that showed only 37 % of Dynamics customers who responded expressed that their needs had been met by their BI project.

So that’s 63% who felt it missed the mark!

Mostly the issues at fault were quoted to be complexity and empowerment – demonstrating a lack of data governance from the outset.

To endorse the importance, the Business Intelligence Guide (4) states that “If the data management capability is not in place before rollout of any BI applications, then there will be no-one on the ground to execute data policies, with the end result of governance in a vacuum – no one enabling these decisions.”

Further discussions with Ed Paice of Harness IT (3) endorsed this view; “I had to interject when one project lead was having sleepless nights worrying about how the new operational systems would support the BI metrics which had been mandated from the board. Something had to give, so I put the team together and getting executive buy-in to the problem added a touch of realism and they were able to agree what could realistically be delivered that was still valuable.”

Moral of this story is don’t get hung up on what you can’t accurately report on and spend time struggling to make the metrics fit but work collaboratively to deliver what is possible.

It’s not all about the Tech

So now we have established the importance of data governance and the need to empower users and remove complexity, here are a few pointers that are less about comparing technology, more about the softer issues that should influence your choices

1.Ensure data access rights are made for business reasons and not governed by technology limitations. A steer here is to look for good self-service analytics dashboards from your vendors.

2. Consider agility and flexibility in your system of data governance, as well as a sound methodology. You need to strike a balance between accurate recording and the needs of the business to change over time.

3. Think collaboration when in design mode across all departments. To achieve a great result, you will need buy-in and cooperation.

4. Think of all of your data sources right at the start of the project, looking at both structured and unstructured data and consider a data warehouse as part of the solution.

5. Some things go without saying, but let’s say it anyway; security, platform, mobile, monitoring and management capabilities all need to be tip top.

6. Visual is important; some people think in Microsoft Excel or Microsoft PowerBI or even Microsoft, so ensure you can deliver everything from a pie, bar and line charts, to heat and tree maps, geographic maps and scatter plots and make sure its intuitive to drill down to the detail.

The scary big name vendors

When considering vendors, there is a scarily long list of providers, including the big names like SAP, IBM, Microsoft, Oracle, MicroStrategy and SAS. They all talk about agility, ease of use and the ability to deploy data governance at every step. However that doesn’t mean that governance is always well executed. This is where your choice of VAR and partner is essential. Further endorsing this point is the MSDynamicsWorld survey (1) which revealed that only a staggering 26% had a data governance strategy in place prior to commencing their BI projects.

Room for improvement

So, although the technology has the tools in place, it appears that there is ample room for improvement in the execution. Technology alone will not achieve nirvana; there needs to be a positive strategy to take data governance seriously and where there’s a will there’s a way!

PrecisionPointWhere data becomes trusted insight

Follow us!

References;

  1. MSDynamics World.com released; Investment in BI and Reporting – by Jason Gumpert, published January 29, 2016 2. Gartner Magic Quadrant for Business Intelligence and Analytics Platforms; published 04 February 2016 | ID:G00275847 Analyst(s): Josh Parenteau, Rita L. Sallam, Cindi Howson, Joao Tapadinhas, Kurt Schlegel, Thomas W. Oestreich, 3. Edward Paice – HarnessIT 4. The Business Intelligence Guide

Contact us