The case for data governance
Think you don’t need data governance? Think again
Eniko Ban | April 8, 2015
Data is the lifeblood of most organizations, a hidden line-item in the company assets. Companies that use data-driven decision-making are five percent more productive and six percent more profitable than their competitors. Yet, without effective governance, data is nothing more than numbers.
Though governance has a negative connotation to some, data governance—the discipline of properly managing and safeguarding an organization’s data assets using mostly policies—is essential. It changes the way that employees create, manage, use, and retire their data assets so they can gain meaningful insights using data analytics. This change often requires the corporate culture to undergo a serious organizational makeover. Successful data governance programs continuously improve data quality, resulting in analytics that can identify actionable insights across the organization and differentiate industry leaders from the competition.
Data governance programs are not (yet) officially valued on company financial statements; however, if harvested, can become a true asset. Data governance provides actionable insights and has tangible and measurable benefits.
The organizational benefits of data governance
Centralized/de-centralized decision-making over the management of all (corporate) data. Organizations generate immense amount of passive and active data. Identifying what data is useful and what is not requires an understanding of what that data represents. Capturing, ingesting, or storing data unnecessarily wastes IT resources, and not harvesting from existing active data wastes business assets. As long as organizations understand the data landscape, who oversees the data is less important than how it’s used. Focus your energies on inventorying all data assets, understanding their use and value, and using data as a financial asset that contributes to the bottom line.
The knowledge base required to profit from the data that is readily available. Prioritize acting on insights over looking for answers, or risk losing to competitors that find ways to generate income from data. Chief data officers—those in charge of data governance—are tasked with monetization of data assets, and companies with a good understanding of how they can monetize their own or someone else’s data can potentially see a significant revenue impact. Look around and make sure that you are not sitting on a pot of gold and determine whether you, too, can repackage information assets you already have.
Consider also this fact: The biggest database any company has isn’t storing transaction, process, or any other data; it is open data available on the net that anyone is free to use, reuse, or redistribute. If your organization is equipped with data scientists who can mine and analyze these large data sets, you may find clever ways to generate additional income for your organization.
Ownership of deliverables tied to the performance of this secret asset. If, for example, a company writes off $500,000 in inventory discrepancies with reason codes of unknown or other—instead of doing root cause analysis and identifying that 42% is due to receiving errors, 16% to material defects, and the remaining 12% to shipping mistakes—it must have the luxury of not holding its management responsible for wasted resources. Incentivizing the receiving manager to not have more inventory write-offs than x% of all received goods may result in closer attention to detail and quality.
Answers to any and all questions around any data. Without an infrastructure in place to answer simple questions like Why do I see my receivables in GBP instead of USD or Why can’t I trace my shipment of xyz medicine back to the container that transported it from the US to India? people can waste hours of productivity chasing answers. Data governance helps establish the lineage of data, providing answers or direction to these types of questions.
Renewed corporate culture. Done right, data governance enables employees to use data to gain insights and make better decisions. The saying “the whole is greater than the sum of its parts” holds true here: Empowering the workforce and harvesting the creativity of everyone is far more promising than relying on a select few.
Making the case for data governance
Executives love seeing quantified costs and benefits, such as the amount of manufacturing waste, the sales growth opportunity by adding a new state, the potential damage caused by a data breach, or the liability amount for inappropriate disclosure of Protected Health Information.
Here are three ways to get that information:
- Survey your organization. Find out if anyone is hurting because of bad data or find themselves in a bind due to another data-related issue. Use open-ended questions, where applicable, and follow up with individuals separately if you need clarification or additional information to put together your use case. Remember, the goal is to quantify the price of bad data, or any other impact caused by issues around data. For example:
- During the last five days/ten days/one month, how much of your time was spent chasing after information rather than mining the data and analyzing it to make useful insights?
- Have you been affected by bad data quality? If so, what data and in what fashion?
- If you came across inaccurate information in the xyz system, would you know who to contact to have the issue fixed, and what would the process look like?
- X-ray your area of operations and make some calculations. This may seem difficult to do if you are not privy to certain details of the departmental budget or are too busy performing your day-to-day duties to allocate time to collect this data, but it’s worth a try. It may be necessary to make some assumptions or estimate some of the components, but as long as you clearly document how you got the numbers, it will provide exactly what you need.
Say you know it took two people two weeks to perform a data cleanup effort of 100 pieces of master data (such as employee records, patient listings, or office locations) at a rate of $40 per hour. You can then calculate the project’s entire cost to your company (2 x 80 hours x $40 = $6,400, multiplied by the total number of master data, perhaps considering gaining efficiency during the process and accounting for project management considerations, such as upcoming vacations).
- Rely on the data quality issue log. And if it doesn’t exist, start one! Data governance helps surface all data-related matters that slow down processes or compromise quick decision-making. It also establishes policies that prevent those impediments going forward. To compile these in an issue log, you need to capture a few key pieces of information, including: the issue, the consequences, those involved, the impact (in USD), the solution, the assignee, and approvers (if necessary).
The issue log tells a lot about a company. If one exists, you can tell how well the process works by looking at the number of issues resolved and the amount of time and effort it took to solve a data problem. You can also identify the subject matter experts for future situations. If it’s incomplete or outdated, proper attention hasn’t been given to this important tool. Companies without a log likely do not view data as a strategic asset—yet could probably benefit from it most.
Eniko Ban is no longer with Slalom.