4 reasons every organization needs a data governance strategy
Organizations across all industries should have a data governance strategy in place. Why is this so important — and what goes into a good plan? Everybody knows data is essential in business today, but handling it effectively and ethically is no easy task. That's where data governance comes in.
In 2017, a survey by First San Francisco Partners indicated that just under 50 percent of survey respondents had a data governance strategy in place. Let's take a look at why that number should be a lot higher.
1. To Organize Talent and Achieve Organization-Wide Insights
The business community finds itself in a bit of an arms race when it comes to data science talent. This is expected to be one of the most in-demand professions in the computer sciences going forward. However, companies need to know what to do with that talent and the insights it can yield.
According to McKinsey, 60 percent of "breakaway" organizations take a cross-functional, interdepartmental approach to data governance and analytical models. This means employing decision-makers who understand the value of data, professionals who can translate from analytical models, experts on user interface and experience, data scientists well-versed in industrial control systems, marketers who can make sense of information gathered, and much more.
A company's data governance model needs to describe how all this information flows into the organization, how it's processed and stored, and who's allowed access to which silos. The goal is to determine how each piece of data is useful for improving the bottom line, streamlining delivery and helping the company do more with less.
In that same McKinsey report, one major financial services company invested in its data analytics strategies over the course of several years. The result was a reduction in fraud carried out via online forms. The analysis wasn't enough, though — it had to close the loop and bring those insights to the front lines. It reinvested in customer service personnel and underwriter training to be more effective at identifying certain types of fraud in the first place.
Data analytics delivered the insights, but culture, data governance and talent management transformed those insights into meaningful and valuable business model improvements.
2. To Better Serve Your Customers
Data is the lifeblood of the modern company. It's critical for every kind of organizational decision-making and goal-setting process, including:
- Improving earnings and profitability
- Boosting customer satisfaction through meaningful insights
- Helping management make timelier and more relevant decisions through performance and market data
The concept of making customer experiences better and more satisfying applies to every kind of company and organization.
State lotteries like Virginia's do a lot of good in their communities. To make that happen, they have to collect and report on a wide variety of data, including total sales, revenue earmarked for public education in each county, operational expenses, earnings for ticket retailers and more.
These days, structured and unstructured data can come from a huge variety of sources, including internal processes, vendors and partners, customer interactions, social media, website analytics and marketing campaign results. Organizations of all kinds need a data governance policy that acknowledges each source and prioritizes it based on how well it supports the mission of balancing earnings potential with the customer experience.
In a manufacturing setting, collecting data on machine service intervals and real-time performance makes it easier to engage in preventive maintenance, avoid breakdowns and continue serving customers without interruption. Companies that can't meet demand tend to find their customers are only too willing to take their business elsewhere. In the state lottery example, collecting county-by-county data on revenue, payouts and public works projects helps the customer base feel more confident they're giving back to the community.
3. To Ensure Compliance With Relevant Regulations
These are dangerous times to be playing fast and loose with sensitive data. From Facebook's spectacles before Congress to Wells Fargo impersonating its own customers to Equifax hemorrhaging personal information on half of all Americans, the consequences for not taking data governance seriously are high indeed.
Under the EU's General Data Protection Regulation (GDPR), fines for carelessness or misuse of personally identifying data can run as high as 4 percent of the company's global revenue. This sounds like an onerous set of restrictions on the business community — until you realize that the American data privacy model makes life more difficult for companies that rely on information to function.
America is a patchwork of voluntary industry standards and federal and state laws. The federal government oversees rules like CCPA (Cable Communication Policy Act), GLBA (Gramm-Leach-Billey Act) and HIPAA (Health Insurance Portability and Accountability Act). Different agencies manage each of these sets of rules.
Given the lack of federal-level consistency, even many U.S. states are writing their own data protection rules — including requirements for disclosing breaches. Naturally, there's no consensus among state governments on what constitutes a breach.
Companies should seek the most stringent data privacy models and craft their own governance rules around them. Remember that the GDPR applies to businesses in other territories that work with EU citizens. It's not a stretch to say this should be the rule, not the exception, as data-driven companies seek wider audiences and a stake in the global marketplace.
4. To Determine What's Helpful and What's Not
Cloud storage is common and cheap enough that some technologists have begun referring to it as a commodity. Smartphones and solid state drives are packing more storage space than ever before, too. However, the availability and relatively low cost of storage space and services mean a lot of companies are becoming data hoarders.
This isn't necessarily a bad thing where companies are concerned. The more data you have, the more context you've got and the more meaningful your conclusions during analysis. Hoarding data that's not useful anymore isn't doing you any favors, though. A good governance policy can help you clear away the clutter, find what you're looking for more quickly and, potentially, save money in the process.
Like that shoebox full of broken odds and ends you're keeping around "just in case," our organizational data can get out of hand just as easily. That's probably why a company as big as Walmart keeps only four weeks' of transaction data around at any given time to fuel its merchandising analysis efforts. Your own data governance strategy needs to guide employees about:
- Best practices for data accumulation based on age, source and usefulness
- Intervals at which obsolete or trivial data is scheduled for deletion
- Procedures for checking incoming data against existing databases to avoid duplicate entries
Of course, there may be times where you want your employees to archive rather than outright delete some of the data they're tasked with handling — and your governance strategy should describe that as well.
When you take these points into account, it should be clear that having a deliberate and detailed data governance strategy in place is essential for your organization. Hopefully, you now have a better idea of what such a plan looks like in practice.