No matter what industry you’re in, your decisions make or break your business—and you can’t make good business decisions without good data.
Unfortunately, good data is hard to find: either you can’t get enough of it, you don’t trust it, or it’s not secure. Moore’s Law for data tells us that the data available to us doubles every 18 months. Without a way to manage it all, we’ll drown in it.
Most organizations grow organically, and so do their tools, processes, and sources of data. In that environment, particularly during periods of rapid growth, it can be hard to keep your data set consistent and well defined.
So companies set up data governance programs to address the issue, spending significant amounts of time and resources to develop processes for cleaning, defining, and structuring data.
But all too often, those projects fall apart—either because they don’t solve specific, measurable business problems, or they don’t engage the business, or they’re too complex.
Never fear: there is a way to regain lost ground on your data management problem.
The beauty of the data management challenge is that it can solve some pretty thorny issues for your leadership chain—particularly around reporting. With good clean data from pragmatic data management, your executive leaders will all be on the same page, no longer inhibited from making decisions because they don’t trust the information. A solid data platform makes it easier to keep the information fresh, accurate, and accessible, lowering labor costs. Best of all, it’s not an expensive challenge to tackle, despite what you may believe.
We suggest taking a concise, thoughtful approach to the problem of data management. Read on for our suggestions.
Focus on a small set of data elements—10 to 15 is best—and confine your efforts to one department or functional area to keep requirements crisp and avoid scope creep. Ideally, the elements you choose will have some impact on the enterprise.
You don’t need to fix the data in all systems at once, as long as users know which systems are aligned with the definitions.
Focus on outcomes: what do the users do with their data? How does their leadership chain use it?
Go after the low-hanging fruit first. Get some wins under your belt and show the value of your work to the business. Celebrate your success—and use it to get other departments excited to participate.
Keep your framework flexible and nimble.
A rose by any other name…
Let’s face it, governance may be the industry-standard term, but it can turn people off. It can potentially convey heavy-handedness and authoritarian structures–it’s just not a word that generates excitement. And your framework needs to be the opposite: light and nimble instead of heavy and burdensome.
So if you’re getting pushback while socializing the name, instead, try calling it data glossary, or data definition, data alignment, or simply data management, as we do here. Whatever you call it, make sure the intent is clear: to make it easier to classify and use data.
Find a sponsor from the business.
If ever a program needed a cheerleader, it’s data management. Ultimately, your sponsor is the one who will bring stakeholders together, ensure that they’re accountable, and drive commitment to the program at a high level.
When you’re looking for a sponsor, look for a business leader. Data management isn’t an IT initiative; it’s a whole-company initiative. While the CIO and technical resources are part of the team, the sponsor ideally resides inside a business team (typically the team that defines the first data elements).
Most importantly, find a sponsor who is willing to make decisions, hold people accountable, and keep commitment strong.
Frame it as a framework, not a project.
Just like our suggestion to call it anything other than governance, we recommend you discuss it as a framework rather than a project. Data management creates standards, provides a forum for making decisions, and makes experts available to assist with decision-making—but it is not a project. It’s a framework.
Here’s the problem with calling it a project: it sets up expectations about end dates. Projects end—frameworks evolve.
Set expectations about an end date.
To be more specific, data management never ends.
Remember, this is a framework, not a project—and frameworks don’t have end dates. If you’ve done it right, you’ve set up a sustainable practice that will carry on into the future. In fact, if you’ve done it right, it’ll be a part of your organization forever.
If the business demands an end date, compromise by setting milestones instead to review changes in compliance laws, available tools and technologies, and business needs.
After all, as the amount of data coming at us increases, so does the technology to manage it, the laws to govern it, and the ways to use it. As long as your business is in business, you’ll be in the business of properly managing data.
If you have a data-management problem, don’t approach it by building complex organizational structures and templates.
Instead, start with pragmatic data management. Be purposeful in setting it up, passionate about execution, and patient in getting results.
Before you know it, you’ll have a data management platform that will grow over time.