Digging your Way Out of Data Debt

Over the past half-decade, you’ve probably spent north of seven figures on multiple analytics and BI projects, looking to drive better, deeper insights into your business, increase market share, and grow your bottom line. Do you feel like you’ve gotten a decent return on your investments so far? Are you getting the answers you need from all those analytics capabilities you’ve deployed? Chances are, you answered no to one or both of those questions. If you did, you’re in data debt.

Data debt is the practice—if you can call it that—of implementing short-term, segmented analytics/BI solutions to solve specific, or departmental problems, without integrating them into your overall business and IT strategy. It’s that gap between the cost of implementation and the actual return on investment (ROI). I call it debt because it’s money you’ve invested unwisely, and you’ll have to “pay back” at some point—whether it’s through fixing the problems you’ve created, or with something more ominous: your career.Data debt

I know you have an analytics/BI strategy. But do you abide by it all—or even most—of the time? You’re not alone if you answer no.

Many companies get caught up in the “me too” syndrome and give their squeakiest wheels the most grease. Don’t do that. All you’re doing is proliferating analytics/BI anarchy and acquiring more data debt. Instead, take a step back and look at what the problems might really be, how you can fix them to avoid incurring more data debt, and what the long-term consequences of your decisions could be.

Here’s how you can begin to close the data debt gap and start earning a better ROI on your analytics/BI investments:

  1. Figure out where your problems lie. Assess the current state of your analytics/BI capabilities. Answer the following questions: Do you need all of them? Can systems be combined? Which systems are being used, and which have been supplanted by unintegrated databases or spreadsheets? Which systems are providing you with the most reliable “version of the truth” about the questions your analysts are asking?
  2. Get reacquainted with your analytics/BI strategy and fix it if necessary. Assess your current strategy objectively with nothing held sacred; anything should be up for debate. Then, tweak it or reformulate it entirely to meet the needs of your business users. It’s critical to get the input of business users on the strategy. As the saying goes, analytics/BI should be business-led and technology-fed. Also, look toward the future. Technology changes rapidly. Your strategy should be flexible enough to accommodate those changes.
  3. Be a good, but stern, cop. Stay true to your analytics/BI strategy. Learn to say no—i.e., stop building customized solutions if they don’t fit the strategy. This doesn’t mean that you can’t look for quick wins—those are critical to proving success and maintaining momentum. What it does mean is that if a project doesn’t fit your strategy—or if it’s the answer to an isolated problem—do some serious thinking about whether to implement it. Typically, you’ll be able to tweak a system that’s already in place to fit your needs. If not, do the math and make sure that you’ll get the answers you want for the price you have to pay.
  4. Rely on an old friend: data stewardship. When no one owns data, it tends toward anarchy. As technology has changed and data has become more unstructured, data stewardship and governance have taken a back seat to the “just get it loaded and read” mentality. This creates a mess that lowers ROI. To fix the data mess, appoint data owners to be responsible for the quality of their data, and implement a data strategy and governance program to put data-quality standards in place across technologies. That way, no matter the technology, you can ensure that your data is clean, and your answers are clear.

If you’re in data debt—and most organizations are, to some extent—it won’t be easy to climb out, but you really have no choice. If your analytics/BI systems become bottomless money-pits that consume your organization’s resources, time, and intellectual capital, the payback is usually pretty high. You can avoid data debt—or greatly reduce it—by figuring out what your problems are, being a good steward of your resources, and building a sound, flexible strategy to help you avoid future mistakes.

I’d love to hear what you think. You can comment or DM me on Twitter, message me on Linkedin, or email me at Anuraag.Jain@thinkbiganalytics.com. Also, if you like what you read, click the buttons below to share it.

Leave a Reply