Beauty Lies in the Eyes of the Beholder
Nearly 2500 years ago, Plato captured data quality’s most fundamental principle, “Beauty lies in the eyes of the beholder.” And when your organization applies this principle to data quality today, your business users are the only judges who matter. That’s because the only quality that matters is the quality that impacts your business.
Try this for a simple data quality lens.
- Poor data quality is any level of quality that gets in the way of your business success.
- Good enough data quality is the degree of efficacy that helps your business succeed.
- And perfect data quality may not be worth the extra investment when good enough is truly good enough.
Whatever your data quality bar (and remember, in today’s dynamic business environment, that bar can be a moving target), most organizations find it difficult to rise above it.
Is Data Quality Management Software the Solution?
According to a recent IDC report, data quality management products are “used to identify errors or inconsistencies in data, normalize data formats, infer update rules from changes in data, validate against data catalog and schema definitions, and match data entries with known values.” The report goes on to say that such products usually include “purpose-built software that can interpret, deduplicate, and correct data in specific domains such as customers, locations, and products.”1
Today, such tools are found in standalone data quality products or embedded within broader data management suites such as TIBCO Unify.
But data quality management software is only part of the solution. Successful data quality efforts span both business strategy and technology.
Start with Your Digital Transformation Strategy
In a data-driven world, business leaders know whether their company’s competitive success lies with customer engagement, innovation, operational excellence, or some combination. As such your digital transformation strategy should inform your data quality strategy.
For example, if your strategy is to grow your cross customers cross-sell revenues, then the only data quality that matters is whatever level of quality it takes to understand your customers—their preferences, prior purchases, current activity, and anything else you need to intelligently offer them just the right cross-sell product at just the right time. Master Data Management, with its focus on customer and product data, may prove more appropriate than data quality management in this case.
Make it Easier for Your Business Users
In my blog, “To Improve Data Quality, Stop Playing the Data Telephone Game,” I describe the problems caused by our traditional approaches to data management where copying data from one database to the next has become the norm. With so many copies to choose from, it is little wonder that different business users come to meetings with different answers to the same question.
If you use Data Virtualization, you can integrate your data without physically copying it. This will substantially reduce the entropy inherent in typical multiple-copy data warehouse and data lake processes. Beyond fewer copies, data virtualization also provides the opportunity to perform metadata-driven syntactic and semantic transformations and enrichments that standardize datasets across use cases so every business user is on the same page. And when things change, it’s a lot easier to modify centrally managed metadata definitions than it is to modify multiple distributed ETLs and database schemas.
Deliver the Easiest Data Quality Win – Reference Data
Reference data is a special subset of master data used to organize and classify the rest of your data. Whether externally mandated or internally authored, your reference data should be unambiguous and non-negotiable.
Business users appreciate reference data’s clarity of purpose. And because they work with it every day, they are perfect arbitrators of its quality. Reference Data Management improves help business users consistently manage standard classifications and hierarchies across systems and business lines. And with data virtualization as the reference data distribution method, organizations can centralize reference data use to achieve quality goals without extra copies.
Building Your CFO Ready Business Case
As you can see above, addressing data quality can be a multifaceted effort enabled by multiple investments, and thus a potentially complex financial case?
Fortunately, my friend and data quality guru David Loshin provides an excellent framework you can use to build your CFO-ready business case in his book, The Practitioner’s Guide to Data Quality Improvement. Chapter 5, “Developing a Business Case and Data Quality Roadmap,” includes sections spanning:
- Return on Data Quality Investment
- Developing the Business Case
- Finding the Business Impacts
- Researching Costs
- Correlating Impacts and Causes
- Estimating the Value Gap, and more.
The Bottom Line
So, do your business users see your data quality as beautiful? If not, you need to start building a strong business case to address your data quality issues and impact the entire organization. To get started, contact us today.
1 Worldwide Data Integration and Intelligence Software Forecast, 2020–2024 Stewart Bond, IDC #US45441220, June 2020