TRILLIUM SOFTWARE INSIGHTS

Dirty (Data Quality) Jobs

By David Binkley, Senior Technical Account Manager, Harte-Hanks Trillium Software

Early one morning some time ago, I was walking through the airport in San Francisco and I happened to recognize the guy walking towards me as Mike Rowe, host of The Discovery Channel’s popular DIRTY JOBS series. 

For those of you who aren’t familiar with the program, DIRTY JOBS profiles the unsung American laborers who make their living in the most unthinkable, dangerous, or often disgusting, yet vital, ways – with Mike usually getting extremely dirty in the course of performing this week’s selected job.

Well, that chance encounter got me thinking why it’s so easy for data quality people like me to relate to the jobs shown on that show.  Yes, sometimes, cleansing data is like scraping ostrich goo out of a tub or shoveling manure. 

Yet, I also asked myself why it would have been so difficult for me to explain to Mike that the type of cleansing I perform and the “dirt” that I’m cleaning is the cyber equivalent of what he’s doing?  Could it be that we data quality people just have a problem with visibility in our profession and its often hard for us to explain exactly what we do and get people (particularly management) to understand the true, vital value of our dirty jobs?

I think we need to all start thinking more about visibility – getting more like Mike Rowe.
So, as a starting point, I’m offering up the following list:

• Let’s find better ways to keep score – you can’t tell if you’re winning the game unless you’ve got a way to keep score – data quality measurement is key.

• Once you have the score, make sure that management knows it and knows what it means.  I know we all don’t have a weekly TV show but we all need to find a way to keep in front of management on a regular basis and show the dirt (and the value).

• Emphasize the good as much as the bad – as nothing turns management off more than someone who only brings bad news. 

• Find a way to be more like Mike to show that, while it may be difficult, you are fully capable and willing to get the job done.  That may be tough to accomplish if you happen to be up to your knees in slop, but you’ll get past that point and show why its so important to the business.

• Work at finding more ways to deliver value to your organization – don’t be shy.  Be like Mike, show up everywhere and get your hands dirty and be loud about it.

Now I have to ask, what have you done today to improve the visibility and value of data quality in your organization?

P.S. Mike, I’m forever grateful that you weren’t doing a show on the restrooms at San Francisco Airport that morning!

Enterprise-wide Data Governance: Walk the Walk

 By Jim Orr, Director, Enterprise Data Strategy, Harte-Hanks Trillium Software

One of the biggest obstacles I see with information quality is that all too often it is confined to a single project. While more and more organizations are beginning to go enterprise-wide with their quality initiatives, too few are actually doing it.  They’re talking the talk, but not walking the walk.

At the same time, companies realize they must branch out across the organization if they are to resolve upstream data issues as well as identify and address other opportunities and risk associated with their data.

We could point to a number of contributing factors why businesses remain stranded at the project layer but the primary one points to the absence of data governance.

Going upstream to mediate data quality issues requires us to work “outside the box” and that means coordinating cross-functional people, processes, technologies, lines of business, and budgets. The only way to accomplish this is on the back of a formal data governance program.

Data governance enables full life cycle and enterprise information quality which in turn optimizes data and business results.

How many of you are embarking on a formal data governance program?  OK, how many even know where to start?  Let me know.

 

Is Data a common language?

By Richard Wilson, Vice President, Product Strategy, Harte-Hanks Trillium Software

Banks operate as large and complex organizations, and often utilize multiple languages to communicate internally and externally.  Cross-enterprise communication, requiring the use of many different languages, is not just limited to geographies or regions though.  Even people who speak exactly the same verbal language around the water cooler tend to develop unrecognizable ‘local dialects’ when discussing their own specific business areas. 

As a result, the all-too-human tendency to develop distinct languages, clustered around a variety of interests, creates multiple ‘information silos’ within companies.

A huge business challenge, therefore, grows from these separate pools of information.  Somewhere – at a point far from the information source – risk management teams must calculate a provable, defensible computation of the exposures underwritten by the bank, and the ability of the bank to remain liquid in times of turmoil. 

Credit risk and treasury functions share the charter to identify the risk exposure from a wide array of products and services delivered across a global landscape.  And they must perform the signatory function of attesting to the accuracy of the information that is being provided to the market, and more importantly, to regulators.

The common ‘grammar’ used for communicating the information streaming from these multiple operational sources is data. But – just like the unique dialects that humans create to address specific issues – data is subject to all sorts of ‘unique qualifiers’ that are appropriate for their specific application.  Those unique qualifiers are often misinterpreted and (many times) simply missed. 

A real ‘Tower of Babel’ problem develops that can threaten the business.

Real problems arise when misinterpreted facts and un-factored exposures are introduced into risk weighting and capital requirement calculations.  A serious danger lies in not being able to explain positions due to poor understanding of the supporting data.  Lack of transparency constitutes a direct challenge to confidence in the data.

The Treasury and Risk functions of the bank must take the lead in developing governance processes to engage the proper resources, and ensure that the fiduciary requirements of transparency, accuracy and provable correctness are met.  Key members of each risk subject area must be empowered with insight and track the quality of the data being reported by them to the risk function so that verified, provably correct data is reported and consistently monitored for defects.

What Basel II or data compliance pain point stories do you have to offer?

 

“Profanity in your Data Quality”

By David Binkley, Senior Technical Account Manager, Harte-Hanks Trillium Software

Many years ago when I was working with a large company, we encountered a situation where our call centers were reporting problems with profanity being entered into several of the name and address fields.  Since the “profane” data usually appeared almost always in name and address, its presence was found to be slightly detrimental to customer satisfaction when the bills were mailed the following month.

As is typical with large companies, several committees were formed to study the problem and formulate a “plan of action” to correct the situation.

The meetings went on for several weeks (seemed like months to me) and we finally concluded that there were several key issues that needed to be resolved:

-  For the most part, what was happening was the customer service representatives were venting their frustration with our customers, and they had nowhere to put “warnings” for other people handling the account.
-  The data quality programs in existence at that time did not have any capability to identify and remove profanity or otherwise deemed offensive comments.
-  Not only was profanity present in the data, but because the large company was international in scope, the profanity happened to be multilingual as well.

Quite obviously, what was originally observed to be just a data quality issue ended up being a problem that had impact well beyond all expectations.  Fixing those issues required changes to existing applications, screens, standards (defining what profanity is), methods and procedures. 

This problem expanded well beyond a data quality tools solution, as the issues quite clearly involved aspects of data governance, standardization, customer satisfaction and security.  It involved various people across departments and divisions, and had the immediate urgency of resolving direct impacts on customer service and trust.  Fortunately we already had a data governance team in place that was able to facilitate all of the needed cross departmental issues, interfaces and development efforts.

I could end this cautionary data quality and governance tale right now, but I need to add a little sub-story about the team that defined the words as profanity.  If you’ve ever been involved in creating standards, you’ll no doubt relate to this.

The first opportunity was finding a team that could properly work the issue – as if the Supreme Court can’t define profanity, how could we do it?  After an extensive search, our team ended up being some ex military personnel and a middle aged female staff member with a rather astounding grasp of the subject matter.

The team promptly went to work, but unfortunately, they didn’t close the door on the conference room where they were working.  And, they got a little bit too “into” the assignment, a little bit too loud “discussing” the terms, and laughter reigned supreme day after day.  Ultimately, one of the other co-workers near the conference room who didn’t know the context of the “profane” discussion, ended up complaining to her manager.

Of course the complaint worked its way up and down through the layers of management, ending up back with me as the analyst assigned to the project.  Yes, in a somewhat circuitous fashion, we found the data quality problem also impacted human resources, employee relations, and internal productivity!

Sometimes you get noticed even when you aren’t trying…