Data Governance Matters

A ‘Franchise’ Approach to Data Governance

Recently I had the pleasure of working simultaneously with two separate groups within the same financial services company. Members of the first group were focusing their Data Governance program on a specific need: ensuring contractual compliance for purchased data sets. The leader of this group is an amazingly focused woman. She taught me a thing or two about the art of looking away from details that were not on her critical path – at least until she had that critical path well defined. Because her success would depend primarily on business process rather than technology, we spent very little time during our roadmapping activities talking to IT folks.

During one such meeting, however, another manager mentioned that he was developing an enterprise data dictionary, and that his group would like to leverage what the first group was doing. He didn’t think IT should “own” his data governance effort, and would the first group agree to be their business “owners” and to take on his effort’s program administrative and facilitation activities.

I knew what she was going to say, of course. “Absolutely not.” She had her hands full, and she needed to maintain focus. Besides, she pointed out, they were different efforts, with different prime stakeholder groups, different skill sets required for their respective data stewards, and different timelines. What did he think the benefits would be of combining efforts, anyway?

His answer was quick. Data Governance can be hard, he said, and it helps if everyone’s using the same terms, concepts, accountabilities descriptions, approach to rules, and scorecards. He thought that a federated model for Data Governance was right for their organization.

So we began a fascinating discussion about the differences between centralized, decentralized, and federated models. What he was actually looking for, I told him, was not a traditional federated model. He was looking to establish a “franchise,” where a new installation of Data Governance with a unique focus could leverage the materials and knowledge of the original installation. He was looking for a tried-and-true template for a program that had already been customized to his company’s culture and environment and was ready to be further customized to his focus and goals.

Bingo. Everyone could buy into that approach.

It turns out my original client and I were way ahead of schedule with our work, so I was able to give him a half day to talk about opening his franchise. We went through the first effort’s “Program-at-a-Glance” materials, discovering how they could be used for his effort. As we worked our way through the briefing page, we slowly substituted his bullet points for her bullet points, until every single box on the page had changed.

But the overall organization remained, along with the look and feel, section headings, and the flow of the materials. An executive reading about her program and his program would experience the familiarity, comfort, and consistency that customers look for when they patronize a franchise. And that executive comfort level and clear understanding of their programs would go a long way toward getting both groups the support they need.

Oh, and along the way, we discovered that both efforts would indeed be receiving oversight from the same executive committee. So, strictly speaking, I guess they do have a federated model.

But the concept of a franchise approach to Data Governance made both teams’ roadmaps much clearer and immediately raised their confidence levels. It made Data Governance seem more real, and less academic. It made their goals appear more achievable, faster.

And aren’t those some of reasons entrepreneurs choose to go with a franchise?

Dictatorships and Data Governance

Recently a guy who was studying different approaches to Data Governance and Stewardship slid a piece of paper across the table from me. It was a picture of the components of the DGI Data Governance Framework.

Here, he said, pointing to the section that calls out accountabilities and controls. This is what I found attractive, because it’s so actionable. But what else should I be noticing? What sets this framework apart?

I circled the Decision Rights component and the Data Stakeholders component. These, I said. The framework is postulated on the idea that before data-related decisions are made, those who have a stake in them will agree on HOW to make those decisions.

He looked puzzled, so I continued. It’s not like you have a democracy, I said. Some decisions – especially ones with legal and compliance ramifications – will need to be made by top leadership, with no arguments from others. But other types of decisions – like where to pull data from in a particular business process, or the best way to structure data so as to cause the fewest problems for downstream users of that information – well, those are “gray area” decisions with lots of stakeholders. I’d hate to be the Data Dictator who tried to dictate a solution to all of them! I concluded.

Why? He asked. I’d think some people would like being a dictator – making everyone else respect their authority. I know she – he said, nodding toward a woman we’d been talking with – has been joking that she’d like to be our ‘Data Czar.’ We’ve been joking about whether to get her a tiara or a whip!

I had to laugh. Both pictures were pretty funny. But then we got serious again. I already had that talk with her, I said. I asked if she liked working here and how long she was hoping to stay.

He cocked an eyebrow. Why? he asked.

Think about it, I suggested. What usually happens to dictators?

The Summer of Love and the Hippie Central Data Repository

Guest Host: Darwin

I like many things. A glass of good wine after a bad day, an intellectual challenge, honest music and anything paisley or tie-dye… but things I love make up a much shorter list. I love my Old Friends. It’s my contention that until a relationship ages, you cannot count the person on the other end of it as more than a close acquaintance.

Gwen Thomas, the DGI President, is an Old Friend . She’s younger than myself, but we still go back aways. We don’t get to spend ‘face time‘ often anymore, but when we can, we make the most of it.

I spent quality time with Gwen the end of last June in San Francisco, after the Data Governance Conference. There was absolutely no way I was going to that wonderful city without a pilgrimage to my old stomping grounds, so Gwen, myself, and the rest of the DGI crew spent a day in Haight Ashbury. We joined a walking tour groupt, I traded memories with the entertaining, informative and attractive tour guide, and then…

I saw it.

The Message Pole. It was more than just a landmark for myself and many others there during the summer of love in ‘67, it was the center of our existence. It was simply a telephone pole, at the time emblazoned with personal notes, flyers and posters. We would all check the pole for our messages and community announcements at least daily. Today, the Pole stands empty except for its glaze of rusting staples, one of Haight Ashburies most treasured landmarks, a monument to the culture of the Hippie.

The Pole revolutionized communication in the Hippie Community. Prior to the Pole, messages between Hippies travelled many paths, were handed from friend to friend, often ending up at the wrong destination or at the right destination polluted with errors. After the Pole, Hippies posted and received information from that one central repository. The results were predictable; less delay in delivery and less introduction of user error which resulted in much cleaner information.

I was in my mid teens that fall in 1967. I was stapling a ‘crash pad available’ notice to the Pole when my epiphany struck. I was floored by the sheer magnitude of the Message Pole concept; that the path information takes can be as important as the information itself.

For the first time, in my life I saw the inherent beauty in design. That moment undoubtably influenced the path I’ve taken; I like to believe the sum of who I am, what I’ve accomplished both professionally and personally, has been shaped in no small part by the existence of the Message Pole.

And it’s still there…

Kumbaya,

Darwin

The Danger Within

A reporter asked me this last week whether I thought private data was safe within corporate firewalls. He wasn’t asking if I thought the firewalls would keep out hackers. Rather, he was asking whether I thought it was safe from internal threats – people who would copy private data.

I told him the truth: that often, it’s safe only because people are basically good. That for most of the company’s I’ve seen, once someone is given credentials and permission to work with certain types of data, there would be very little to stop them from misappropriating information. We talked a bit about the possibility of corporate espionage – how easy it would be, and how great it is that company’s aren’t out there creating data breaches just to hurt their competitors.

And yes, we did talk about new technologies that detect improper patterns of behavior by credentialed users. We talked a bit about catching bad guys – the stuff that security freaks revel in. But when we were done, what I remembered most about our conversation is that people really are generally good, trying to do the right thing. And that’s what I like best about helping companies with their Data Governance programs; for the most part, its about helping good people do good things better.

‘Squiggly Line’ Data Flows

Recently I was asked to review an executive-level data flow diagram. It was orderly, simple. Data in, data moves a few times, data out.

I asked about one of the arrows: Is this really a single set of data? Now 1 thick arrow was replaced by several thinner arrows, since information was coming in from the web, an application within the corporate firewall, and several FTP feeds.

OK, I said, pointing to another arrow showing data moving from one database to another. Is the data simply moved from one to the next, with no transformation?

There was laughter: no such luck. There are a lot of things going on: translation, cleansing, transformation, filtering. Are any of these activities providing value other than making the data usable by the next database? Not really, no.

So, on top of each of the large, neat arrows on the diagram I drew messy, squiggly lines. These squiggles represent waste – non-value-add activity – I said. Is this an accurate picture?

When they agreed, I put down the marker. THIS, I said is what Data Governance can help with, reducing this waste. How? We begin by uncovering the data-related reasons for the squiggles along the way. Chances are, we can change one of these fat arrows early in the data flow so it will continue to accomplish its primary goal but now does it in a way that won’t require so many squiggly lines downstream.

Won’t that make that first process more difficult or expensive? someone asked. It might, I said. Then someone else said that if there’s an ROI, we should do it, even if those first guys don’t like the idea. That’s when the real conversation about Data Governance began.

The Elevator Speech and Motown Records

Guest Host: Darwin

It may surprise people who know me to discover I like Motown music. My tastes generally run a little more to the esoteric side; the psychedelic innovators I grew up jamming with and listening to, obscure jazz, classical music, acoustic folk, but there is something infectious about the sing-a-long finger-snapping Motown sound.

The reason I mention this is because I just had a revelation while in the process of helping a colleague ‘hone’ a new elevator speech. I’ve always been considered a good communicator, the go-to guy for these sort of little tasks. It was after 5 o’clock, we had adjourned to one of my favorite places, the not-too-noisy one with the comfortable chairs, great food and classic rock tunes playing in the background. My colleague had just finished reading me his ‘first draft’ which was a 5 minute technical dissertation on his latest project. During that pause after he finished, while I was formulating my reply, I noticed we were both tapping our fingers on the tabletop in time with the background music, that great Smoky Robinson and the Miracles classic, ‘Tracks of My Tears.’

“That song was 3 minutes of music from Motown Records,” I mentioned. “The companies founder and president, Berry Gordy, had a very ‘hands on’ management style. He had many artists under contract, vocalists with tremendous range and technique, and writers of incredible talent. His message to each never varied. ‘Keep every song simple, singable by the average person, and get right to the Hook, the most memorable part of the song.’ Vocalists with over three octave ranges were always complaining his music did not challenge them, yet, he was the most successful independent record label of all time, and an astonishing percentage of his records remain classics today.”

“You need the Motown Sound and a ‘Hook.’ I continued. “When your project is implemented, what will the end benefit be for the Enterprise?” He told me, we had a Hook, and a very short time later we had a 30 second powerful and memorable message.

Our IT department needs a Berry Gordy.

I wonder if he consults…

Kumbaya, Darwin

Does our organization need a Data Governance framework?

All organizations need to be able to make decisions about how to manage data, realize value from it, minimize cost and complexity, manage risk, and ensure compliance with ever-growing legal, regulatory, and other requirements. Management and staff need to make good decisions - decisions that stick. They need to reach consensus on how to “decide how to decide.” They need to create rules, ensure that the rules are being followed, and to deal with noncompliance, ambiguities, and issues.

In short, they need to do more than manage data; they need a governance system that sets the rules of engagement for management activities. Small organizations, or ones with simple data environments, may be able to succeed in these goals through an informal system of governance. They may not even be aware of when they are switching between making management decisions and broader governance decisions. On the other hand, larger organizations, or ones with more complex data or compliance environments, generally find that they need to step back and agree upon a more formal system of governance. The DGI Data Governance Framework is designed to assist these organizations.

Yes, Data Governance Matters

Does Data Governance really matter? someone asked me recently. At first I didn’t answer – I thought it was a joke. When I realized she was serious, I answered carefully. Where you work, I asked, does data matter?

When she nodded yes, I went on. Does the way that information is managed matter? When she nodded again, I asked whether managing that information involved options, and decisions about which option to choose?

Of course, she answered. OK, then, I said: does it matter HOW those decisions are made, and by whom, and when, and following what process, and with whose input and approval?

She paused, then said she’d never really considered it, since she wasn’t the one making most of those decisions. But yes, of course those things matter.

Those things, I said are governance. And when they’re applied to data, they’re Data Governance. So yes, Data Governance matters.

Later, I realized that Data Governance Matters would be a great name for my new blog. So here it is. I’ll be doing most of the posting myself, but I’ve been promised frequent guest posts by Darwin the Data Dude (see Total Disclosure).

Talk to you soon.