The Role of Data Governance in Analytics

Connect with Dwight:

• 
• dwight.brown@turetskyconsulting.com

Connect with David:

• 
• 
• david@turetskyconsulting.com

Resources:

• The Data Governance Institute
• LinkedIn: The Data Governance Institute
• DATAVERSITY
• DAMA International
• HR Data Labs Season 2 Episode 2
• HR Data Labs Season 1 Episode 25

• “You can’t have analytics without some sort of data governance.” – Dwight Brown
• “Data governance is going to be the differentiator between good data and bad data.” -Dwight Brown
• “Having no data is better than having bad data… As the steward, you need to have a system that helps you, a system of checks and balances that helps you, to be able to ensure that you have good data integrity.” – Dwight Brown

David Turetsky: 0:26
Hello, and welcome to the HR Data Labs podcast. I’m your host, David Turetsky. And like always, I try and find fascinating people with fascinating topics in and around the world of HR data and analytics. Today, I have my friend Dwight Brown from Turetsky Consulting. Hello, Dwight.

Dwight Brown: 0:43
Hello, David.

David Turetsky: 0:44
Good. How are you today?

Dwight Brown: 0:45
I’m great. How are you doing?

David Turetsky: 0:47
I’m awesome. For those of you who don’t know, Dwight. Dwight comes to Turetsky consulting with over 22 years of healthcare experience. Dwight, why don’t you tell us about that?

Dwight Brown: 0:57
Yeah, so I started in healthcare, almost straight out of college and worked for the Mayo Clinic in Rochester, Minnesota, most of my time, I spent a little bit of time down in Scottsdale, Arizona and our site down there. But working in various administrative roles there, I worked in HR and finance and a lot of my time was spent just in our quality area. But about half of that time was was spent in the analytics arena. And I had various analytic leadership positions, first with our quality management services area, working with our quality analytics, our quality of care analytics. And then working with our enterprise analytics, doing the leadership of our back end support functions there for a couple of years. And long with that I also worked on the side as a paramedic did that in my spare time doing that for probably about 10 years. So

David Turetsky: 2:02
Dthat’s great.

Dwight Brown: 2:04
Left mayo and and 2019. And when I went into the consulting arena, and had been there ever since.

David Turetsky: 2:12
So you’re doing something in South America, is it something you can tell us about?

Dwight Brown: 2:18
I can tell you, but I’d have to kill you, right?

David Turetsky: 2:19
Oh, no, please DON’T.

Dwight Brown: 2:23
No, you’re right. You’re right. I actually so my, my educational background is is international. I was a Spanish major and my masters in international management. Then when I left Mayo and got into consulting side, I decided to do international consulting. So I threw a dart at a map, knew I wanted to go to South America didn’t know what country but I threw a dart at a map ended up in Cali, Colombia, and worked for an emergency medical services provider, they’re doing business strategy, and then have also been working on some other ventures in the meantime. So I spent probably, I’ve spent probably half of the most last year and a half down living and working in Cali and working back and forth between Cali and and the United States. So it’s been a great, great deal. Great gig.

David Turetsky: 3:20
It sounds like it. One fun thing you may not know about Dwight is that he is an adrenaline junkie. Please tell me you’ve been safe, please.

Dwight Brown: 3:31
Well, you know, part of the adrenaline side is you got to put safety. You got to look at the prioritization of that but no, I I am Yeah, I am safe. I you know, I do things that are that are a little out on the on the edge but

David Turetsky: 3:49
literally out of the edge.

Dwight Brown: 3:50
Yeah, literally. Exactly. Exactly. And sometimes I go right over. Yes.

David Turetsky: 4:00
But but you have a parachute, or your paragliding or something, right. Yeah.

Dwight Brown: 4:05
Yeah. So I I actually when I went down there the first time I was doing some tourist stuff and I I was looking at and I came across paragliding as an activity and I thought, oh, god that looks that looks cool. I had never even believe it or not even though I’m an adrenaline junkie. I’d never been skydiving before. But I looked like it was pretty close to skydiving. And so those of you who don’t know paragliding is essentially jumping off of a mountain with a parachute on your back and flying for long periods of time using the using the updraft from warm air. And so I try to do that at least once every time. I’m down there sometimes I sometimes I’m not able to but

David Turetsky: 4:49
yeah, yeah, Dwight. Please don’t do it anymore. We want you safe and you back here with I wouldn’t say both feet. On the ground, but I’m a hockey player. So I’d say both feet on the ice and you know,

Dwight Brown: 5:03
yeah. Yeah, exactly. Although I don’t know which one’s more dangerous hockey or paragliding, that one could be a close.

David Turetsky: 5:12
it depends on how you play. That is true. So today’s topic, Dwight is talking about our most valuable asset, which is our data. And talking about the role of data governance in analytics. And it’s a real, it’s a passionate topic for me, I really believe in data governance as a beginning strategy for anybody who’s undertaking an analytics path, or they want to go down the analytics path, especially in HR.

Dwight Brown: 5:43
I would agree, I would agree. It’s, you definitely, you can’t have analytics without some sort of data governance.

David Turetsky: 5:54
So that brings up our first topic, Dwight, what is data governance?

Dwight Brown: 6:01
Well, that’s a great question. Excuse me. And I think it depends, it depends who you ask. But you know, to give the to give a formal definition, I, I look at the data governance Institute, which is a data governance standards body sort of one of the gold standards out there and Data Governance Institute describes it as a system of decision rights and accountabilities for information related processes, executed to according to agreed upon models, which can describe who can take what actions with what information, when, and under what circumstances using what methods,

David Turetsky: 6:45
We will put a link, by the way to that, so that when if people want to actually go to that and check out the standard, we can actually put a link on the podcast?

Dwight Brown: 6:55
Yeah, definitely. And I would, when I look at data governance, you know, at a, at a very basic level, data governance is really putting a set of rules around your data, how, what it looks like, how you use it, who can use it, you know, it’s a way of being able to control the data. And it’s a way of being able to have data integrity, essentially. And there, you know, it can be anywhere from very basic to very complex. The, you know, the other more formal piece that goes with that, that, that was also part of the materials that that I saw out there says the most common label to describe accountability and responsibility for data and processes that ensure effective control and use of data assets. Stewardship can be formalized through job titles and job descriptions, or it can be a less formal function driven by people trying to help an organization get data from it value from its data.

David Turetsky: 8:02
Yeah.

Dwight Brown: 8:02
Which brings another another topic altogether, which is data stewardship.

David Turetsky: 8:07
Yeah. Well, I think anybody who’s ever been part of either an HRIT organization, or payroll organization, will tell you that the lines blur pretty dramatically when you’re trying to maintain the data in HR between the data owners, like a manager or an employee or a business owner or business leader, and then the stewards like they’re talking about who are the people who manage the data, manage the flows of the data and manage the structures of the data?

Dwight Brown: 8:38
Yes, definitely. So and depending on the organization, sometimes those roles are are different roles. Sometimes they are blended roles. I’ve seen a lot where the the owners of the data oftentimes end up being data stewards, as well. But it depends on the size, complexity of the organization, all IT systems.

David Turetsky: 8:59
Sure. And I think one of the most important parts of this is that there are going to be times when you’re going to need to work with people, even outside of the HR world, whether it’s in IT, whether it’s in finance, whether it’s you know, because data governance is really a multi disciplinary issue. It’s not strictly HR, and it’s not strictly it either. It has a lot of process implications as well, isn’t it?

Dwight Brown: 9:25
Oh, definitely. Definitely it. And you’re exactly right, it oftentimes that means reaching across boundaries to other areas outside of the HR organization. You know, the the data point that I like to use as an illustrator for this zip code, you you take something as simple as zip code. And number one, there are different ways of denoting zip code. You could go by the five digit or you could do the five digit dash four digit sir and the zip code is fairly unique. versatile, it can be used under an employee, it can be used under a customer. In our cases, it could be used under a patient. But coming to a common definition of how you’re using that zip code and how you’re denoting that in the system, you oftentimes have to go across disciplines. So I would agree with you, I think you’re exactly right on with that.

David Turetsky: 10:27
And we get into other topics, like, you know, zip code is a really good kind of bridge to what I would like to call, you know, the problem that HR gets itself into where, you know, there are issues where you could say, Where is an employee domiciled? You could say, well, you know, what’s the work location? And then what’s their home location? You could talk about the word address, what does the word address mean? Is the work address? Is it the office address? Is it the home address? You know, what does it actually mean? data governance sets it up very carefully, and very cleanly to say, these are the different types of address you need. These are the owners of address, these are the people who can update address. And this is where addresses stored and where can be referenced, right? I mean, right, I’m missing out a lot of detail there. But right. But But the reason why that’s key is because then the general ledger, the HR system, the security system, the workforce planning systems, the time management systems, all of those different systems can utilize that definition of the word address in all those various forms, and then be able to, quote unquote, talk the same language.

Dwight Brown: 11:39
Right? Right. And everybody has sort of a common knowledge on on who, what, how, when side of the equation that goes with that. But having a system of data governance is the only way to only way to get that you get some informal systems, a data governance, where that happens, you know, there’s that, that common knowledge kind of stuff that happens, but, you know, to really do it, right, it’s, you’ve got to have that.

David Turetsky: 12:10
So that brings up the next question Dwight, which is the what’s in it for me, data governance, you know, as a as whether HR user, whether you’re a leader, whether you’re an employee, whether you’re, you know, the CFO or the chro, what’s the what’s in it for me on data governance.

Dwight Brown: 12:27
So, you know, here’s, here’s the way that that I describe it, data governance, is going to be the differentiator between good data and bad data. And that’s the piece that that people don’t understand, but especially us as data geeks, we really get it and we get it from the school of hard knocks, because we’ve we’ve been there we’ve dealt with that data before. And and, you know, I don’t know about you, but it I’ve, I have one specific time that I remember where I was presenting data to one of our executive leadership groups, and somebody brought up something that really shot holes in the data that I was presenting, because we really didn’t have we didn’t have consistency in our data. We didn’t understand the definitions, and you know, the and so with data governance, the what’s in it for me is that, number one, it’s a differentiator between good data and bad data. Number two, the best case of bad data is that your integrity is blown apart. The worst case is your business being blown apart. I mean, data and working in health care. I want I also saw was the data can literally between the life and death that’s in it, unless you unless you have the data governance systems in place. You know, it, it really, it really can be that the other thing you know, memories are long get it right the first time.

David Turetsky: 14:07
Yeah.

Dwight Brown: 14:08
And the what i when i was leading different groups, and we do system rollouts, or report rollouts, or whatever that might be what I always told my group was look, having no better no data is better than having bad data.

David Turetsky: 14:28
Yeah.

Dwight Brown: 14:28
And I think that that’s a piece that people people really forget about sometimes, especially people who are new to analytics, reporting and some of that,

David Turetsky: 14:39
Yeah.

Dwight Brown: 14:39
So the the what’s in it for me is not is basically the fact that you are the steward of putting data out there and putting data insights out there for the organization. And so as the as the steward, you need to have a system that helps you as a system of checks and balances that help You to be able to ensure that you’ve got good data integrity out there for what you’re using and reporting.

David Turetsky: 15:06
Yeah, let me just comment on that last part. It’s that, you know, when when someone’s the steward, they really need to utilize the people around them to be able to help make sure that the data is in the shape necessary for analytics and reporting. You know, we look at stuff so often that we miss a lot of things, because we’ve been looking at it for so long, it’s kind of like when you’re writing the news, when you’re writing a paper, and you keep looking and they keep looking at some of those, hey, you know, you spelled that wrong. And you’ve been looking at it for so long that you just never see it? Well, in the case of the data, you live with this data day in and day out, you don’t see the problems, and they exist in there. So make sure that you have people along the way who are helping you by making sure that they’re looking at as well getting those second set of eyes on it. You know, you talked before about the example, the example I have is I had a conversation, what wasn’t more conversation was more of a yelling at when I the compensation leader brought into a meeting with the CFO, a set of headcount numbers, and the CFO was started yelling at me that that’s not the right numbers that they had the right numbers from the general ledger. And I said, No, no, these are right. I counted them myself. I literally called managers to make sure the count was right. And they said, but but yours is wrong. And I said, No, they’re both right, you’re looking at a different definition of the word headcount than I am. I’m looking at it as butts and seats, you’re looking at it as FTEs. Or you’re looking at it as some type of FTP equivalency. And so you can get into those conversations, and you’re getting to data governance in the conversation. But unfortunately, it starts out as a yell fest for a little bit. And we hate those meetings, and you hate to be called into those meetings. I actually love those meetings, because in the end, it leads to discovery. And it leads to consensus. And I tell people not to fear, you know, you’re talking about an adrenaline junkie. That’s where I’m an adrenaline junkie and a data geek, because I go into those moments, you know, looking for that rush where they say that you’re wrong, and I go, No, you’re right. And they look at me and they go, What the hell are you talking about?

Dwight Brown: 17:19
What are you on?

David Turetsky: 17:21
Exactly yeah, I’m on data. You’re insane. Yeah, but that’s okay. But But at the end of the day, at the end of the meeting, they come out of it with a full appreciation, which is what data governance is trying to uncover from the beginning. It’s trying to give everybody a level playing field and a basis for not only understanding but interpretation of different forms of data, because it’s got this standard that data governance provides, correct?

Dwight Brown: 17:50
Yeah, I would, I would agree entirely. That’s I, and like you I love those conversations. It’s Yes, they can be painful. But I think you hit the nail on the head that you’ve got to be out there talking with people and say that great data governance analysts are some, you know, you’re getting on the phone, calling somebody saying, here’s what I’m thinking about this, here, I’m looking at this specific piece of data, what might I be missing? Or what do I need to know about this? And, you know, it’s, so often what I’ve seen is, is where a data governance analyst might sit down, and they’re just looking at the computer system, they’re just looking at the raw bits and bytes that go with it, but they’re not getting because, like you say, we look at this data day in and day out, we’ve also got a lot of information about the data turned around in our heads that might not necessarily come out and might not be documented or whatever. And until somebody picks up the phone, or shoots an email across and says, tell me more about this. You don’t know or And now, the yell fest meetings that you were talking about, you don’t know.

David Turetsky: 19:05
But not only that, but you can’t have data governance in one person’s mind. It doesn’t work

Dwight Brown: 19:11
Exactly.

David Turetsky: 19:12
A data governance strategy means that people know it, it’s a published document, or a published set of documents or a published set of standards.

Dwight Brown: 19:21
Right. Exactly. Exactly.

David Turetsky: 19:25
So Dwight, that brings us to our third topic, which is, you mean data governance seems overwhelming because there are so many sources of data, where does somebody start?

Dwight Brown: 19:37
You know, it really it really can seem overwhelming and some of that can determine it can be determined by the size of the organization, the number of systems that the organization deals with the number of data points, whatever that whatever that might be. And, and actually what I’ll what I will also preface this with is is the like to joke that data governance is the end of the data world, it can really put a lot of people to sleep, right? And it’s, you know, it takes a special, it takes a special person to love this. And, you know, that in and of itself can make it overwhelming because some people see it as It’s just boring as all get out and, and that’s okay.

David Turetsky: 20:25
Well, it is okay. And it’s not okay. It’s okay. Because there are other people who will take the mantle of the data. Exactly,

Dwight Brown: 20:33
yes

David Turetsky: 20:33
It’s not okay. Right, right, right. It’s not okay. Because if those people then start showing their data to other people, and they get yelled at, they’re gonna wish they didn’t fall asleep during the data governance. So hopefully, you’re paying attention listener.

Dwight Brown: 20:50
It only takes that one time, it really does.

David Turetsky: 20:54
You know? It’s, yeah,

Dwight Brown: 20:56
no, go ahead.

David Turetsky: 20:57
I was gonna say it only takes one time to be in front of that leader who says, I don’t trust you. I don’t trust your data. Right? I don’t trust you, you’re telling me because I know how many people I have in my group. And this is wrong. And it’s wrong by a lot. And unless you know, the origin story, unless you know, why it is, what it is, and what it is, and talk about correction of it or talk about at least where the derivation was, you’re gonna be, you’re gonna be in bad shape.

Dwight Brown: 21:25
Yeah, absolutely, absolutely. And that there’s nothing worse than those times when you when you hear those words from a leader, and it’s it, it really can be devastating. And so but it is, it can be the impetus for action. And, you know, so in terms of in terms of where to start, I think that the best place to starta system asset inventory or a data asset inventory, start cataloging, what systems do you have in place, and then within that, what data is there and then start to add some of that data and put it into macro categories, That, to me, the key to data governance is starting high and moving down low, because the overwhelming part of it is that you look at how many data points the average organization is going to be dealing with. And it can be, oftentimes it can be in the millions, and we start looking at it like that. It’s kind of like, gosh, where do I start, but if you start to figure out, if you don’t know what, if you don’t have an asset inventory, you’re never going to get there, if you don’t know where your data is what it is, then you’re never going to get there. So start start at the macro level. So here are the systems that we have here are high level categories that are in those systems, and then start to break the data down. And, you know, the, and I don’t want to go too far afield here. But if you start to if you look at the concept of what’s called Master Data Management, we’ll start another topic in and of itself.

David Turetsky: 23:10
Absolutely.

Dwight Brown: 23:11
Data is essentially, what is that? What are the the source of truth, high level pieces of data that we have, and then pushing those pushing those down? And

David Turetsky: 23:24
it’s basically like, how do they work through the piping? You know, if you consider, you know, Master Master Data Management, where does the data come from? Where does it go? And where’s the piping? That takes it from and to it’s almost like blockchain for right for data?

Dwight Brown: 23:39
Yeah, exactly. Exactly. And so if you look at Master Data Management, the I found a great example out here that, that, that next step down after you have your systems is looking at the macro level categories for it. And so, you know, your macro level might be customer, product, location, employee and asset. And you start from there and then you start, then you start slating your data into those categories. And from there, you can you can then start to work down. And, you know, really the next step becomes what’s the high value data of that, that we have? And then work with that and then keep going from there. So it really you know, I think, in terms of the where do you start? That’s a good place. The other thing, the other thing that I would say is, start by start by trolling the internet learning more about data governance, what is it it’s not that everybody has to become an expert on it, right, but you can’t summon your organization, if you as leader don’t have at least a little idea of what data governance is, and they’re all you know, all kinds of resources out there. To do that, and,

David Turetsky: 24:59
and we’ll put some We’ll post some on the podcast. So if somebody wants to learn more about it, they can go to that.

Dwight Brown: 25:05
Yep. That’s perfect. That’s perfect. Yeah, and then determine what your priorities are. And yeah, based on your asset inventory and, and just keep working from there. This is very doable. Don’t try to bite it off all at once. If you don’t have a data, data governance system in place in your organization, start small, you know,

David Turetsky: 25:33
they can definitely make progress. I mean, you know, starting small, it’s critical, because, you know, even midsize or small and midsize companies, data is so plentiful now. We keep so much because because, you know, hard disk space, oh, gosh, hard disk space, the ability to store data, God, how old am I hard disk space, the ability to store data is so easy now, especially with AWS, where it’s so cheap to store, you know, anything. And so we hold on to things for so long, even though we don’t need them. And that’s actually another problem, is it? Good data stewardship means that you’re getting rid of things when you don’t need them anymore. You don’t hold on to things for too long, which I think some of some of the people I’ve worked with in the past may forget, is you don’t need history from day one of your organization, unless day one was maybe 10 years ago.

Dwight Brown: 26:27
Right. Exactly, exactly. And you determine what that looks like what the timelines, look what the timeline looks like. And now it’s like my closet at home, I’ve got to start cleaning it out at some point and doing it periodically.

David Turetsky: 26:43
Yeah, actually, way too many skeletons in my closet.

Dwight Brown: 26:49
Amen to that. The The other thing that I would that I would say is, don’t assume that you’re not doing data governance, if you are ml system in place, because I have yet to find an organization that doesn’t have some sort of data governance in place. It could be as small as the Report Writer documenting the definitions of some of the data points that they’re using. That’s data governance, that’s, that’s a piece of data governance. And, you know, like we’ve talked about, you may have people with a good amount of about a particular data point or a metric, or whatever that might be, how it’s used. And they’re, you know, oftentimes their informal agreements that have taken place along the way about how to use specific data, right. So it’s oftentimes there, it’s just a matter of starting to formalize it. So definitely would would say that getting past that overwhelming feeling is also realizing that you probably have something in place, it just may not be as formal as a formal data governance system.

David Turetsky: 28:03
And if you need help on a formal data governance system, please call Dwight at Turetsky Consulting. Exactly.

Dwight Brown: 28:08
Exactly. I’m here, I’m here to help.

David Turetsky: 28:14
Yes, you are. So Dwight, we’ve talked about a bunch of things around data governance. First of all, we talked about what it is, what is good data governance, and what is just data governance. We’ve talked about what’s in it for me and what you really could harm yourself, if you don’t actually have a data governance in place, or some kind of data governance in place. And lastly, we talked about where to begin and starting small and, you know, some kind of data governance is better than no data governance and getting started, you know, should be on understanding what your assets are. What else would you like to impart to our listeners before we close?

Dwight Brown: 29:03
I think it I think it really is the the piece just reiterating for people that you know, don’t get hung up on feeling too overwhelmed by this just, again, start small, realize what you’re what you’re currently doing already in this space. Look for people who look for people who have an interest in this also and then earmark those people as being sort of the champions, abet and and that that is one final will bring out as a key to success in any data governance system, and that is having effective champions. That means it doesn’t necessarily thing with that piece of data day in day out. Oftentimes it’s a leader within the organization and somebody who is interested and passionate about this, but is also well thought of in the organization and like any other effort, having a good key champion like that will really help you to, to make great strides with this,

David Turetsky: 30:10
I’ll add one more thing, which is make sure that it is a cross functional team. And that you get some champions that are in all the different aspects. Because in HR, we kind of think of things very HR. But there’s HR data that we get that affects all different areas of the organization, and that we collect data from all different areas of the organization. So make sure that you have people who can speak the language of the organization, and be able to uncover pieces of data that we rely on, even outside of HR.

Dwight Brown: 30:44
Yeah, I could not agree more, I think you hit the nail on the head with that. Be sure and be sure and look across borders with it and get those people and, and, you know, oftentimes, oftentimes what you get with that is that that person, just by the nature of working across borders, they will work across other borders and and work with their colleagues to help to get robust definitions and the definitions or whatever that might be that you’re working towards. So I I couldn’t agree more with you on that point. I think that’s key.

David Turetsky: 31:25
Thank you for agreeing. And thank you for joining us.

Dwight Brown: 31:28
Thank you for having me. It’s really been a pleasure. This has been fun.

David Turetsky: 31:32
Yeah, that’s been fun. We got to do it again sometime.

Dwight Brown: 31:34
Most definitely.

David Turetsky: 31:35
Awesome.

Dwight Brown: 31:36
I’ll look forward to the next one.

David Turetsky: 31:38
Okay, awesome. And thank you very much for listening. If you enjoyed please hit subscribe. Also, if you have friends who might enjoy it, please send it on to them. Thank you very much. Take care, and stay safe. Alright, I’m gonna hit stop now.

All content © 2021 Turetsky Consulting LLC .