Why Business Leaders Can’t Afford to Ignore Data Governance in 2025

×

Greg Frasca
Senior Solutions Architect
Senturus, Inc. 

Greg leads our data governance and strategy practice. He brings nearly 25 years of experience in data and analytics spanning numerous disciplines, industries, technologies, and lines of business. At present, his passion is ensuring that organizations are focused on the right strategy at the right time to maximize the impact of becoming data-driven.
 

Steve Reed-Pittman
Director of Enterprise Architecture and Engineering
Senturus, Inc. 

Steve’s career in data and analytics spans 20 years and includes clients from the Fortune 500 to the Inc. 5000. He is adept across many facets of analytics including data architecture in physical, virtual and cloud environments. Steve transitions easily from the server room to the boardroom, speaking the language and understanding the priorities, risks and complexity that accompany the selection, implementation and operation of technology in fast-moving enterprises.   

Read more

×

0:14
Hello everyone, and welcome today’s Senturus webinar on why business leaders can’t afford to ignore data governance in 2025.

0:24
Thanks as always for joining us.

0:26
Before we head into the material, just a quick overview of today’s agenda.

0:30
We’re going to do some quick introductions of our presenter.

0:34
Then we’re going to talk about data governance and the Senturus point of view.

0:37
On the subject of data governance, talking a bit about the business case for implementing DG within your organization, what it takes to obtain business buy in and how to go about implementation and achieving lasting adoption.

0:53
Finally, we’ll talk about how to get started on the DG path within your organization.

0:58
And as I said before, we’ll do some live Q&A at the end of today’s presentation.

1:04
Quick introductions our presenter today is Greg Frasca.

1:10
Greg is a Senior Solutions Architect here at Senturus.

1:13
Greg leads our data governance and strategy practice, and he brings nearly 25 years of experience in data and analytics, standing numerous disciplines, industries, technologies and lines of business.

1:27
Greg’s passion is to ensure that organizations are focused on the right strategy at the right time to maximize the impact of being data-driven.

1:35
So welcome, Greg.

1:36
Thank you for being here today.

1:38
And I’m Steve Reed Pittman, director of enterprise architecture and Engineering.

1:42
I’m here to do the kind of opening and closing, but Greg is going to be your expert on data governance.

1:48
And with that, we’re going to launch a quick poll here before we get started.

1:52
So the poll question is, what is the current state of data governance within your organization?

1:58
And I’m going to go ahead and launch that now.

2:01
So you should be able to answer that question within your Zoom interface.

2:04
And the options are, is your DG initiative just nascent?

2:08
There’s no formal DG yet.

2:10
It’s complicated, you know, maybe a little bit, maybe some kind of people working at cross purposes.

2:18
Do you have formal DG already implemented or are you in crisis mode?

2:23
And what we find in terms of crisis mode is the triggering event for ADG implementation often can be a compliance problem and audit.

2:32
So, you know, something really like legally or financially significant that can force an organization into needing to implement some quick data governments.

2:42
And with that, I’m going to go ahead and wrap up the poll, share those results out.

2:49
So again, about half of you don’t have any data governance initiative in place yet or, or you’re just at the very beginning of that.

2:56
Some of you are in that complicated space that, you know, maybe a little bit of data governance may be struggling with how to kind of carry that forward.

3:06
And a little under 1/5 of you have a formal program already in place, go ahead and stop sharing those results.

3:15
And with that, I am going to hand it over to our presenter.

3:19
So Greg, take it away.

3:23
Thanks, Steve.

3:23
Appreciate interesting results there.

3:25
Not all unexpected, but I know that it is complicated for a lot of you or there have been multiple tents in launching a data governance program that have kind of died on the vine a bit.

3:36
And that’s kind of what we’re here to talk about today.

3:39
For those of you that have a formal data governance program implemented, I still think this is of a lot of value in terms of how do you go about internally marketing and acting as ambassadors for that formal data governance program during your organization or obtaining business buy in for an initiative that comes out of a data governance program, such as a data cataloging or data dictionary effort or data quality effort or, or what not, right.

4:06
So First things first, one of the table set what data governance means to Senturus in the context of today’s webinar means a lot of different things to a lot of different people.

4:18
I just want to make sure we’re all on the same page going in first, right?

4:22
So first, I really want to address the why we are here.

4:25
And as the webinar title communicates that business leaders can’t afford to ignore data governance in 2025, the first example I’d like to talk about is mitigating risk, specifically centering this around an AI perspective, right as AI is moving forward at a quite a frenetic pace right now and organizations are all looking at how exactly fits into their strategic direction or they’re not looking at that and they just want that new shiny thing right now.

4:56
It’s really important to understand the risk of giving unfettered access to your AI systems of your, of sorry, your source systems to AI without any safeguards or governance in place, right?

5:07
You don’t want to be running with scissors, especially when it comes to AI, which there is a bit of a trust fall on implementing something like a chat bot externally or using AI internally for a number of purposes.

5:18
You want to make sure everything’s in place and properly considered when before implementation.

5:24
So there’s this common misconception that AI is going to cure data quality issues.

5:31
And while it can help you identify data quality issues, it’s by no means a fully self healing kind of system, right?

5:39
Human interaction is still 100% necessary with just like with any data and analytics initiative.

5:45
And it’s kind of more important than ever when it comes to AI.

5:48
Thinking about the garbage in garbage outs analogy, which I’m sure many of you, if you’ve been a practitioner in this space for a long time or in leadership, you’ve heard garbage in, garbage out.

5:59
And it applies to AI better than as well as any example, right?

6:04
And if the data feeding these models is flawed or incomplete, the results that AI produces are going to be unreliable.

6:14
So if you’re relying on an AI system to make critical business decisions like predicting customer demand, setting pricing for a product, if the data you feed into that system is incomplete, inaccurate, or outdated, your results are going to demonstrate that and, and be reflective of that.

6:32
So if you’re missing key market trends or looking at human error from old spreadsheets or, or taking into account things that have not been approved at an enterprise level, the outputs are going to become unreliable.

6:46
And what does that lead to ultimately that leads to a, it could hurt your bottom line as an organization from a revenue perspective and cost.

6:55
It could confuse your teams or worse your customers, which is going to inevitably erode trust.

7:01
And that’s you’re going to see a common theme in that today is that without data governance, there is a risk of not only the common things you see like a compliance or regulatory infraction or spitting out the wrong information to make critical business decisions, but more of the soft side of that which is eroding trust in your data asset.

7:25
So it’s really important to consider, especially as we go into this phase of artificial intelligence and machine learning algorithms, that your data is in a good state for it to be consumed by AI and automate.

7:38
Those processes that are there should be automated and will be automated.

7:42
As you know, the inevitable March of time moves on.

7:44
So compliance and security risks are another often go topic.

7:49
Like I mentioned, you know, ADG program won’t be responsible for understanding the full legality of each regulation.

7:56
When these regulations change daily, they’re very fragmented.

8:01
If, if you’re operating specifically within the United States, States have their own regulatory, you know, CCPA leading the way in terms of the Privacy Act, but other states are, are implementing their own version of that or have implemented their own version of that.

8:17
And then when you get into international business, you know, all, all bets are off.

8:20
There’s, there’s a lot of different permutations you need to consider when talking about compliance in that regard, right?

8:28
So the DG program will be responsible for knowing where your sensitive, sensitive data lives, if it has the proper tags and controls on it, how it reverses through your organization and how you safeguard it to provide it to only people that need it for business critical operations.

8:48
Right?

8:50
Next piece is data lineage.

8:51
And data lineage is key to understanding the how of an AI solution.

8:56
If you don’t have traceability back to how AI garnered an answer or blinked an answer, you’re really flying blocked and you’re really flying blind into how AI is coming up with these answers that it’s providing not only externally but internally to your user base, right?

9:12
You need that traceability in order to train the model to, to not look a certain place or look a certain place where it’s not looking at that current time.

9:19
So the, the permutations of this are endless.

9:23
If you can’t answer the how, you really need to be answer the how.

9:26
And a good data governance program will be responsible for having that data lineage well documented and well traceable for the organization to look at.

9:37
Next piece is siloed AI.

9:39
We’ll spend a ton of time on this.

9:40
But as you can imagine, sometimes in my experience, that sometimes siloed data will be from the cause of a specific line of business going rogue in terms of coming up with our own analytic solution.

9:57
Not to pick on marketing, but I’ve seen in, in a lot of times in my experience that marketing will go rogue and use vendors to drive their analytics and drive their customer outreach programs and their customer retention, acquisition, loyalty, etcetera.

10:14
There’s a lot of reasons behind that we won’t get in today, but if they are producing data that the organization is consuming and they are defining things in their own specific way, that’s fine.

10:26
That’s perfectly fine, but something like AI needs to understand that this is marketing’s definition of what even something as simple as revenue, This is marketing’s definition of how revenue is measured versus how this is finance’s definition of revenue is mentioned, right?

10:41
Two entire they could be two entirely different things, but they’re very important for different aspects of the organization.

10:49
So next piece is model drifts.

10:53
You can think of model drift as your information going stale.

10:57
And this would apply not just to AI that we’re talking about in the context of this slide, but just like launching a new dashboard or a new report over time, the data elements that make up that dashboard or report for that AI system are going to become stale.

11:13
You’re going to have whether you’ve changed product lines within your organization, you’ve created a new date revenue stream, those sorts of things need to be accounted for in your KPIs and your metrics, just like they need to be accounted for in AI, right?

11:28
So it would be the job of a data governance organization to be looking at those things and making sure that those are prioritized that that you’re your daily flash report or your, your dashboard are reflective of those, just like your AI system is also reflective of those changes.

11:46
And hopefully they’re all, you know, playing from that same sheet of music in, in that regard.

11:51
And lastly, this is sort of the, the one that everyone kind of associates data governance with is regulated industries demanded specifically, you know, if you’re on this call coming from the healthcare, finance, insurance, I don’t have it on here, but education industries, you need to be, you are subjected to audits at any time and you need to be auditable.

12:19
So you need to make sure that that data is properly provided and at the ready for those audits or else you’re going to be spending a lot of manual time and effort going through that audit process and, and obviously it’s not going to look well for you.

12:34
So really a really good example of everything I’m talking about in this slide goes to something called Zillow Offers, if you’re familiar with it.

12:43
So in 2021, Zillow launched a new program called Zillow Offers, and it was an AI home buying program and its intent was to streamline real estate transactions by using machine learning to predict home prices and automate the buying and selling process.

13:01
So the goal was for Zillow, they’re sitting on an absolute gold mine of real estate data at Zillow.

13:07
And the goal was to rapidly flip those homes for profit that they can buy, right?

13:12
They were using the data that they had at their fingertips and at their disposal just by the nature of their business model and utilizing that to turn it into a new revenue stream.

13:21
Great idea, right?

13:23
However, they rushed it to market.

13:26
The models were not properly trained.

13:28
They were not looking at the right or any market in some instances, any market trend kind of data.

13:34
They weren’t looking at a number of attributes like, you know, does the home have mold or does the home smell like cats or all those types of things, right?

13:43
And they were at some level, they just did a trust fall with AI that AI would purchase these homes at the right price.

13:52
And as you can see by the numbers on the right, that absolutely did not happen.

13:57
These numbers are somewhat staggering and they almost seem unbelievable.

14:00
But they did suffer half a billion dollars in financial losses from buying properties that they were unable to sell.

14:07
That was the majority of it.

14:08
Obviously, there’s a bit more in terms of implementing the program and administering it, but $500 million in financial losses, which ultimately led to 25% of their workforce layoffs.

14:20
In other words, some other considerations of how they came to 25% of workforce layoffs, but that was definitely the lion’s share of how they lost that money.

14:28
So it’s a good example of how you really need to make sure that everyone in your organization at an enterprise level is looking at data governance and has the right processing, auditing, traceability in place before launching a major program that could really, I mean, devastate the bottom line of your company.

14:49
Now, Zillow has recovered over time, but not only that, but it really kind of hurt the trust in the organization as well, which is a common theme you’re going to be seeing throughout today’s presentation.

14:59
So the next piece, you know, we talked about more like some of the technical reasons, but another big reason why data governance is non negotiable is the necessity for process, right?

15:11
Nobody likes to hear the word process.

15:13
It sort of invokes a feeling of a lack of progress, but it doesn’t need to be that way, right?

15:19
So you can tactfully and thoughtfully in front process that is unique to your organization.

15:26
And that’s a key theme as well, is that every organization is different.

15:31
It has its own culture, its own group of people, its own history and its own general data maturity levels, right.

15:38
So finding that sweet spot between having a formal process that will be accepted and is adaptable and adds value to your organization while not competing progress is a bit where the art meets the science in this case, right?

15:52
Is implementing process at the risk of not impeding progress.

15:59
So we talked about how human intervention is still a very necessary piece to your overall data state.

16:07
And this is exactly what process is right.

16:10
So first you need to process.

16:13
You need process to catch both human and system mistakes.

16:18
And sometimes this comes in the form of embedding a data governance program.

16:21
You know, your organization may have a great change control and change management process already in place.

16:28
Data governance should only be something that you append to that in in a in a positive way, right?

16:36
But sometimes in in lieu of that, if you don’t have that, a data governance committee may lead the way.

16:40
So you need to really look at your organization and what your current capabilities are, where your strengths are and adjust to how exactly your organization is currently running.

16:50
Whatever makes the most sense for your organization to prevent careless errors and ensure compliance and control.

16:58
The overall propagation of data is really important.

17:01
So you know corporate data is key in making decisions for your business and how it should be run.

17:09
It should not be allowed to grow in the wild in any sense of the word.

17:13
So it should be properly cared for and that’s where process really comes in.

17:19
Another good example of where a process breakdown has cost an organization a lot of money and caused red face is around Citigroup.

17:30
This happened in 2020.

17:34
So Citigroup, I won’t get too much in the detail, but Citigroup was acting as a loan agent for the company Revlon and Revlon had taken out a large loan.

17:46
They were in financial straits and they had taken out a large loan and Revlon intended to send just a routine interest payment of about $7.8 million to their lenders.

17:58
However, due to a lack of process and some antiquated software, they accidentally wired the entire principle of the loan, which is around $900 million directly to the lender using Citigroup’s own funds.

18:12
So they spent the next two years trying to get payments back from the lenders.

18:17
Many lenders did willingly, you know, allow the transaction to be reversed, but a lot didn’t.

18:24
And they said, you know, this is what we’re owed and we’re going to take it.

18:28
Some may say rightfully so.

18:29
We’re not here to judge the legality of that.

18:31
But inevitably it costed them about $500 million upfront of what the lenders were allowed to keep.

18:40
Now, they were able to scratch and cost some of that back, but initially it was $500 million that they were on the hook for.

18:46
And additionally, which is more painful as they would have paid that money back anyway, just not on their terms, is that the Federal Reserve and a few other federal agencies came in and audited them and uncovered a slew really significant risks in the company and determine that they were unable to determine their own risk and find them $400 million in Federal Reserve funds.

19:14
So huge number once again.

19:16
And but just what really is more of what you don’t see here?

19:20
We’re going to get into hard ROI and soft ROI in a minute.

19:23
What you don’t see here is that the lack of trust in the organization to act as a loan agent on major loans has really hurt their business right now.

19:33
Again, their city too big to fail and they’re coming back.

19:36
But, it’s really a very cautionary tale.

19:40
I encourage everyone on this call, if you’re interested to go look at the Zillow and the city case because they’re pretty fascinating if you’re into this sort of stuff.

19:50
So successful data governance is a bit of a paradigm shift for an organization.

19:56
And you really need to ask yourself if you believe that the cost of operating a data governance program is worse than the cost of not having one.

20:04
And recently he was on a call with our CEO, John Peterson, and he said something that was a bit profound to me.

20:09
So that data governance is like incorporating fiber into your diet.

20:13
It’s a necessary lifestyle change that you’re much better off for in the long run, right?

20:17
It’s something you’re not necessarily thinking about.

20:20
But as soon as you get that message from your doctor that you’re not incorporating enough fiber into your diet, you need to make that change immediately, right?

20:28
So data governance is being proactive about making that change and never having to hear that, that warning from your doctor, right?

20:35
It’s challenging.

20:37
Throughout my career, I’ve been through a lot of data strategy and data governance engagements.

20:44
And I can say that not necessarily starting a data governance program that’s easy, but maintaining it and nurturing it and adhering to it over time.

20:52
It’s one of the more challenging, but one of the more rewarding things that I’ve been a part of.

20:56
Right?

20:56
So eat your Wheaties, take your fiber, and only good things will happen.

21:01
There’s a couple more things just to set the table is number one is data governance.

21:06
I want to get into exactly what it is for the purpose of today’s conversation and then a little bit into what it isn’t.

21:14
So #1 it’s a structured approach to managing data as a strategic asset.

21:21
As I mentioned, data needs TLC and attention just like anything else in your organization.

21:28
So some level of structure is required, right?

21:31
And I know that I mentioned earlier that could be a bad word sometimes associated with data governance.

21:36
So how you approach that, how you define that, how you communicate that to your audience is really important.

21:42
We’ll get you set a little bit later.

21:45
Goes into #2 applying policies and controls.

21:47
Another couple of other naughty words for people who just want to move data forward quickly.

21:54
But applying policy and controls to ensure data quality, consistency, definition, security and compliance is key and is necessary for a successful data governance program.

22:06
Not all these things will be equal depending on your industry and your line of business and the aspect you’re coming from, but all these things need to be considered when considering data governance.

22:17
And lastly, and maybe most importantly is it embeds clear ownership and accountability across the enterprise.

22:25
So cross functional data, which is becoming more and more the vast majority of data at a company, it could be a bit of a hot potato at times of who owns the data and who’s responsible for the data.

22:36
And it’s important to make sure your organization understands and is naming those people and making sure that is communicated throughout the organization of who is accountable for, you know, you can call it your CRM, your ERP, your HR, whatever your key source systems are, you need to make sure that there’s clear ownership and accountability.

22:54
And that speaks to the roles of a data owner, a data steward, and all the different roles that that make up the data governance committee.

23:02
So what data governance is not, it is not the sole responsibility of IT.

23:09
And I cannot stress this enough, we’re going to cover it in more detail, but if IT is solely responsible for the function of data governance organization, it’s not going to succeed.

23:19
I can confidently say that it can help get a program off the ground and I see could be one of the biggest champions of the program, but cannot shoulder the burden of the program #2 is data governance is not a one time project.

23:32
It is a continuous capability and you must adapt to the environment around you, to your organization, whether you’re in heavy acquisition mode, whether you’re going through changes in your product lines and what you’re delivering to your customers.

23:47
It’s really important that your DG is constantly adapting to that and your industry as well.

23:52
You mentioned regulatory procedures, compliance, the legality of it all.

23:57
It’s changing every single day, and you need to make sure that DG is aware of that and adhering to those changes over time.

24:03
And lastly, just what it is not is an impediment of progress.

24:07
This is really the biggest misconception in my experience of data governance.

24:12
So if it’s done poorly without proper oversight, it can definitely impede progress.

24:18
But done correctly and done right, it can drive agility for the organization and leads to a lot of economies of scale and accelerate returns on your data and analytics efforts rather than delays, right?

24:31
It’s a little bit of a heavy lift up front, takes a little bit of work kind of like the third little pig and the three little pig story.

24:37
But once you have that House of bricks built, you are able to unlock new potential for your organization and your data assets.

24:47
So that being said, with the table set, I want to talk a little bit about the business case for data governance.

24:53
It’s really important.

24:54
Again, that is not solely a function of IT, but you need, if it is a directive of IT to start a data governance program, you really need to internally sell that to the business.

25:08
Should be multiple lines of business or whether or not it’s one line of business clamouring forward upfront, but IT should be the enterprise overall.

25:17
So why is it so?

25:18
I found this Gartner study from 2023 that was very interesting to me and said by 2020 680% of organizations will have deployed multiple data hubs within their data fabric to drive mission critical data and analytic sharing government.

25:34
How?

25:34
Governance, excuse me?

25:36
However, 20% will implement a single data and analytics program to unify and automate discrete data.

25:46
So what that means for organizations?

25:49
Is that why you may be dodging bullets in the short term?

25:53
A lack of an enterprise data governance program is a ticking time bomb in a lot of ways.

25:59
It’s not always about the compliance or regulatory hit that a lot of people attach to that.

26:05
In the vast majority of cases, it’s really a very common symptom of a lack of trust or loss of trust in your data asset, which if you haven’t experienced, you may have heard about, you know, a lot of organizations just don’t have trust in their data.

26:22
The data is there.

26:23
It’s not there to fool you or manipulate you.

26:25
It’s there to be used as an asset.

26:28
And once people lose trust in that asset, it’s hard to get it back.

26:32
It takes twice the effort to gain back the trust in data than it does to set it right in the 1st place, right?

26:42
So it’s really important to set those boundaries first and set those guidelines upfront to make sure that you’re using a universal data governance program and make sure that everyone in the organization is aware that your data is an enterprise asset, not just something specific and silo to their organization.

27:02
So how do we do that?

27:04
You know, as with any business case at any organization, whether it’s any data and analytics initiative or any initiative, it’s really imperative to assign expected ROI to your new program when you’re trying to, you know, sell it up the food chain, so to speak.

27:18
So we define hard ROI in three buckets, which is increase of revenue, decrease of cost or risk mitigation.

27:28
There’s a number of soft ROI categories such as number of hours, manual hours spent fixing an error, what that can do just to morale of some of your really valuable employees and such that those are the soft ROI that we’re not going to cover today, but are in my estimation very close to or equally as important as the hard ROI.

27:47
But for the purposes of selling data governance program to your organization, we’re going to focus on the hard ROI today.

27:53
So a lot of people don’t associate data governance to increase revenue.

27:59
But just by the nature of having clean, understandable and cleanse data, you’re already putting yourself in a position for increased revenue.

28:08
And data monetization is a really good example of that.

28:11
People are coming up with new ways every day to monetize their data.

28:14
Scary and exciting, but that data has significantly more value if it’s accurate, consistent, standardized, and timely.

28:24
Think about, if you’re going to trade in your car, you want to make sure that that car is squeaky clean inside and out, that you’ve done everything you possibly can to make that car as attractive as possible to the dealership or the buyer of who’s going to purchase it on a much, much grander scale.

28:40
The quality of your data is really important to that.

28:44
And additionally, you know, having those types of characteristics of clean data is really going to allow you to consider new revenue streams.

28:50
It’s going to make, it’s going to remove the noise and really allow you to make better decisions, think about new revenue streams, see things in a whole new way without that noise cluttering your thinking.

29:02
The next phase is having a number of ways to decrease costs.

29:08
And the most important, I won’t dig in too deep on this, but the most important is eliminating redundancy in your efforts.

29:13
Without a data governance program in which everyone has standards to run by, is that you’re inevitably going to get analysts or people within each line of business that are coming up with their own business rules.

29:28
And they may be, you know, if you’re looking at an event, they may be 80% in agreement on those things, but they’re not 100% in agreement on those things.

29:34
So putting everything together and making sure you have all of those rules of the road, so to speak, defined is really going to decrease your costs across the board.

29:47
And additionally, as a segue into the risk mitigation pieces, if you centralized all your data policies, you can streamline your compliance.

29:57
And that goes back to the auditability and traceability I mentioned earlier, like the most apparent form of good data governance that people talk about the most is the one associated with risk mitigation.

30:11
So a sound governance program is going to prioritize your data cataloging effort early, very early understanding what do we have, what are we working with, right?

30:22
And through that exercise, you’re going to identify your personally identifiable information or otherwise sensitive information to the organization that not only should definitely your external users be seeing, but your internal users as well.

30:37
And hopefully what comes out of that is an initiative to tag secure and only provide that data the parties that absolutely need it for business critical operations.

30:50
And again, this is really tied to your industry.

30:53
There’s a slew of regulatory and compliance factors that need to be considered depending on your industry.

30:58
Hopefully if your company’s large enough, you would have access to a legal team that can help you with that.

31:02
But if not really understanding that and making sure leadership is understanding of that you can’t move forward with a new initiative.

31:11
If there’s any risk of personally identifiable information being distributed to the organization to, to people that it should not go to, right.

31:20
So it’s very important.

31:24
So in building that business case, so you at a high level at least at this point quantified the expected ROI in one or hopefully many of those buckets we just talked about.

31:36
The next best step would be to clearly illustrate the financial and operational impact of taking action first, the potential cost from doing nothing, right.

31:46
So Citigroup was a really good example of that and asking yourself how and asking yourself how the Citigroup very extreme example might apply to your organization.

31:57
For example, if you’re AB to C firm and you’ve committed API infraction at some level that has gone public, what are the implications of that in terms of the thousands, 10s of thousands, hundreds of thousands or more of customer records that you’ve exposed to the public, right.

32:14
And once one case goes public, people are going to start looking at how many other cases may have, how many other fractions may have occurred, right?

32:24
So it’s really important to, I don’t really like to operate from the fear principle as a starting point, but it’s absolutely imperative that you, you do that at some regard when it comes to governance, because governance, you know, at its core is mitigating those types of things.

32:43
So a few other pieces on this slide is number one because you want to start with an individual use case.

32:51
You know, data governance is a prime example of the phrase don’t boil the ocean.

32:57
You don’t want to start with a monolithic giant data governance program that you expect to blanketly cover all of your initiatives going on at your organization.

33:08
Find something that’s relevant and timely to your organization that’s getting a lot of attention right now and attach yourself to that.

33:15
Attach yourself to that.

33:16
Wrap your data governance efforts into that without impeding progress again, right.

33:21
You know, kind of playing from the side lines and saying, you know, this is how we should be doing things to move this into production and move this into a valuable production state and really go around that because that’s how you’re going to get the most attention again.

33:34
If you build something that is not, it’s not an example of if you build it, they will come, right?

33:42
It’s actually quite the opposite of that.

33:45
So next piece I really want to hit home on is finding a coalition of data champions.

33:50
Every client I’ve been at in my many years, I know who these people are.

33:56
They exist in every organization that I’ve had the privilege of working with.

34:02
They understand the data in their domain better than anyone and they’re absolutely incredibly important to the organization.

34:08
And I’m imagining that many of you on this call listening right now already have these names that are coming to mind.

34:14
These people that understand their domain, whether it’s the marketing analyst, the financial analyst, or what not.

34:21
The people that understand the data, understand the nature of data, how it traverses, who the key players are, assemble that coalition, build that team of those folks.

34:31
Because I can almost guarantee you, not necessarily, but I can almost guarantee you that they will be very willing to listen to what you’re trying to talk about in terms of data governance because it’s probably a huge impact on their life, a lack of data governance and the fact of the manual effort they have to put in every day to data.

34:52
And lastly, you know, and this comes more when you build the program, but demonstrate the value of the program early and often.

34:59
It’s really key to do that and make sure that, you know, once you have a data governance program bought into that you’re not just, you know, kind of going into a vacuum and building out the data governance.

35:12
Make sure everyone understands what you’re trying to do, how you’re prioritizing, how this is changing between meetings, whether or not you have a data governance committee that meets weekly, semi monthly, whatever, you know, make sure that whatever comes out of that meeting is being communicated throughout your organization, downward, sideways, upwards.

35:30
You are communicating that have an internal marketing plan.

35:33
Even I’ve seen organizations I work with the great organization in the retail and gas industry that they went so far to have their marketing team build out, you know, posters that they put up around, around the hallways and then the Cafe of their organization explaining what they were trying to do.

35:53
And they would change those things out monthly or so in terms of their initiatives and what people should be looking for within the organization.

36:01
I thought it was a really great application of, of demonstrating the value of the program.

36:06
So to whatever extent, just make sure it might, doesn’t have to be that extreme, but to whatever extent, make sure you are communicating throughout the organization how data governance is making a difference within your organization.

36:18
And it might not catch fire immediately, but if the more you keep doing it really can catch fire over time.

36:26
A small spark can start a great fire.

36:28
So next piece I want to get into.

36:31
Now you’ve made the business case, you’re ready to go to leadership and you’re ready to obtain business buy in, right?

36:39
You’re ready to sell it internally within your organization.

36:42
It really is an organization.

36:43
It really is exercise of sales to sell this within your organization.

36:51
And the first thing I wanted to talk about, we’ve touched on it already, but when you’re ready to secure executive sponsorship and business buy in, the best possible way to do this is to attach your initiative to a strategic use case or a set of strategic use cases.

37:07
So interesting separate Gartner study I found from 2024 predicted that 80% of data governance initiatives will fail due to a lack of prioritized business outcomes.

37:20
So not just data governance, this slide and this statistic speaks to data governance, but many other data and analytics initiatives.

37:30
It’s not a case of again, if you build it, they will come.

37:34
It’s the opposite.

37:35
You really need to attach yourself to something that is getting attention and is a top priority organization to apply data governance to demonstrate the value of data governance overall.

37:47
And if you do this and you do it correctly, the next initiative that starts is going to ask for that assistance and ask for that ride along from data governance to, you know, if you don’t have that mandate to ride along and be part of the initiative.

38:02
So, you know, I’ve seen time and again IT wanting to build this mature data governance program before they bring it to the business and apply it to those real world examples.

38:14
It does not work.

38:16
the IT team can provide the foundation, they can provide the guidelines, they can set the rules, but the business really needs to drive the priorities and direction of the group.

38:26
And the best possible way to do that is to get the attention of the business through demonstrable examples of relevant conditions, right?

38:35
So, keep that in mind.

38:38
Again, you don’t want to build a giant bureaucratic type of data governance organization.

38:45
You want to do it slowly and surely over time.

38:49
One way we do that when you’re presenting a case for brand new data governance initiative or continuing or enhancing an existing data governance program is to find how you’re going to X execute.

39:01
So this slide is a format I’ve used many times in my career to explain the reasons of why you’re justifying an effort and the business value in the first place.

39:13
So first I’ll explain the concepts here on this slide and then I’ll go into a real world example.

39:19
So these five categories is how we will categorize any use case and help prioritize use cases.

39:29
Overall, when we’re doing a use case solicitation and prioritization exercise, The first one is economic value.

39:35
What is the hard ROI?

39:36
We covered that earlier, increase revenue, decrease costs or mitigate risk.

39:40
Make sure you have that as a number one in your back pocket when delivering this.

39:46
Number 2 is strategic alignment.

39:47
We talked about that aligned to a specific use case.

39:51
Next one is economies of scale.

39:53
So if you take on an effort, where will you find accelerated returns in the future?

39:57
And let’s say you want a master data management effort for your product domain down to the SKU level, right?

40:07
If you do that for a specific cause, for a specific department, what other departments is that going to help, right?

40:14
Theoretically that is going to cascade into assisting other departments greatly, right?

40:19
So understanding the economies of scales and the accelerated returns you can get is key Level of effort, you know, if those three things check, but the level of effort’s going to be, you know, 12 months to bring something to the table, you might want to de prioritize that to some extent, right.

40:36
So looking at that as what is the overall level of effort maybe getting to that product normalization piece is, you know, implementing 12 different source systems that you are tasked with through, you know, maybe through acquisition or others efforts, right?

40:55
So maybe doesn’t make sense right now to tackle that, right.

40:57
So really look at level of effort and then look at dependencies.

41:00
And that’s really just a feasibility study.

41:01
Is it feasible to do at this time.

41:08
So next piece is establishing formal data governance from this is the real world example.

41:15
So this comes from ABC retailer I worked with in the past.

41:20
I’ve scrubbed it and genericized it to very much to make sure it’s sort of abstracted from the business.

41:29
But really we looked at, you know, the economic value of establishing a formal data governance program was reducing the cost of doing business and allowing for trusted data to confidently flow throughout the organization.

41:41
I should comment on the bottom left.

41:45
What they really wanted to get to in those priority rankings was an advanced analytics platform deployment and they knew that their data was not in the right form to be able to do that, right.

41:56
So they needed to really take inventory of their entire data state and understand what they had.

42:02
So the first step with that would be establish a data governance committee that would go in there and prioritize those and had the mandate to kind of spider web throughout the organization and find out where all those data assets lived, who owned those data assets and what those data assets contained.

42:20
So the next piece was strategic alignment again getting set advanced analytics platform 2 economies of scale, faster deployment of AIML models and seamless integration of new data sources over time.

42:34
If they can do it for one and set the ground rules if they wanted to acquire a new company, it’s much less of an effort to integrate that data into an enterprise wide data strategy.

42:47
If they already have the rules of the road set in place, right.

42:52
It was a bit of an effort upfront to name roles, train staff, garner adoption.

42:59
So we estimated about 6 to 12 months.

43:02
And really for this step one, establishing the formal data governance program, the defences weren’t luck.

43:08
They needed executive sponsorship.

43:10
That is an absolute again, non negotiable things that you need executive sponsorship to give you that mandate cross functional buy in is that you know the company is in it as a whole and act and that leads to that access of siloed data and change and management.

43:29
So this is a good way to start these five categories.

43:32
We have a scoring system at Senturus that we apply to this that will come up with what we call the voice of the business scoring just based on what we’ve heard from everyone and our approach to interviewing specific people, whether line of business or as a whole and applying values to all of these five categories.

43:54
So just a couple minutes here.

43:56
I want to stay spend on implementation.

43:59
We’re not going to spend a lot of time here, but what does the actual implementation look like and what are the characteristics of it?

44:05
And then show some specific collateral that you can, if you download the deck will be available to you.

44:13
So the first piece is what a holistic data governance program looks like.

44:18
So a holistic data governance program connects an organization’s data strategy to its practical application.

44:24
What you see here is the operational data management side, right?

44:29
And we’ve talked a lot about that today.

44:30
And that’s the collecting, processing and maintaining real time business data.

44:35
And that is sort of a baton pass from your source systems, right?

44:39
Marketing is initially responsible for CRM data, who your vendor is, how the, how the data is created and created within your source.

44:48
And then it’s, it’s sort of a baton pass to your data and analytics team of how they go about maintaining that data, organizing it and defining it for the business.

44:58
So that that is a shared responsibility between IT and the business.

45:01
The next is data privacy and security, very much a function of IT and your infrastructure teams.

45:08
You probably other than maybe if you have a legal team defining what the privacy and security rules are and IT and the business from determining what they know is out there.

45:19
It’s really a function of IT to be responsible for, for making sure that data is secure and private and only being presented to those that need it for business critical application data quality.

45:33
This is very much a shared responsibility between the business and IT to maintain accuracy, completeness, consistency, timeliness and validity and uniqueness.

45:42
Sure, a lot of you have heard those exact words uttered many times on webinars such as this, but it is important to call out.

45:50
And lastly, metadata.

45:51
I would strongly recommend if you’re coming from that place of Nason, see we talked about earlier of starting with a cataloging and defining of your overall data inventory, right, Going through understanding what you have.

46:05
And just one piece I want to bring up here is that this does not need to, you don’t need tooling for this.

46:12
A tooling can be very helpful, especially as you mature tooling is it becomes more and more important.

46:20
But don’t let a lack of tooling get in the way of this cataloging cataloguing exercise.

46:25
You can easily do this in Excel or Google Sheets or whatever your spreadsheet tool of choices of cataloguing everything.

46:32
And the great news is if you start there all of any tool you buy on the market, any best of every tool has the capability of importing the data you’ve already created.

46:42
You’re not starting from scratch.

46:43
You’re going through an initial import exercise to bring that data into your tool, and then you’re curating it and, and, and making it, you know, you’re manipulating it to fit the tool, but the vast majority of the work is done once you have the data catalogued and defined throughout your organization.

47:03
And then lastly, you’ll see on the left, organizational change management.

47:07
Absolutely.

47:08
I live and die by the saying of adoption is the hardest part.

47:13
And with data governance that applies no better than to data governance is that organizational change management is absolutely key.

47:20
And it’s something you need to be thinking about from the inception of a data governance program or the inception of any data governance related use case you want to look out for.

47:30
So change management and adoption are absolutely something you want to look for.

47:35
I subscribe a lot to the pro sci organizational change management philosophy, but there’s a lot out there.

47:43
I suggest you look into it and I suggest you make sure that change management and adoption are absolutely top of mind when implementing anything, anything from ADNA perspective, but specifically data governance for the purposes of today’s conversation.

48:00
Couple of things quickly here is that, you know, we really subscribe to, and this was demonstrated by previous rest of the webinar, is that we really subscribe to concept of non invasive data governance, which was a book that I strongly, if you’re interested in this looking up.

48:20
So data governance, really non invasive data governance, excuse me, really takes a practical approach to data governance.

48:29
A lot of what I said already was inspired by this book of data governance, but it really defines, really clearly defines what the roles are, what the tools are, what the levels are of people that should be, what they should be responsible for when implementing a data governance program.

48:46
It talks about how it should meet the need where it is, which alludes to the use case example I used earlier, minimize resistance to foster adoption, which is don’t boil the ocean as I mentioned earlier and should be democratized.

48:58
It should not just be a function of IT or anyone specific line of business.

49:01
It should be the voice of the business holistically of anyone that touches any enterprise data within your organization.

49:10
And lastly, before I hand it back to Steve we have a sample data governance racy that we use.

49:16
And once you download the deck after this presentation, you’ll be able to see the unblurred portions of this.

49:23
But I just really wanted to point out one thing is, you know, you can see here the consulted version of what it should be responsible for only consulted on in terms of a data governance program, right?

49:37
So establishing those policies and standards that should be done by the business, the ownership and stewardship, etcetera.

49:45
So when you download the deck, you can see that if you have any questions, you can reach out to me.

49:50
And before I hand it back to Steve, just a quick recap of everything we discussed today.

49:57
Basically data governance is it’s becoming more and more no longer and nice to have.

50:02
You can look at Gartner, McKenzie, top rated people in the data and analytics space are all in agreement right now at some level of data governance.

50:15
IT it’s non negotiable at this point for an organization of any size.

50:20
If you really want to be data-driven and monetize and use your data as an asset.

50:24
Overall the key drivers for data governance adoption.

50:30
When you look at that hard ROI, increase revenue, cost reduction, risk mitigation to improve regulatory compliance, business buying and implementation, make sure you get an executive sponsor cannot bring that home enough.

50:45
Find someone that is going to be your ride or die on a data governance initiative, right.

50:51
Make sure they are clear in their alignment with the business objectives and you have a structured approach that integrates both IT and business teams in a lot of ways when done right, it can it can really mend some the gap between IT and the business overall.

51:06
So sustained adoption and best practices, make sure you have clear ownership and accountability and that you are focused on adoption and change management at all times, at least until you know you take the training wheels off of a data governance program, make sure that there’s sustained adoption around it.

51:25
And lastly, getting started, right?

51:27
So best way to get started, if you answered that question of Nacency or it’s complicated, really step back and take an inventory of what your current data governance maturity is.

51:39
What’s your organization’s appetite for data governance is, what your business objectives are that could data governance can be applied to and what those use cases are that can demonstrate value early and often and glob onto one of those and make sure you’re part of one of those.

51:56
And this is where myself, some of my colleagues at Senturus could really help you and would love to help you on this is getting that assessment and that clear picture of where you stand now and how you go about implementing data governance at your organization.

52:11
So thanks everyone for your time today.

52:15
Really appreciate it.

52:16
I’m going to hand it back to Steve for a few comments and hopefully we still have time, but it can yap too much for some questions and answers.

52:23
Thank you.

52:25
All right, thank you, Greg.

52:26
And yeah, just a reminder for everybody, if you do have questions, go ahead and drop them into the Q&A panel.

52:32
We are coming close to the top of the hour.

52:34
So if we don’t get to all of your questions today, we will reach out to you after the webinar.

52:38
So don’t hesitate to those questions in there before we get into quick Q&A #1 taking the next step.

52:45
You know, what do you do from here, right?

52:47
How do you, how do you begin working on DG in your organization?

52:52
Here at Senturus, we offer 1/2 day workshop to help you understand the current state and the desired future state of your data governance initiative up to find the organizational goals and use cases and prioritize those.

53:04
You know, Greg emphasized this a lot earlier.

53:05
You really need to identify and prioritize.

53:08
You don’t want to boil the ocean, start small and work bigger from there.

53:14
Deliverable from that workshop is an executive readout of our findings and recommendations.

53:19
And really a big piece of this workshop is to help you identify, you know, what are the existing risks within your organization and really build out the business case and establish the ROI for pursuing data governments.

53:33
Little bit of just kind of information for everybody to be aware of.

53:37
Upcoming events and resources here at Senturus.

53:39
If you’re going to be in Las Vegas in about a month for the Microsoft Fabric Community Conference, stop by and visit us.

53:45
We will be there.

53:47
You can also find resources on our website related to data governance topics.

53:51
We’ve got a blog post on whether your data governance program will hold up in the era of AI, modern AI.

53:59
We also have an on-demand webinar available on Master Data Management.

54:02
That was another Greg Fresco presented webinar, so check that out.

54:06
You can see that on our website at Senturus.com/resources.

54:10
And we have a wide variety of other blogs, webinars and resources available at the same location.

54:19
Again, lots of stuff on the resource page.

54:21
I’m going to have Greg just slide right by this one because we’re short on time, but do visit Senturus.com stuff resources for lots of free info.

54:29
A little bit about us as a company, Senturus and what we do.

54:33
We are a data and analytics company obviously, but we actually have a full spectrum of analytic services and enablement in addition to proprietary, proprietary software solutions that help accelerate hybrid BI environments as well as assist with migrations between BI platforms.

54:52
We cover high level analytic strategy as we’ve been discussing here today, all the way down to the nuts and bolts of detailed implementation of your reporting environments.

55:03
We also provide support and training on all of our supported BI platforms.

55:09
And as I said, we have accelerator tools to assist with hybrid BI and migrations to new BI platforms.

55:17
We’ve been in.

55:18
We’ve been in business now for 24 years.

55:20
Hard to believe I haven’t been around for the whole 24 years, but I’ve been around for a lot of it.

55:26
We’ve got 1400 plus clients, over 3000 projects.

55:30
You probably recognize some of the logos there.

55:33
Some of you may even work for some of these companies.

55:36
So we’ve been very focused on analytics for a very long time.

55:40
We’ve got him that’s big enough to meet your needs, but we’re small enough to provide dedicated attention.

55:47
With that, let’s jump just quickly in the Q&A.

55:50
We just got a couple of minutes here.

55:52
There’s a question from Sri Raman.

55:55
So I’ll just give you the question, Greg.

55:58
And Sri Raman’s question is about how would you prioritize ADG program when really the bulk of the company’s resources and time are devoted to other areas like core system replacements, critical integrations?

56:12
Like what are some best practices and recommendations for ADG committee that’s trying to just kind of get started in an environment where most of the attention is on big system upgrades and maybe there’s not a lot of thought being given quite yet to the need for government?

56:30
Yeah, that’s actually a really good question.

56:33
I think a good approach would be to just integrate the thought and process of data governance into, you said an upgrade, I’ll take an upgrade or a migration probably, probably similar in nature in terms of, OK, what do we need to think about in terms of this upgrade or migration?

56:55
What are the data elements that are being considered and what do we need to consider now what can we?

57:00
It goes back to the garbage in, garbage out example earlier of, you know, if you upgrade a system, if you’re porting bad data from system A to system B or upgrading it from version A to version B, you’re not going to see a lot of the benefits of the new upgraded system, right?

57:23
So now is the time to mind your data.

57:27
Look at your look at your data, look at your information and make that leap now to curate your data better and present a better picture of your system to the organization once it’s rolled out, right?

57:41
Because a lot of times users won’t feel the they won’t feel the benefit of a new upgraded system if the data looks exactly like it did before, right?

57:51
They, they just, you know, your typical business end user who doesn’t live in this world that we live in isn’t going to be able to contextualize that this upgrade did is not supposed to fix my data.

58:06
The data needs to be fixed individually in and of itself, if that makes sense.

58:10
So really it’s, it’s actually the perfect time to leave in the need for government governance and the need to improve your, your data quality across the board, right?

58:20
So it’s a great catalyst for it overall.

58:25
Thanks, Greg.

58:26
One other quick question before we wrap up here.

58:28
I know we’re out of time.

58:30
You mentioned in one of the earlier slides you gave the example of this big retailer data governance solution and one of their, one of their early priorities after just establishing a program was data cataloging and developing a data dictionary.

58:46
Is there a typical kind of starting point for, you know, like does a, when a company undertake something like this it common to start like with a specific subject area or how do you sort of bite off pieces of that dictionary cataloging?

59:01
Well, yeah, this particular retailer, the catalyst for everything you said word to get was all around customer.

59:11
So it was customer acquisition, it was customer retention, it was customer loyalty, it was lifetime spend.

59:17
They want to know all those things, but they didn’t feel like the customer data was in the right position for them to do that.

59:26
And what that started doing was that sort of started expanding to like, OK, what products did they buy?

59:31
Well, we don’t have a good view of what products they bought because our SKU management system is messy right now.

59:38
So they had a really good use case to go in there, like, OK, we need to catalogue what we do have right now and understand what the full crux of the problem is before we understand how to fix it.

59:49
And those were essentially our recommendations of, you know, they wanted to be able to launch these, you know, hyper quick customer campaigns to their best customers.

59:59
But they kept missing, they kept targeting the wrong customers.

1:00:04
They kept targeting customers with the wrong information based on what they thought they bought and then kind of a little egg on their face.

1:00:11
So that was a really a big catalyst for.

1:00:14
For how like, OK, the first thing we need to do is catalogue customer demographics, anything we have about customer attributes and then their transactional buying behaviour.

1:00:24
And then as we did that, we came across like, OK, we have their buying behaviour, but we don’t exactly know what products they bought in good confidence because the ski management system was a bit faltered.

1:00:36
So we needed to go into that.

1:00:37
So that sort of just spider webs across the organization to the fact that like let’s just take it, it doesn’t need to be the most complete comprehensive inventory of data upfront.

1:00:49
You don’t want to get wrapped around the axle in that regard.

1:00:52
But let’s just take a quick inventory and cataloging of all the data we have through across all the domains and see where the main pain points are and can start to hammer away at fixing those.

1:01:03
So that was really the point.

1:01:05
But again that speaks to starting from an individual use case and I mentioned acquisition, loyalty retention, customer retention was where it started and that spanned into, you know, as you might imagine retention spans into loyalty expands into, OK, now that we have this, how do we acquire customers using these models etcetera.

1:01:25
So having all that information was really beneficial to them.

1:01:31
Thank you, Greg.

1:01:32
So we are past the hour, so go ahead and advance to the next slide.

1:01:36
And if we didn’t get to your question, as I said, we will reach out to you after the webinar.

1:01:41
So if we didn’t get to answer your question live here, you can expect to hear from us.

1:01:47
And with that, Greg, thank you.

1:01:49
Thank you so much for presenting today.

1:01:51
It’s great to have you there.

1:01:52
Lots of great info and thank you to everybody for attending.

1:01:57
Reach out to us by any means you prefer.

1:02:00
You know, we’ve got an 800 number still.

1:02:02
If you like to use telephones, you can also reach out to us. [email protected] and visit our website at www.senturus.com.

1:02:13
And with that, we’ll call it a wrap.

1:02:15
Thank you everybody for attending and we look forward to seeing you again on a future Senturus webinar.

1:02:21
Have a great day everybody.

Read more