Is Data a First Class Citizen in your Company? w/ Wendy Turner-Williams

Media Thumbnail
00:00
00:00
1x
  • 0.5
  • 1
  • 1.25
  • 1.5
  • 1.75
  • 2
This is a podcast episode titled, Is Data a First Class Citizen in your Company? w/ Wendy Turner-Williams. The summary for this episode is: <p>Everyone says they want to be a data-driven driven company. To do so, data must be a first class citizen, at the heart of all strategy and reporting to the CEO. Is that really happening at your company? Wendy Turner-Williams will discuss how data is at the center of AI, Ethics, Privacy, Security and more. </p>

Tim Gasper: Catalog and Cocktails, presented by data. world. It's your honest, no- BS, non- salesy conversation about enterprise data management with tasty, sometimes very colorful and fruity beverages in hand. My name is Tim Gasper, longtime data nerd, product guy, customer guy at data. world, joined by co- host Juan Sequeda.

Juan Sequeda: Hey Tim. I'm Juan Sequeda, principal scientist at data. world. And as always, it's Wednesday, middle of the week, end of the day, and time to take that break, have a nice drink and chat about data, data, data. And today, we have an exciting topic, exciting guest because we're really going to be hearing for the first time what our guest is going to be announcing. And I'm really, really excited to have Wendy Turner- Williams, who's the former CDAO of Tableau, and now is the CEO of The Association, which we'll get to hear much more about it soon. Wendy, how are you doing?

Wendy Turner-Williams: I am great. Like you said, it's hump day. I'm ready for a cocktail. I'm ready to talk data. So super excited to be a part of the podcast and to have this conversation. It's a perfect-

Juan Sequeda: Fantastic. So let's kick you off. What are we drinking? What are we toasting for today?

Wendy Turner-Williams: I'm toasting The Association and changing the world and driving data left as a first class citizen. What about you guys?

Juan Sequeda: Ooh, that's nice.

Tim Gasper: I will cheers to that as well. Data as a first class citizen. Sounds like you're doing your part. And I am drinking a cocktail called" It's Mint To Be." I asked for the hotel to give me something that was interesting. And" It's Mint To Be." I think it has Rosé and strawberry chunks-

Wendy Turner-Williams: It looks yummy. This is a Rosé champagne, A nice French Rosé champagne. Yes. What are you drinking, Juan?

Juan Sequeda: I made a margarita. I have some tequila, I had some inaudible liqueur, some lime, and that's it.

Wendy Turner-Williams: Yummy. Cheers.

Juan Sequeda: inaudible.

Tim Gasper: Cheers.

Juan Sequeda: inaudible to making data a first class citizen. Cheers.

Tim Gasper: That's right.

Juan Sequeda: So talking about first class citizens-

Wendy Turner-Williams: Tequila's my arch enemy, Juan. So we're glad I did not break out the tequila because there's no telling what would've happened on this show if inaudible.

Juan Sequeda: My wife is Mexican, so a lot of my life has now been very influenced by a lot of tequila, a lot of Mezcal, a lot of spicy food. That was not me 10 years ago.

Wendy Turner-Williams: What's funny is I loved it. It was my favorite... Tequila shots was my favorite thing in college and young. But as I've gotten older, and after I've had too many nights of" I'm not sure what happened" when the tequila comes out, it doesn't come out really anymore for me.

Tim Gasper: Yeah. I remember all the tequila and the lime and the salt. That was the go-to.

Juan Sequeda: Last episode, I was sipping on Mezcal, because I had a nice bottle of Mezcal, so I do like that one. All right. Well, talking about first class citizens, we got our warmup question today is, what are the first class citizen things in your day- to- day life?

Wendy Turner-Williams: Yeah. I think that's a really good one. I think being at the table, being a part of the conversation, being respected. I think there's a respect factor when there comes with first class citizenship. If you think about society or around the globe and what's perceived as first class or second class or third, I think there's a respect factor that comes with it. I think that investment comes with first class, right? Because normally, you're investing because again, there's a respect and epitome that comes with that label. And I think all too often, when it comes to data, which we're going to get into, don't always hit that first class bar, so far is the reality. What about you, Tim? What do you think first class is?

Tim Gasper: Well, you know what? I think you had some good thoughts there on what it means for something to be first class. And where it's really top of mind. Is it a focus, right? Is the only thing I'd add. One thing that I think about is, what's first class in my own life? What's the personal lens on first class? I know we're going to talk about first class today with data. I think about my mornings and how my morning is very precious. I want to have my coffee, and I want to be able to just meditate. I want to look at what my goals are for the day, and I'm hoping the kids don't wake up too early so I have my 30 minutes of calm and peace. So that's an example of something that's first class in my personal life.

Wendy Turner-Williams: No, I think that's a good example. What about you, Juan?

Juan Sequeda: So for me, the last couple years, something I've treated as first class is my health and working out. And it's finding that balance. Because it's not like I need to workout... I want to work out, but I also want to eat well. I want drink and inaudible. So it's finding that balance. So something that I did, was we started the podcast during the pandemic and once we were opening up a little bit, I would go to the gym, and it was my tradition. I would go to the gym every single day at 5: 00 or 6:00 and go to the gym after the podcast. And I'm still doing that, if I'm not traveling. So even though I have a drink, it is 4: 00 here, I am going to the gym at... We end up 5: 00. I have to go to the gym. We have class at 5:15. I inaudible my health and my gym and workout, first class right there.

Wendy Turner-Williams: Which is back to commitment or focus or investment, whether it's time and commitment. Same concepts, right? But really, it's about, I am deliberately, intentionally focusing on this particular area with investing, whether it's time, money, resources, whatever it may be. Even if it's your personal self. Making that commitment there. Yeah.

Tim Gasper: Exactly. Investment, commitment, intentionality.

Wendy Turner-Williams: Exactly.

Juan Sequeda: This is a good segue into... All right, Wendy, honest, no BS, what's up? What are you announcing here with The Association? Tell us more.

Wendy Turner-Williams: So I am so excited about this because frankly, I don't think anyone has done anything like this before. As a longtime data practitioner who has really focused on driving culture and driving change and driving collaboration across companies, I just announced last week the association. ai, now called The Association, where we're going to bring together practitioners, by practitioners for practitioners, in the AI, data, ethics, security, and privacy space, with the intention of A networking, like you would do in a normal business networking group. But also publishing within ourselves thought leadership opportunities, branding opportunities, job placement components. But more importantly, we actually want to create a cross discipline, cross industry governance channel that starts to define what ethical AI and data looks like in this next transformation, by practitioners. So if you think about this, I think that when people think about what's happening in the world with AI, lots of focus, lots of... Everyone's looking at how they invest, how do they define their strategy, et cetera. And the reality is that AI is not a standalone. I mean, data's the foundation for security, for privacy, for ethics, for AI, period. And each one of those things are almost like a domain area that attaches off of data. And we need to bring the practitioners together in regards to, how do we actually scale this thing? How do we actually share data across companies? How do we actually handle ethics? And it's just something that the government can't really regulate. And frankly, the big tech companies can't regulate, because they're the. 0000001% of all of the companies out there. So the goal is to give practitioners a voice for change, and to really help drive what's going to happen with the future of AI.

Tim Gasper: I love that. It is so true that folks are so excited about the change that's happening about AI and the impact that that can have. But not only is it confusing to figure out how we're going to handle that and what the right way to handle that are, it exposes the cracks in our foundation inaudible-

Wendy Turner-Williams: The skeletons, right?

Wendy Turner-Williams: That's right. I mean, there's a lot of skeletons

Wendy Turner-Williams: in the closet. I mean, there's been all types of statistics, reports, published by a lot of the research facilities. But I think one of the most common ones that I've heard that really make me open my eyes when it comes to AI was around data, where 92% of businesses failed to scale data analytics. And out of that 92%, 95% of the people in that 92% basically blamed culture. When you think about the business opportunity related to AI, and you think about the impact and the risk associated to trust or your brand or mission- critical release of secrets like Samsung or all types of things, if you can't scale your data, what makes you think you can scale AI, which is driven off of the data? Most people don't understand that AI is actually 100% dependent on the data. So if that's the failure rate, what's the failure rate going to be with AI? Probably the same, right?

Juan Sequeda: inaudible going on. I think now with the era of AI that we're in right now, everybody's thinking about AI. And then people are thinking about, okay, the data, the training data for the AI. But then it's like, wait, it's not there. You have to go back. How did the data come up to be, to create that training data to do that?

Wendy Turner-Williams: That's right.

Juan Sequeda: And there's all this lineage, all this governance behind it that is critical, and we're brushing that, " Oh, somebody else is doing that," or-

Wendy Turner-Williams: Sweeping it under the floor, just to say we have AI.

Juan Sequeda: We just wanted to jump into the AI thing. So this is the epitome of hype, get up all hyped, we're going to go down and things are going to fail and then we're going to go start, I mean, pointing fingers all over the place. But we know right now what we need to go do.

Wendy Turner-Williams: Well, and back to, again, that first class citizen mentality, digital transformation has always been about data. So I never loved when they turned it to digital, even though I know digital is all encompassing as far as the type of impacts. But what digital transformation really is, is about the automation around job processes and job flows and collections of data points that you can then start to share across the organizations and partners and business forms, et cetera. And to me, all AI is just the next crank of evolution on top of that transformation. It's still a part of digital transformation. Now we're just going into artificial intelligence to automate some of those workflows based on decision trees. But the key to all of it is the data underneath. If you're not collecting the right data sets, if you're not conforming the data in ways that remove bias or have quality, or again, if you don't know what's your data versus your customer's data, and it's just pumping through these models, then the amount of risk associated to AI is a lot higher than just flat digital transformation as it's been pre- generative to AI release.

Juan Sequeda: There's five things that you talk about here for the focus for the association. AI, data, ethics, privacy and security. So I'd love to, if you could, go through these and why are these the ones and how do those intersect?

Wendy Turner-Williams: Yeah. So when I think about these are all correlated, cross- dependent functions. So if you think about things like privacy, right? Privacy really is a domain of data. What type of data do you have that you're collecting where there is PII or personal information that you need to anonymize and collect and handle in a separate way? That's different than financial data, that's different than your operations data. It's just a data type to a degree, from my perspective. Security is really about, how do you protect that data from bad actors? How do you make sure that your data is treated as an asset, that it is protected in the right way and handled from a controls perspective to meet compliance bars related to that protection? So again, it's not about what's in it. It's about how do you protect it. When you think about things like ethics, ethics is just another use case scenario. How am I using this data, for what purposes? What's it fit for, and fit for intent? So to me, ethics is just another area that you are looking at intent, right? What's the ethical intent around that data? Are we using this to drive DEI? Are we removing bias so that there's no bias within it? What are we doing around those pieces? And back to AI, I don't really think of AI much differently than data, because to me, AI is a part of data, because machine learning has always been a part of data, frankly. And data science has always been a part of data. So to me, AI is really just an advanced data discipline when it comes to, how do you actually generate and make decisions? And whether it's artificial or whether it's through the models or what those things are. So they're all just intersecting components with data right in the middle of the target. You think about throwing darts. Data's in the heart, these are concentric rings around them and they all impact one another when it comes to policies or decisions or your customers.

Juan Sequeda: This is a really nice explanation inaudible. I ask myself, is there anything else? Is there anything else that could have been included here? Or why is it just these five? Or, inaudible the four.

Wendy Turner-Williams: Well, I think there could be more, right? There's always business functions as far as, how are you handling it for different types of wings of your business? There's always different types of industries. But to me, these are the disciplines and then how you spin them from an industry perspective, by your discipline to feed your industry, or your particular company as far as the use case, et cetera, or where you are from a maturity perspective, that's independent per company and per your industry. But to me, these are sister disciplines that are very, very much dependent on each other when it comes to actually using data. I'll give you an example, right? Privacy, GDPR. All of a sudden, GDPR starts launching, what is that? Seven, eight, maybe nine years ago, eight, nine years ago. And you see a ramp up all of a sudden of chief privacy officers. Okay? Never really existed before then. So some companies hired chief privacy officers. Some people handed it over to a CDO or to a CIO, you name it. But what their focus was on, how do we actually produce policy for how we're going to actually support GDPR, and how are we going to roll out those policies within our organization and make customers aware of how we use their personal data, and allow them to opt in, opt out? Well, from a data perspective, you had to operationalize those things. So again, from the people who are consuming and pulling data from the source systems and storing that data and curating that data and then supporting the analytics use cases or the hot path, real- time use cases or the support cases, et cetera, they still have to apply those principles into the data itself and make sure that the data is actually tagged appropriately to support things like right- to- be- forgotten or who owns it? And I'm software, right? That's mostly my background. So it gets very confusing between are you a controller? Are you a processor? Are you a sub- processor? So that goes back to data tagging. Well, there's no one better in regards to understanding the data being collected and how it's being used than the data people. And they have to operate it. So same thing with security. It's another example. I spent four years, I think it was, at Salesforce reporting to a CISO. And for me, it was eye- opening from many, many perspectives because A, we're opposites of a coin. I want to help everyone get the information they need and the data they need to do their jobs. And CISOs are very risk focused. Let's put everything in the closet until every component and thought process and control is in place. But the reality is, is that things like NIST, which everyone wants to have a lot of massive NIST uplift and they want to conform to NIST because that will meet a lot of their compliance boundaries, about 41% of it is an overlaying a dependency with data management maturity. So if you focus on the data, and your job is about protecting the data, well the data exists there first for you to protect it, you already know how people are using it and for what purposes. So things even like classification. How do most security teams classify the data if they don't know how the data's used or the risk associated to that data? If it's your high emission critical business value or if it's your financials, et cetera, they depend on the data teams. So that's why I selected those five, is that to me, those that's the sisterhood, or that's the band of brothers in regards to true cross dependency in relation to AI success and trust and ethics, and how do we implement those things?

Tim Gasper: I love what you're walking through as an example here, Wendy, and I think that you're your example of classification actually highlights some of the problem that I think we face here, which is that you have the CISO, you have the head of privacy, you've got the CDA, CDO, and maybe you even have a split between governance and the analytics office and maybe, every single business unit has their own analytics office with their own principles and things like that. There's these silos that form in our organizations, both from an org structure standpoint, but also from a discipline standpoint. And you're talking about brotherhood and sisterhood and I'm curious about, how do we get there? I love the five things that you've noted, but I think in many companies, they're handled in a very disparate and disconnected way.

Wendy Turner-Williams: They are. And that's part of the means for bringing together the practitioners, is to share ideas of success, et cetera. But for me in particular, I always make an approach of... I love hub and spoke models. I'm a big fan of self- service. I just don't think data can be centralized at most companies, depending on the size and scale of the company. And like you said, there's different specialty areas by business or even by a function, you name it. But the reality is that certain things like data management or your data strategy has to be a shared component. It has to be embedded in your culture. Otherwise, what you're doing is that your operational costs and your funding costs associated to those operational costs, whether it's hardware or the software, and then your risk for actually using that data to drive your overall corporate strategies gets lost. So to me, we're going to talk about the honest, no- BS, I think the reality back to that data culture, most people don't understand what data culture is. Is it having data? That's not data culture. Just having data just to have data. What you want to have is high quality business decisioning data that actually aligns to the strategies that the company is focused on. And enabling the people to have those strategies. Now what that needs to do from a literacy perspective in order to be effective for people is that you've got to break down the business funnels and actually help people understand how the business is interrelated. So something like, I've got a lead and I've got an opportunity and a lead to a deal, to a renewal, to your product and how that flows in. Each one of these business teams lives in their little silos with their blinders on. But the reality is that the data that they collect drives downstream processes. And so starting to look at strategy in a different way and starting to remove those silos to a mentality that we're all on the same ship, because your corporate success is yours is my success and yours is my success. And it's all of our success if we're looking at a corporate perspective, is what needs to happen. But to do that, we got to shift to first class citizenship.

Juan Sequeda: It's a standing ovation here. Actually Tim, back channel me. She is singing your song. You can see my smile, for those who are just listening to us. This is what I call the time that the data teams need to have business literacy. Literally what you just described. You need to understand, first of all, what are the different business functions? How do they operate together? There is a lead. After a lead, you have this thing goes to this and blah blah. It's called this thing. And they use new systems and so forth. And by the way, this is the objectives and the strategy of the company. This is how we make money. If we don't get this right, we lose money. Anyways.

Wendy Turner-Williams: Let me give you a real life example. Okay? I'll give you a real life example that I dealt with years ago. This was pre- Salesforce. People can look up where I was at. But anyway, the big new area for a major software company was all about usage. Driving usage, driving usage. And I set the team that was responsible for shared metrics and KPIs, and we were having quality problems with our usage metrics. Well, digging in, it was my job to dig in and figure out what was going on in my teams. Basically there was a disconnect in the business processes between what we sold and the billing related to it. So think the marketing, the sales, how we sold, and then the commerce platforms, how we build, to where A, you didn't always bill based for usage. Sometimes you gave free trials, sometimes people had an incident or something and they got some type of credit, whatever those things were. But the reality underneath it was that we had a big master data mess where about 18 years of tech debt related to products and then services as they moved from box products into cloud services was basically creating a lot of different lack of conformity in regards to the roll- ups of the data. And what would happen is that as new products were rolling out for software, the marketing team and the sales team were having conversations about what they're going to call the product about, where they're going to market it, about what the SKUs were and what you got. But on the commerce side, what would happen is that you had basically operational people who would free type into the commerce platform meters. And the meters would spend to generate the usage. So they would do this sometimes months in advance of the product launch. No one would ever go tell the operational team that, " Hey, you had FU in as the product? Because FU is the project name because it hasn't even gone live. And we actually are launching it as, whatever, X, so go update this." So you had some products that had split lines of usage into 20 different meter names. Massive confusing to customers in regards to billing and what they signed up for. Massive confusion internally in regards to how much usage did this spend? Or is it available? Or what's the customer journey and are the features working? Because it was tribal knowledge in regards to, how do you aggregate these things up? Nonetheless, then look at incentive comp for yourself and your partners or financial forecast and how do you do that? Something so simple as just the collection of usage and how you name it was a major thing that impacted every single BU in the company. We literally fixed that problem within I think three days, and restated and got to zero unknown usage. Then drove the largest commerce platform cleanup in history for that company, which then led to a new product called the billed usage API that now they sell to customers based on that. And all of that went to a master data system that I built years ago that now the product teams own and the operational teams consume what product says so there's no in- between and it's an automated process. But that's what I'm talking about. People don't understand who's impacting who. Or another spinoff off of this, one more thing, there was a tax team that if you go to... This was Azure, okay? I'll just call it out. But if you go to azure.com and you look at the products and the SKUs and you get to the rate calculators based on geographies, there was a whole team that would spend two weeks out of a month manually inputting rates by geographies to do an update. We just extended that off of a dimension. We automated the inaudible process and we just consumed it from a source because there was no need to physically type in, anyway. So we gave tons of time back to independent teams. But the reality is people don't understand what they do and how it impacts the downstream teams from a business perspective. And the data that triggers around all those processes. And that's where literacy is. I mean, literacy... Tableau was a major focus on literacy. There's a lot of talk about literacy. And I have my own definition because literacy is not learning SQL. Literacy is learning the tools that your company specifically uses and understanding the business systems and the business strategies that your company is currently operating against and knowing how to ask the right questions, and then how to answer them with the right tools. You know what I mean? And to me, that's literacy.

Juan Sequeda: So right now, you've given some fantastic definitions on what culture is and literacy. So we need to pull that out as good snippets for later on. So what does it mean, or what would be the ideal scenario or example... Ideal company set up where data is a first class citizen? And I postulate, does that mean that we have a data office, a CDO that... Who reports to the CDO or CDAO and where does the CDO report into? How would that look like?

Wendy Turner-Williams: Yeah. So first I'll caveat that I don't know that there's a perfect definition for all, right? I mean, you've got different companies at different sizes with different amount of data volume. So I don't think that there's a one- for- all answer. I would say though that in my world, right, in the software world, look at most of the big tech companies. So look at Microsoft, look at AWS, look at Google. You don't have CDOs today. Okay? Number one, that should raise a lot of eyebrows around that. But to me, the CDO is the CEO's best friend. You're defining strategies. You've got a COO who needs to operate. The COO needs the data. They're a customer of the data. You've got the financial team who's looking at, how do we actually get more revenue, how do we reduce our cost savings, et cetera? They've got their specific business strategies. What the CDAO does is cut all of those different business strategies and brings the data to life underneath it to support them all. Right? So they're this cross- functional component, not the CEO, or the CIO, I'm sorry, who tends to focus on, what's the containers? What's the containers? This is the people who focus on, what's in the containers, and why do we need it, or why do we not? Why do we not need it sometimes? You know what I mean? From a trust perspective or a usage perspective. But that's how you bring your strategies to life, is by enabling the insights with clear business decisioning capabilities, which means you're you've got to be at the C- suite label. You've got to be at the table while they're having business conversations. You have to align your strategy across that as a big picture perspective. And you've got to create space for the CDOs to operate, right? With dedicated funding and others to support those things. Otherwise, what you're doing is you're just not using your data to really amplify what could be across your organizations or where there's new opportunities or where there's upsells across products or where there's a new cost savings that weren't identified or realized. You're really leaving a lot of your data value, and frankly, your ROI and your infrastructure investments at the table. You're not realizing that.

Tim Gasper: For companies that aren't doing this well, whether it's because they don't have a CDAO, or they don't have the right strategy or the right approach in place, what are your recommendations on how to get there? Does it start with saying, " Well, you do need a CDAO, and you got to empower them with the strategy capabilities and let them build that relationship with the CEO," or do you approach it a little bit differently?

Wendy Turner-Williams: Yeah, I think there's two ways to go about it. A, I think again, if you're a mid- size to large company, I think you need a dedicated CDAO, right? But I always start with, personally, I start with a data management maturity type of assessment. So whether you're using the cloud data management maturity assessment or you're using DCAM or... Gardner's got one. I use an industry framework to bring in third parties. So it's not perceived as a personal assault by the CEOs. Data's political. Again, let's not shit about this. Data's political. So again, you bring in a third party and you do assessments of each and every organization and their data teams and those core five disciplines as well in regards to how you're operationalizing data and the policies around those things. And you start there. That should start to define where you need to focus. So an example, when I joined Salesforce, for example, they had never had a centralized data team. They hadn't even had a data policy when I joined. Okay? So the first thing I did in the first six months was establish a data management maturity assessment to go in and identify where our low hanging fruit was, what our strategy would be the first year, over the next two years, three years, four years, and how we were going to drive that maturity. And things like platforms, things like data governance, things like catalogs, things like glossaries, even your strategy. Most of those things are baked into those assessments. So those assessments will point you exactly at where you start, and how do you start, and what's the best return on your investment for that start? And that's always where I start. That's ground zero.

Juan Sequeda: I'm curious here from a reporting perspective. So we've been doing these honest, no- BS centers. You've been part of one of them in Seattle. I mean, I actually to talk to people like, " Hey, who do you report to?" And usually people end up reporting directly to the CIO, right? Or the TSO. Where do you see that the data office usually starts out reporting? And I know you said they should end up reporting to the CEO. What are you seeing? And to somebody listening right now, what is a red flag? Like, " Oh, crap, I need to figure out how to get out of the mess I'm in right now."

Wendy Turner-Williams: Yeah. I think reporting to the CIO is normal. I think that's where people usually start. I've also seen two other flavors of this, reporting to the CFO and reporting to the CMO. And really, it depends on the company and what their number one strategies are. Right? I've seen companies who, they're all in, go to market, and growth, growth, growth, growth, growth. So they really want to focus on that marketing angle, really the monetization, how do they go? And so they tend to park it under a CMO. I've seen people who, again, they're more concerned about the financial aspects or the governance components related to SOX or whatever it is, and then they'll park it under the CFO. Which I actually... I've been under the CFO many times and there was value in it. I personally don't mind being under a CFO because all roads lead back to finance. And so being there is not a bad place to be in regards to funding or driving change. But they're very siloed in how they think about data from their purpose. But the norm is usually the CIO. And I think that CDOs.... It really depends on the company. It depends on the relationship between the CDO and the CIO. But what tends to happen is that the CIOs are focused on infrastructure transformation, and that's what they get funded on. That's what they get bonuses on. That's what they get their board discussions on. And what tends to happen is that what's within that infrastructure just becomes not a priority. So you're getting the crumbs, right? Related to the funding or the tooling, et cetera. Even though the whole point of the infrastructure is to get to the data. It's the entire point. So again, it's this lopsided view because people just don't understand. Data is so complex. I mean, data's big and it's complex and every org has it. And do you own financial data or does finance own it? Well, if finance own it, then do I really need a CDO if finance owns it and marketing owns it and sells- owns their pieces? Because then what's the point of the CDO? Well, the CDO point is to be a servant, right? You're an enabler servant to make sure those functions are successful because they aren't data experts. They're business functions. And they have business strategies that need that data input in order to shift change, be agile, or make sure their strategy's even on right. You know what I mean? So it's really this discipline that people have not realized is distinct. It's a distinct discipline in itself.

Tim Gasper: I think this is some really good advice around thinking about how to set a strategy, how to think about your organization, how to make an impact. Even if you're starting in a different part of the organization, which I think is very helpful. I'm curious about... Going back to the main title of our session today around first class. If you were going to give some sort of final advice or action steps for listeners on how they can make data a little more first class in their organization, take some good positive steps there, where would you direct them?

Wendy Turner-Williams: I would direct them to back to the data management maturity type of frameworks, to educate yourself on them. I would also direct them to understanding their internal org structure. Data success is all about networking, in my opinion. You've got to know who the players are. You have to know their strategies. You have to be creating opportunities to basically get involved and do work if they're not currently supporting you, and to drive change and drive impact. I think that piece around networking is one of the biggest components around, how do you build success and how do you create a story? Who do you influence? Who gets it and who doesn't, right? And who influences them? So that you can actually... You don't want to push a rock up a hill all the time. Eventually you want that energy to where others are helping to advocate and drive that energy and the hill's just starting to roll down in regards to change and the work. The other thing that I would do is... So part of the reason why I called out the five teams, one of the biggest things at Salesforce that was beneficial for me to set up an enterprise function was aligning it to trust. Okay? Companies have values. Most companies have some type of value. What are their values, and how does data help enable their values? Is it agility? Is it innovation? Is it trust? Is it customer success? And then start to understand, again, what those business strategies are and initiatives, and try to understand how data has a role and how you can help others. Right? Back to being that servant. You're a change agent, but we don't own the business functions. We support the business functions and bring them to reality in regards to the change. And there's an art to that. So back to, know who the players are, know what the initiatives are, and have conversations. Create space. Figure out where the fans are. Get involved. Build community internally. One of the things we did at Salesforce is we created Data At Salesforce. It was one of the biggest things I did, which was a forum. So it was like a Slack community in which all of those desperate data teams and those silos and organizations were all there. But we also then started to use that as our marketplace where you could find policies and you can find literacy, and you can find our platforms, and you can find our glossaries. And so it started to ingrain it and became a tooling aspect in regards to any new hire, if you're doing data work, you just got to Data At Salesforce, and everything's built right off of it.

Tim Gasper: Where did you say you built that into? You built it into the community itself?

Wendy Turner-Williams: Into Salesforce.

Tim Gasper: Into Salesforce?

Wendy Turner-Williams: Yeah. Yeah. So if you went to Salesforce and you go to Aloha, which is their Salesforce homepage where you see your paychecks or whatever, that's where Data At Salesforce is. And it's a marketplace, not just for tools, but for exchanging ideas, seeing stewards, seeing lineage, doing whatever. Make data intuitive. Make it task oriented. Don't remove the abstraction, and make sure it is something that is easy to digest, easy to deal with, and pick the right projects to show value. And quantify, quantify, quantify what you're doing. Right?

Tim Gasper: That's good advice here.

Juan Sequeda: There is so many great nuggets... There's so many just saying I want to put on a T- shirt. Tim, we got to start this... A T- shirt store off of-

Wendy Turner-Williams: We do. We should. We should.

Juan Sequeda: Yeah, we should.

Tim Gasper: Sure. inaudible Wendy T- shirts? There's a lot of good quotes here.

Wendy Turner-Williams: What's my startup thing? I have one. We're shifting data talkers to data walkers. Back to the no bullshit. I get so tired of talk about data when it's not the first class citizen at most places. Build the community, invest in the community, invest in your employees, invest in the output that comes out of your infrastructure, and invest in your strategies. And that's about shifting data to the left. And that's about creating data as a first class citizen. Data walkers. We want data walkers. We don't want data talkers.

Juan Sequeda: In other words, no bullshit. We just want honest, no BS.

Wendy Turner-Williams: Yeah.

Juan Sequeda: inaudible. We're going to do our AI minute. You got one minute to rant about AI, whatever you want. Ready, set, go.

Wendy Turner-Williams: Oh, okay. So my rant is that AI is not new, one. That AI is all about data, and it's the next gen of basically data transformation. That AI is not here to steal all of your jobs. You should be learning how to use AI because it will actually enable you to do new things and focus less on the operations. But AI needs to have guardrails and needs to have high trust in quality and ethics. And that we need a community around AI. So we need AI deputies. There's no one sheriff. We're all practitioner deputies for AI. I shot the sheriff, but I didn't shoot all the deputies because we're the ones that have to actually deploy this crap.

Juan Sequeda: I love this. I love this so much. All right. Well, all right. Let's go to our lightning round questions here. I'll kick it off here. So can there be big AI impacts on your business without investing on the core data foundation?

Wendy Turner-Williams: I think that's a good question. Of course, I think there can be. But I think you need to weigh what's the impact versus what's the risk if you get it wrong. So again, to me, there's a quality concern. There's that bias or ethical concern. Can you just run it and see what it does and operationalize some components or get some operational efficiencies? Of course you can. But again, you're trading off. Potentially, you're trading off your brand and your market as a result.

Tim Gasper: That's really good advice. There's trade- offs here in everything that you're deciding around this. And different businesses maybe have different risk profiles.

Wendy Turner-Williams: That's true.

Tim Gasper: Second question. So does the chief data officer or chief data and analytics officer own data ethics?

Wendy Turner-Williams: I think so. I think there's a difference between the ethics in regards to policy and there's a lot of ethical officers that are spinning up. But I think that there is no point of ethics if you can't apply it in the data as far as how you can use it for fit for purpose. And that requires the CDAOs.

Juan Sequeda: All right. Next question. Will the excitement about AI accelerate progress around data ethics and governance? Or are we just going to be heading into some bubble burst winter, inaudible?

Wendy Turner-Williams: Well, I think that if you heard Microsoft last week when they did their, or two weeks ago, when they did their earnings, I mean, Satya was like, " This is all about data. Data's everything." You know what I mean? I just think that most people don't take that to heart. And I think GDPR was proof of that. I thought GDPR would've sped up the focus on data and the governance, and it did for a window, until GDPR got deployed at companies and then it all just went back to the norm and they only focused on that particular domain. So smart folks, it should. I hope so. But proof in the pudding with GDPR says probably not.

Tim Gasper: That's a good point around GDPR. Obviously that's still a big motivating factor for many organizations. But the initial burst of activity that was much broader has subsided into something a little bit more niche.

Wendy Turner-Williams: That's right. They treat it like a project when data's not one and done.

Tim Gasper: Yeah. Interesting. All right, fourth and final lightning round question. We're actually going to do this a little different. This is a twist. I call it good term, bad term. So I'm going to give you three terms. And I'm curious, good term or bad term. Data literacy-

Wendy Turner-Williams: Okay. Do I just say what I think about it? That it's good or bad?

Tim Gasper: Yeah. What do you think about it? Good term or bad term. Data literacy, data- driven, data culture.

Wendy Turner-Williams: Data literacy, good term. But again, I think that people don't understand what it really means. You got to understand the business and the flow. It's not just about the tools. Data- driven, I hate that term, because everybody uses it. And the reality is that most of them aren't data- driven. You know what I mean? And what was the other one? Data culture?

Tim Gasper: Data culture.

Wendy Turner-Williams: So again, I think that that is... Why do we have to define culture? Where do we have anything else defining culture at most corporations, right? I think that it's almost become this blanket bandaid that just because you have data that now you've got a data culture when you just have a culture. And the culture should be about growing together and collaborating and being informed. And why does that mean that your data culture? You're just a culture, right? For your company. inaudible-

Juan Sequeda: That's an honest, no- BS take. I love that.

Tim Gasper: We don't talk about marketing culture. I mean, I guess you could. But it would be weird, right?

Wendy Turner-Williams: You don't with anything. I don't know of any other culture things that people call out when it comes to inaudible.

Juan Sequeda: There is a sales culture. That one I would argue there is.

Wendy Turner-Williams: Well, but they don't say it when they're in-

Juan Sequeda: That is true.

Wendy Turner-Williams: When you're in inaudible-

Tim Gasper: Sales culture or something like that.

Wendy Turner-Williams: Well, like Salesforce. Salesforce is a sales and marketing culture internally. But they don't say that when they're talking to their employees. They don't say that when they're talking to their customers. They talk about how they have a data culture and that you need a data culture. To me, data culture's become a marketing tag, and that's all it is. And there shouldn't be any definition around culture. You have your culture. Is your culture to grow your company and to operate effectively and to remove silos and understand your business processes and to make sure that you actually have the best thing for your customers, or is it not? If it is then you need data. You know what I mean? And that's it.

Tim Gasper: Well said.

Juan Sequeda: Wow. Perfect. Love this. This is the hot takes we wanted. All right. Takeaway time. Tim, take us off. Takeaways from Tim.

Tim Gasper: Time for our takeaways. So Wendy, amazing conversation today. You started off by announcing The Association, at theassociation. ai.

Wendy Turner-Williams: Look online.

Tim Gasper: Which is awesome. Very exciting. Bringing together practitioners for networking, for thought leadership, job placements, but most importantly cross- discipline, cross- industry conversations, collaboration around especially the areas of AI, data, ethics, privacy and security. And very exciting. I think we need that community badly. And you mentioned that there's a lot of motivating factors here. The rise and the excitement around AI is just one of those things that's a big driver here. But another, I think, really important thing you mentioned is that you threw some stats out there. 92% of businesses failed to scale data and analytics, 95% of the 92% blamed culture. That seems like a little bit of a scapegoat here. There's more under the hood. If you can't scale data, you can't scale AI. And there's work to do here, and we can do it together as a community.

Wendy Turner-Williams: Totally. Totally. Totally. I mean, we're the ones on the ground. So the more that we help each other, the better. The more you can amplify your impact internally.

Tim Gasper: Yep, exactly. And you went into a little bit today around these different five areas, these five sisters or brothers, five siblings. They're all correlated functions. And so privacy, it's more around the type of data. Security, around how you're protecting from bad actors and meeting compliance so it's protected. Ethics. What is the data fit for? What's the intent? And are we really doing the right thing by our business, by our customers, and by the broader stakeholders? AI is that more advanced data discipline. And then data is the connective tissue. It's right in the middle. It's the heart, as you mentioned. And use cases cut across all these things. Classification was an example you gave where it really affects all these different five aspects. They have to work in concert with each other. And whether you're centralized or you're decentralized, or you mentioned hub and spoke model, things like that, that are really effective for larger companies, ultimately, your strategy does need to be unified, and it needs to bridge across these five disciplines if you're going to create a cohesive approach to be able to handle those use cases.

Wendy Turner-Williams: That's right. That's right. That's a good take down.

Tim Gasper: And then lastly, and then I'll pass it over to you, Juan, the reality is that most people don't understand what data culture is. And I thought that that was really well said. And you really tied it back to, you need to understand the business, and you need to understand that literacy is not just about learning SQL, for example, or learning Python, as another thing that I sometimes hear people throw around. It's about learning the business, the business strategies, and knowing how to ask the right questions and then the right answers, how to get to the right answers using the tools, using the technology. But that's the last part. The other things have to come before it. That is data literacy.

Wendy Turner-Williams: That's right.

Tim Gasper: Yeah. So Juan, what about you?

Juan Sequeda: We discussed about what would an ideal org structure look like and where does data fit in? And then very clearly, there's not one size fits all around here. And you brought up inaudible, if you look at these big software tech companies, the big tech, that they don't have CDOs. So that should raise a lot of eyebrows right there. Now, a CDAO should be the CEO's best friend. The CEO is the number one customer for the data. It's really a cross- functional component that supports all the different pillars. And the goal here is to provide clear business decision capabilities and to have a seat at the table to enable that. Now, how do we get started? This is a phrase I want on a T- shirt. Data is political. And this is the honest, no- BS that we have to acknowledge and accept it. It's political and you have to play that game. So how do you get started? First, use industry frameworks to understand the maturity of your organization. And actually have third parties, because if you don't have that third party, people are going to think, " Oh, it's you pushing that and turning into politics right there." Make assessments of each business unit. Understand how the lay of the land and these frameworks really help you define, what is your strategy, what is the low hanging fruit? Maybe it's simply creating your first data policy or whatever. This is what you need to understand. And talk about reporting. It's really normal for the data teams, the data office to report it to the CIO. It's where where you start. It's really focused on infrastructure and transformation. But you can maybe report it to the CMO, depending on the strategy. If it's in a very growth stage, that's why you may report to the CMO. Reporting to the CFO is not bad because hey, all roads lead back to finance at the end. But the ultimate goal here is to really be the best friend to the CEO. I think that's a sign of success here. Now we ended up asking you advice to turn data into first class citizens, and so inaudible. Data maturity frameworks, educate yourself. Understand your internal org structure. Network with everybody there. Understand their strategies. Go create opportunities. Who do you influence? Who gets it? Who doesn't? What are the values of the organization? And how does data help to enable those values? How can data help others? Because again, data supports the business functions. Find the fans. Build community internally. Make data intuitive. Make it task oriented. Remove the abstractions. Make sure it's easy to digest. Pick the right products to show value. Invest in the strategies. Shift data to the left. Quantify, quantify, quantify. And I want to close out with yours. We're shifting from data talkers to data walkers.

Wendy Turner-Williams: Yep. Yep. What a world we live in if that becomes a reality. I'm excited for it. We're going to do it.

Tim Gasper: Really. Bring it on.

Juan Sequeda: How did we do? Anything else we should... Anything we missed in our takeaways?

Wendy Turner-Williams: No, I think that was a great recap.

Juan Sequeda: Awesome. So we'll throw it to you to wrap up. Three questions. What's your advice? Who should we invite next? And what resources do you follow?

Wendy Turner-Williams: Okay, what's my advice? So my advice is, join The Association because we're going to change the world. Number two. What was number two again? Who should you have on?

Juan Sequeda: Who should we invite next?

Wendy Turner-Williams: You know I'm about to talk to Moham Araf. I think that people need to hear about the opportunities that could happen when you bring business semantics right on top of stores. Okay? And what changes that can drive.

Juan Sequeda: Yeah, we're going to be definitely talking a lot about AI and semantics and knowledge graph. And actually, there's a call- out to a previous episode, almost a year ago, with Bob Muglia, the former CEO of Snowflake. And he was on the board of Relational AI. This is what I tell people. If the former CEO of Snowflake, the world's largest cloud data warehouse, is talking about semantics and knowledge graphs and relational knowledge graphs, you got to pay attention to that.

Wendy Turner-Williams: Well, just to give everyone a little tidbit, I mean, what's the point of a lot of business applications if you can define the business logic and add it right into the stores with knowledge graphs and decision trees? So a little tidbit about where things are headed, right? And Bob, as you know, is on the board with Moham. But it's an interesting topic and I think it's definitely a direction for the future.

Juan Sequeda: And one of the things that we'll be talking a lot more about here, because this is where we're seeing about AI, is the combination of AI and all these large language models with the knowledge graph, the semantics, the business, the context, less up.

Wendy Turner-Williams: You want to talk about automated of business decisioning, right? For operations, there you go.

Juan Sequeda: So much inaudible. All right. And finally, what are your resources that you follow? People, blogs, podcasts, books and conferences, whatever.

Wendy Turner-Williams: I tend to follow a lot of the standards. I'm very active in a lot of the DEI type of efforts. So women in data, women in technology. I do a lot of the Gardner, McKinsey publications. I'm not a huge podcast person. I hate to say that. But I'm not. I'm just too busy.

Tim Gasper: No offense.

Wendy Turner-Williams: inaudible. But I think that I... Learn from the people who interest you. We all have different things that we're interested in. I like big thinkers. I like role changing types of concepts, and those are the people I tend to follow.

Juan Sequeda: Fantastic. All right, well just a reminder, next week we're going to have Simone Steele talking about data sustainability. Simone, she gave a fantastic talk at the CDOIQ conference, and she's right now a free agent. She's a very experienced CDAO and she's taking a break right now before she's going off to her next gig,. So she has the opportunity to be very open about all her thoughts. So that's going to be a very thought- provoking episode next week. And with that, Wendy, thank you so much. And I want to remind everybody, check out theassociation. ai. And as always, thanks to data.world who lets us do this every Wednesday.

Wendy Turner-Williams: Thank you guys for having me. Cheers again. And congrats on this launch. And congrats for everyone who joined us live. Not that it's congrats. But thank you for joining us live. And I can't wait to see where this goes in the future.

Tim Gasper: Awesome.

Juan Sequeda: Cheers.

Tim Gasper: inaudible. Cheers.

Wendy Turner-Williams: All right. Thank you, guys. Bye-bye.

DESCRIPTION

Everyone says they want to be a data-driven driven company. To do so, data must be a first class citizen, at the heart of all strategy and reporting to the CEO. Is that really happening at your company? Wendy Turner-Williams will discuss how data is at the center of AI, Ethics, Privacy, Security and more.

Today's Host

Guest Thumbnail

Tim Gasper

|VP of Product, data.world
Guest Thumbnail

Juan Sequeda

|Principal Scientist & Head of AI Lab, data.world

Today's Guests

Guest Thumbnail

Wendy Turner-Williams

|CEO @ Culstrata-ai