As CDO at Tableau, Wendy Turner-Williams manages enterprise data strategy, data platforms and services, data governance and management maturity, data risk, and data literacy. Wendy and her team help to fuel data-driven business innovation, transformation, and operational excellence across all areas of Tableau’s business.
As the Co-founder and CEO of Alation, Satyen lives his passion of empowering a curious and rational world by fundamentally improving the way data consumers, creators, and stewards find, understand, and trust data. Industry insiders call him a visionary entrepreneur. Those who meet him call him warm and down-to-earth. His kids call him “Dad.”
Wendy Turner-Williams: (00:00) I am someone who, as soon as I join a team and every day, I have the org chart up every day, every moment it’s always on my laptop. And I’m always trying to understand: who’s this person and they roll up to what org, and they do what, and who else do I know there? And wait, I think I remember this other team was doing something similar, and who was that again? And let me connect them.
Satyen Sangani: (00:24) That was Wendy Turner-Williams, Chief Data Officer at Tableau. Salesforce acquired Tableau in 2019 for $15 billion. At the time, Wendy was Vice President of Data Governance for Salesforce, already a massive, sprawling multi-billion-dollar software company.
Before working at Salesforce, she was at Microsoft on the Cloud and Enterprise Analytics Management Team. I can think of few people that have such deep experience at the intersection of software and data. Today, Wendy’s goal is to integrate Tableau into every part of Salesforce and its culture.
That takes more than technical knowhow — it requires incredible political savvy. Like a politician, you have to figure out how to create a coalition across the enterprise that will work together for a common goal.
And if you think data is complicated, people are far more complex. Wendy doesn’t fight that reality. She works with it to achieve her goals. Wendy and I discuss how data is political, because it’s just a reflection of people and their power. We also discuss tactics for turning that dynamic to your advantage.
Producer Read: (01:41) Welcome to Data Radicals, a show about the people who use data to see things that nobody else can. This episode features an interview with Wendy Turner-Williams, Chief Data Officer of Tableau. In this episode, she and Satyen discuss the relationship between data and strategy, the importance of personal relationships, roadblocks for building a data culture, and so much more. This podcast is brought to you by Alation. Data citizens love Alation because it surfaces the best data, queries and expertise instantly. The result? Folks now know how to use the most powerful data with guidance from the experts. And with Alation, you don’t have to choose between data democratization and governance. By embedding governance guidance into workflows, Alation welcomes more people to great data…fast. That means your data strategy can play both offense and defense. Learn more about Alation at alation.com.
Satyen Sangani: (02:37) Tell us about what it’s like to work on data inside of a software company, where you’re working on data, but also responsible for producing software to manage data.
Wendy Turner-Williams: (02:45) I think that that is a very interesting role to be in when it comes to being a data professional within a software company that produces data. The reason behind that is because A, there’s a reality of the complexities related to data, meaning quality, or latency issues, or integration gaps, or even sometimes our infrastructure when it comes to supporting kind of enterprise cross-scenarios. Or even the complexities of the different org structures and the business units, and their priorities that come into the play, that often when you’re developing the software, you’re thinking in bits or features and you don’t always have your fingers on the nuances and the turn in the system around the soft points that come with data.
When it comes to software engineering, software engineers don’t necessarily always use their software nor, again, understand these competing types of priorities or conflict points or touch points that actually kind of drive a lot of influence into the actual internal data processes as far as actually impact of scalability or delivery.
Wendy Turner-Williams: (03:56) So for me, I try to really center myself, and present myself as almost like a customer zero. I’m like, if we can develop software that meets my needs and my needs are holistic and cross-enterprise in nature, and I’m looking at different analytics approaches, platform approaches, data management approaches, advanced analytics approaches, you name it. And if we can create software that makes my life easier, and I can provide that feedback loop as far as the experience or the features or the use cases that are gaps, then I try to do so because I feel like it’s my responsibility, being in software, to help make that experience better for all.
Satyen Sangani: (04:40) I asked Wendy, what strategy means to her as CDO and how data informs her strategy at Tableau.
Wendy Turner-Williams: (04:46) Strategy is something that is a community thing, meaning there’s different pieces that make up a holistic strategy for a company. Meaning there could be financial avenues, there could be emerging market avenues, there could be product and product features avenues, or scalability of infrastructure or your workforce. You name it. There’s different things that all make up part of a strategy for you as a company.
And to me, the role of the CDO is somewhat like a, it’s kind of like a glue mechanism. It’s a seat at the table. It’s a seat at the table as each one of these strategies are discussed, who’s basically trying to define and decide what data is needed in order to support those strategies in a proactive way, as well as the infrastructure or the governance or the data handling, other different aspects that are also needed to do that.
Wendy Turner-Williams: (05:38) So to me, it’s again, how do you actually make that strategy come to life in a way that you can measure that business value in the ROI, in return toward that strategy in regard to the investment that happens.
So again, sometimes I find it really interesting when there’s certain companies who make blanket statements around “I have done X,” but how do you know you’ve done X? Did you measure that you did X? And when did you do X? And there’s different pieces that come to that question in regard to was X actually done, or the definition of “done,” or did it actually land the way that you expected it to do so? It’s kind of an interesting thing where data really plays this underlying, almost like a foundational role in regards to bringing strategy to life through questions. And how do you measure these things?
Satyen Sangani: (06:32) One of the hard challenges in talking to so many chief data officers is that often with data, you can be in this place where you go in 15 different directions. You can try to document every source. You can try to get all of the lineage of all of the data. You can try to cover compliance use cases and offensive use cases.
And it’s super easy to get lost, but this idea of like, “Well, look, my strategy as a chief data officer is to just follow the company strategy,” and to enable that could guide or could be a mechanism to guide so much of the investment dollar and so much of the bridge to the business. And I think that’s a really interesting observation that I don’t think happens as often as it ought to.
Wendy Turner-Williams: (07:17) I really think about data and chief data officers as they’re like a supporting actor, to the corporate entity as a whole. And you’ve got to have those partnerships. You have to have those breakout moments where you’re working in unison with these particular business units who are executing on these strategies all up. You need to be able to talk to others. You have to be able to have a learner’s mindset. You have to understand what different teams and functions do and how they play into a bigger picture so that you can get into cause and effect. And then when you start to do that, you have a lot more ability to actually have impact.
Satyen Sangani: (07:57) Any tips or tricks on how to do that?
Wendy Turner-Williams: (07:58) I’m a big believer in culture, and I’m a big believer in, kind of, networking. So one on ones I think is a big avenue. I am someone who, as soon as I join a team and every day, I have the org chart up every day, every moment it’s always on my laptop. And I’m always trying to understand: who’s this person and they roll up to what org, and they do what, and who else do I know there? And wait, I think I remember this other team was doing something similar, and who was that again? And let me connect them. So it’s not even always about my own networking.
Wendy Turner-Williams: (08:31) I’m constantly playing glue between different teams and different functions in regards to where they can have mutual interests or where they can learn from one another as well. And to me, that starts to build kind of a reputation as a go-to person where people start to come to you. So I don’t have to necessarily seek out. People actually come as well. But I use councils, I use external engagement mechanisms, whether LinkedIn or other social avenues. And then I schedule a lot of one-on-ones and keep them. I think when you build relationships that can’t be one and done, they’ve got to be an ongoing conversation.
Satyen Sangani: (09:12) Tableau recently published its data trends report for 2022. It uncovered some fascinating gaps in trends about data literacy today.
Wendy Turner-Williams: (09:20) One of the findings that we had was that 39 percent of organizations surveyed basically say that data training is available to all employees. But if they went back and actually spoke to a lot of the decision makers, then I think it was 82 percent of decision makers expect basic data literacy for all of their employees. So, in other words, there’s over a 50 percent gap between the amount of literacy and training that’s actually being provided to the employees versus the employer expectation about these employees actually having basic literacy skills, which seems to be a huge disconnect.
Satyen Sangani: (10:02) Why do you think that is? I mean, Tableau’s been around now for 20 years. Alation’s been around for 10 years. Why haven’t we all figured this out?
Wendy Turner-Williams: (10:10) I think there is just somewhat of a disconnect in regard to what a corporation or what an employee perceives as an employment or an employee actual avenue, as far as their own self-learning and self-training versus what a corporation should actually provide to make sure that their workforce skills maintain being relevant. So I think that there’s just a disconnect between what they need to do.
Everyone knows that we need to have training related to safety or security or various things like that. But training related to “how do you write a SQL query” or “how do you actually create a report” is somewhat considered optional instead of being a mandatory type of training to make sure everybody is grounded in some type of basics.
Satyen Sangani: (11:05) And training’s always a tough thing for managers to invest in because the ROI isn’t always apparently immediate, even if it is super obvious that it’s needed. So how do you convince your customers? And even internally within Tableau, how does that happen? How do these programs get funded? Are they getting funded more often?
Wendy Turner-Williams: (11:27) Within Tableau, of course, this is a huge effort for us. A, we’re definitely focused on funding because not only are we looking at funding internally, we’ve also made a commitment or a pledge to basically skill up, I think, 10 million data skills over the next, I think it’s 3 years, by 2025. So, I mean, that’s a public actual pledge that we’ve done of which we’re living that internally.
So one of the ways that we’ve invested is the creation of this CDO function. And within this CDO function, we actually have a data literacy effort and a whole focal group, who’s building out curriculum, who’s working with educational institutes or third-party partners in regard to literacy programs as well — while also looking to package up our own internal literacy about our own culture or our actual business processes all up, and the various software and products that we use across the board. So that is something that is an explicit function, that has been actually tied to our corporate commitments this year with actual resourcing that has come with it.
Wendy Turner-Williams: (12:33) And as far as ROI, one of the things that we found in this Forrester survey was that nearly 80 percent of employees were more likely to stay at a company that basically offered data skills. So you think about ROI: 80 percent is a huge number in regard to actually becoming happy within that company and wanting to stay and just think about the actual influence that happens, too. And at this time of the Great Resignation and at this time where we’re having generational shifts around just the amount of workforce that we have, it’s incredibly important.
Satyen Sangani: (13:10) Yeah, it’s funny because it’d be really hard as a CEO or an executive at a company to say, “Hey, we really don’t believe that we need more data. We’ve got all the data that we need.” That would be an odd statement to make.
And you also have now data that says, “Hey, look, if you invest in your people, they’re more likely to stick around, in particular for these roles where there’s negative unemployment.” I mean, it is really hard to find a data scientist or a business analyst. And if you have those two realities, you might as well invest in the training because it may show up in lower attrition and better business performance, but you’re not going to be able to measure it immediately.
Wendy Turner-Williams: (13:47) That’s right. That’s right. I mean, again, I think every company’s at a different point in time, depending on even the industry, you are probably related to your attrition rates, et cetera. But I’m with you. I think that there’s a correlation that’s easy for companies to actually measure when it comes to seeing the ROI based on those attrition rates versus the investment, as well as even again, back to strategy. It’d be interesting to tie literacy to not just attrition and employee happiness, but also back to strategy execution. And how those progress and how those plot in some type of line against one another. I think that’d be a really interesting kind of data set.
Satyen Sangani: (14:30) So what is a data fabric? Wendy shared how Tableau approaches and defines that concept.
Wendy Turner-Williams: (14:35) When I think about a fabric, I think about this kind of middle-tier angle that kind of wraps around your end-to-end kind of data processes, whether it’s compute and store up to your orchestration engines and your query engines, up to your actual KPIs, metrics, and visualizations. It’s something that kind of goes throughout the process that helps you to basically manage, have insights, and to govern that information kind of appropriately.
So again, if you think about Tableau Prep or you talk about Catalog, et cetera, I think of these as more extensibilities into your more upstream data points so that you can understand full pipeline health. You can understand what type of data is available to actually enrich the existing kind of KPIs and metrics that you might have visualized within Tableau all up. It just gives you more of the ability to both enrich as well as to find and discover as well as to manage
Wendy Turner-Williams: (15:32) So I tend to think about things like, trust kind of angles, think like GDPR or CCPA. Waiting to understand a catalog from a last-mile perspective does not help you necessarily support things like right to be forgotten or DeleteMe. You need to be able to understand a full pipeline stack and you need to understand your role around that data. Are you a processor or are you a sub-processor? What is your role, so that you can actually kind of operationalize a lot of the compliance support needed to actually support GDPR, CCPA, all up?
Satyen Sangani: (16:11) And I guess in some sense, it’s a new paradigm to attack an old problem, which is data governance. There’s always been this sort of element of process and procedure and a heaviness to it. Do you see that evolving or changing over time?
Wendy Turner-Williams: (16:26) It has evolved, thankfully. I, for one, am not a process-for-process person. I am a process automation person. And I don’t know the last time you read a policy. Most people tend to read them or not read them and not discover them. So very much, I’m a person who believes in right-size governance and also believes in kind of governance automation. I think when it comes to scale and volume of data in the modern world, manual processes just don’t scale, whether it’s for quality or whether it’s for policy, or stewardship, in stewardship, driving access control, this is all about automations and all about the platforms to automate those things.
Wendy Turner-Williams: (17:11) From my perspective, I tend to focus on: What are those processes? What are those automation points? And then how do I actually build those in an intuitive way, almost like an application experience that really bakes into an intuitive user experience so that the process piece or the governance aspect is almost buried into just the data engineering life cycle or the data analytics life cycle? And it just becomes an intuitive piece to where we’re guiding people to what they want, and what they need, and the processing that they need to do. And governance is almost like baked in around them, in a way that they can’t quite see it, but all of the right things are happening from a handling perspective or a policy perspective intuitively.
Satyen Sangani: (17:57) So we’ve talked a little bit about data literacy. We’ve talked about data fabric. Let’s talk a little bit about the how. So you implement the tech stack for data at Tableau. What is that tech stack and how have you designed it, and how do you pick your priorities in terms of what to build as you augment that tech stack?
Wendy Turner-Williams: (18:14) We kind of live in a hybrid model. We have a mix of on-prem and cloud, multi-cloud even substrates. So we’ve got a mix of AWS. We’ve got some Azure. We’ve got some GCP. Because I’m doing cross-enterprise-in-nature types of analytics, we’ve got different people that are on different funnels depending on the business unit. And then we’ve got a mix, other cloud services like Snowflake, et cetera, that we’re using for things like curation store and modernize ETL services.
We also, as you know, are a big Alation customer as well. So we’re using Alation as part of our data management stack, as well as some other services around data quality and lineage, et cetera, to really make sure that we have kind of a shared data services model that can support any of our internal data teams based on their own agility and their own business priorities.
Wendy Turner-Williams: (19:11) So I tend to separate from a priority perspective, I tend to separate what I’m doing on platform and the services from what I’m doing from a prioritization related to data engineering or analytics. Those are two different functions, two different teams in which A, we’re focused on where we can actually improve our service health, or scalability, or availability related to our platforms as well as adding new feature sets that bring more benefit to the collective good versus siloed kind of positive cases.
Wendy Turner-Williams: (19:45) And then from an analytics perspective, we really try to focus on enabling our business partners to define their own priority and us executing against that priority to support the strategies all up for those business teams. When it comes to, of course, since we’re a shared service, we sometimes get prioritizations that are conflicting in nature. So when that happens, we’ve got mechanisms internally where we greenlight, we do it kind of in a very transparent way, in which we have all of our business partners and we bring them together to make shared decisions related to priorities when we have conflicts.
Satyen Sangani: (20:27) One of the things that it strikes me about you — or at least that I’ve observed over the years — is that you’ve probably picked, and used, and tried, and kicked out more technology than, frankly, anybody that I’ve ever seen or know. What are the lessons and the learnings there?
Wendy Turner-Williams: (20:42) I think that people need to understand what’s the business value that they’re intending to actually bring? And how do you actually measure that ROI? I think too often people get focused on implementing tech and they think that just the tech solves the data problem. But the tech is just a component to actually get to the value of data. There’s actually the value itself, the data itself, and there’s also again the culture and kind of the atmosphere that you’re creating as well.
And so I don’t think that tech always solves your business problems. I think that tech is an avenue to help you, as long as you understand your actual data need and your business need as well. So my advice is always like: “Fail, and fail fast if you’re going to fail. And then adjust and move, iterate, don’t be afraid of change, don’t be afraid of asking questions.” And again, ultimately your job from a technology perspective or a software perspective is to solve business needs. And if it’s not solving the business needs, why? Make sure you understand clearly those requirements and adjust.
Satyen Sangani: (21:53) Makes perfect sense. And you earlier in the conversation talked about the thing that this podcast is all about, which is data culture, or culture specifically. But obviously we here like to talk about data culture. How do you define data culture? And what’s a good one and how do we work towards it?
Wendy Turner-Williams: (22:13) Data culture is really trying to ensure that you have the people and the aptitude with the skill sets who really want to make data-conscious decisions, meaning not shooting from the hip, not shooting from your gut, but making informed, educated decisions based on information. Now, I think the big thing with culture is that just literacy, just like governance, just like strategy, it’s not something that’s one and done. There, it’s a continuous process, and you’re always trying to ingrain it so that it’s the same as breathing versus an extra step — I’ve got to be in my data culture hat, so that I think this way. You want to make it intuitive, you want to make it something that’s just natural. And that it’s something that everyone groks, not some people, but everyone. Using community to feed good behaviors or to make shifts, as well, I think is important. So how do you feed this cultural kind of animal and keep it alive, and kicking, and constantly growing and thriving? I think it is one of the biggest challenges.
Wendy Turner-Williams: (23:28) The other piece, I think, when it comes to culture — it’s kind of the thing that data professionals never love to talk about, but data’s very political. I mean, it’s incredibly political. And I think that sometimes when it comes to culture, you’ve got to have a culture that also not only wants to be data-driven and use information, but you have to have a culture that, again, wants to have a learner’s mindset and isn’t afraid to ask questions, and isn’t afraid to challenge, either. Like sometimes things should be challenged. And sometimes that is very positive for the collective whole. And having a culture who is grounded in the fact that that’s accepted, and that you’re learning from one another, and the culture as a whole is going to grow from the result of that, is something that is, I think, hard to do. But it’s also something that can differentiate really solid cultures who are very, very successful and those who are not and tend to silo off.
Satyen Sangani: (24:32) So that statement’s super interesting and I’ve got to dive into it. So tell us a little bit more about what it means for data to be political. Why is data political? And what does it mean to be political?
Wendy Turner-Williams: (24:45) Data can be shared. It’s like this living entity thing that can be used across multiple teams for multiple different purposes. But then sometimes with data comes a lot of heavy accountability. So you get into situations from a political standpoint where sometimes it makes sense that data is owned within certain functions, but then there’s no accountability for that ownership because either they don’t have the bandwidth, maybe they don’t have the capacity, maybe they don’t have the knowledge. Whatever reason, there’s no real kind of ownership that happens.
Wendy Turner-Williams: (25:22) The other thing that happens is that data is power. To me, again, people who know how to use data and know how to actually make decisions based on that data and amplify that message tend to sometimes create fiefdoms or silos of knowledge that can happen around that data. And sometimes that is really hard to break through because instead of working collectively against a common good, you could sometimes have people that go off into, again, kind of political atmospheres or Machiavellian mindsets and not work for the broader good for the shared usage of that data.
Wendy Turner-Williams: (26:03) And that happens all the time in data roles. It’s something that I think that data professionals — especially data management professionals — are constantly trying to break down silos or fiefdoms, or clear up accountability gaps, or sometimes bring clarity to accountability issues as well, or quality issues is another big one. A lot of times, people want to be the source of truth for data until there’s quality issues associated with that data. And then there’s this scurrying away from that problem set. And again, so back to a political standpoint, sometimes the data management role or the CDO role is bringing structure to some chaos. And I think there’s just a lot of times, political chaos that happens across the board.
Satyen Sangani: (26:55) That is such a great observation. And it explains a lot of the “why” for what you talked about earlier, which was, “Look, I look at the org chart, and I go and build relationships because I’ve got to figure out how to get people on my side and really on the company side.”
Wendy Turner-Williams: (27:11) Yes
Satyen Sangani: (27:12) Right. To go figure out how to use the data. Any other tips and techniques that you would recommend to folks about breaking through those politics and breaking through those motivations?
Wendy Turner-Williams: (27:22) Always try to align on what is strategically advantageous to your company. Like always err on the future state versus the existing state. I try to do another trick for me personally, not just networking, but I try to understand who are the partners? Who are the people that really get it and who are the people who don’t necessarily get it? And what I try to do is create enough energy in the system. It’s almost like when you’re rolling a rock up, especially like when you first start, some big initiative, like coming brand new into a company or some major problem that’s going to cause a lot of internal change, or structural change, whatever that may be. What I try to do is I try to understand who are the players? And of those players, who tends to get it and be aligned, who’s kind of a middle ground, and then who is someone who is opposed, or doesn’t have the time, or seems like they’re not really heavily engaged, or bought into the direction?
Wendy Turner-Williams: (28:30) And I almost tend to break them up into three groupings for myself. And then what I try to do is I try to network around those groupings to understand who influences who around those things. And then I start to plant seeds and play influence games around what is that direction in which we’re headed and why? And who are those influencers around that person? And create enough motion in the system to where the energy of the positive progress starts to basically overtake any type of negative energy or any type of political postures that may also come at play. So again, I tend to use systems to drive conclusion and drive alignment and drive progress.
Satyen Sangani: (29:13) Literally just what you described sounds almost exactly like a political campaign. You have to find the initial people who care about the issues that you care about, and then you’ve got to advocate for them. And then you’ve got to build a larger constituency. And people — it’s funny because as technologists or as engineers, you might look down on that, say, “Ah, that doesn’t sound like work that I want to go do.” But if you want to drive change, it’s the work you’ve got to go do.
Wendy Turner-Williams: (29:42) It is. And that’s why I was saying data’s political and that’s like the nasty thing no one wants to talk about, but it’s the reality of what our roles are. It’s a change-agent role and change is hard and change isn’t always popular. But when the change starts to happen and things start going in the direction — the positive direction of which the intention is — then those changes start to be embraced.
And that’s where, again, I think you can build enough relationships and once you start showing good and you start showing those things kind of holistically, even the naysayers start to become some of your biggest advocates. Especially once you could start showing them the benefits to their personal space and the things that they couldn’t do before that they can now, or the amount of time that they were spending answering questions, or doing operational support, or supporting trust insights or whatever those things would be. Versus now the ability to focus on some type of new analytics or new strategy because you’re freeing that time up, the more and more that they become a huge advocate.
Wendy Turner-Williams: (30:53) And for me, I mean, my job, I don’t own any data. My job is almost like a … it’s like a conductor over an orchestra, really. You’re trying to get all these different things from business units’ perspective and all these different teams to play beautiful music together. And that’s what the company needs, is that beautiful music and your job is to help try to coordinate and bring those things to reality.
Satyen Sangani: (31:15) I love it
Satyen Sangani: (31:20) Change is hard. It’s often unpopular. If you can motivate people to change in a way that helps them grow, they’ll probably thank you for it. So how do you launch a data culture? How do you lead change with confidence? If you’re like Wendy, you start with people. Know your org chart like the back of your hand, and be very clear about how transformation will affect all the relevant stakeholders. Meet with people, listen to them, motivate them, show them not just why change is necessary, but also how they will benefit. Thank you to Wendy for joining us for this episode of Data Radicals. This is Satyen Sangani, co-founder and CEO of Alation. Thank you for listening.
Producer Read: (32:01)This podcast is brought to by Alation. Are you curious to know how data governance might actually be good for your business? This webinar with Gene Leganza, Research Director from Forrester, explains how to align people, process, and technology for growth-oriented governance initiatives. Check it out at alation.com/youtube.
Season 2 Episode 12
With an approach that combines show-and-tell with “So what?,” Wendy Batchelder drives success as the CDO of Salesforce with her experience and empathy. In this episode, Wendy shares her passion for data and equity, why some data trends are merely fads, and how starting with a data culture ends with data literacy.
Season 2 Episode 9
Ashish Thusoo has been on the leading edge of a data culture, whether it’s as a founder of a data lake startup, developing the Hive data warehouse at Facebook, or in his role as GM of AI/AML at Amazon Web Services. This discussion traces the evolution of data innovation, from big data to data science to generative AI.
Season 1 Episode 23
Centralizing data was supposed to produce a single source of truth. Why did it fail? Zhamak Dehghani shares why she created the data mesh, and reveals how this socio-technical approach decentralizes data ownership to address growing complexity at large organizations.