Tim Gasper [00:00:31.333] Hello, everyone. It's time once again for Catalog in Cocktails. It's your honest, no BS, non-salesy conversation about enterprise data management presented by data.world. We've got tasty beverages in our hands. I'm Tim Gasper, longtime data nerd, joined by Juan Sequeda. Hey, Juan.
Juan Sequeda [00:00:48.533] Hey, Tim. How are you doing? It is always great to take a break during the day and have this chat about data. We've been doing this now. This is, again, season eight, year five doing this. I am always so excited. And I feel what gets me really excited is that we always have the opportunity to talk to so many phenomenal people. Amazing people. And today, finally, we get to have on the show, Tiankai Feng, who is a data strategy and governance lead at ThoughtWorks Europe and is the author of the brand new book, Humanizing Data Strategy. Tiankai, how are you doing?
Tiankai Feng [00:01:23.953] I'm great. Thank you so much for having me. I've been a fan of this podcast for a while, so it's really nice to be on there as a guest now. So thanks for having me.
Juan Sequeda [00:01:31.393] It's super cool. I know we've met in person a while ago last year, I think in San Diego, and it's great to be able to continue having these conversations. Conversations uh hey so let's kick it off i mean i know just as a reminder with folks uh. Five years into the podcast so we've always been doing it live and now this season we're gonna have to do some stuff pre-recorded so we're gonna be still streaming this you're watching us streaming live right now streaming uh Wednesday at 4 p.m even though it's pre-recorded so that means that we don't have a cocktail right now honest no bs so no cocktail but we can still talk about cocktails i don't know uh uh in the in our in the last couple months Tim anything else you've discovered uh tian kai what's your favorite cocktail mocktail.
Tiankai Feng [00:02:10.220] Well, I'm a fan of a good old mojito. I think I really just like the mint part of it. I mean, I like mint in everything, basically. That's kind of what I like about snacks as well. So basically, a good mojito would be my choice of drinks.
Tim Gasper [00:02:24.060] I love a good mojito. Lately, I've been doing a lot of whiskey cocktails, and I've been trying different combinations of whiskey and different simple syrups and things like that. But lately, I think I found my favorite old fashioned, which is a pretty big update, which is I've been using Woodford Reserved Double Oaked. And then I've been using Demerara syrup and I've been using some special bitters that have some cardamom in it. And overall, I'm just like, whoa, like I just elevated my old fashioned experience. So that's my latest on the cocktail front.
Juan Sequeda [00:03:02.740] That's fascinating. I need to get more into like trying with the different syrups and bitters and stuff. I've been trying kind of a spin of an old-fashioned with adding some Aperol. And that's a bigger kick into that. I haven't tried that. Yeah, that's a good one. Try that. All right. Warm-up question. We're talking about humanizing data strategy. What other things do you think need to be humanized?
Tiankai Feng [00:03:31.580] It's a great question. I think, for me, humanizing means to just think more about the people that are impacted by something, right? And I think if I look at my own weaknesses, I really hate queuing up. So whenever I'm queuing up somewhere and I have to wait in line for something for a long time, I just think that could be more humanized, right? Like, either you could maybe anticipate the demand a little bit and open more cash registers, for example. example, or maybe you could put certain processes online so people don't have to actually pull literal paper with numbers on it and sit there for two hours. I think there's a lot of things where I would just say, if you keep the people more in mind that have to do it, then you could help humanize it more. So maybe that's more from a venting point of view, but that would be my point of view.
Tim Gasper [00:04:14.700] I like that. That resonates a lot with me as a chief customer officer at Data.World, right? You really want to think about the humans that you're impacting and making their lives easier, right? And maybe I'll answer this question in a slightly non-business way, which is like, I think we should humanize our politics more. We're in the middle of election season, right? And sometimes it's us versus them, and it's about the policy and this and that. And we have to remember that it's about the American people. So hopefully that's something that we can humanize more as we go forward.
Juan Sequeda [00:04:46.660] And I'm going to take a completely different spin. And I think in places where humans are that we are too, uh, we, we spent too much time on our phones. So the thing that I do all the time, I walk into an elevator. Everybody gets in and then they pull out their phone and i'm like i catch myself doing that i'm like no i'm gonna stop and i'm gonna just stare but nothing not even i could wait i can wait, 10 20 30 seconds a minute and not have to look at my phone so and then hopefully you actually start conversations with people in the elevator so yeah
Tiankai Feng [00:05:18.176] But often you don't even have reception in the elevator so i don't actually know what the people are doing right so they're probably
Tim Gasper [00:05:26.836] People just don't want to talk right so they're like uuhhhh
Juan Sequeda [00:05:31.656] All right all right let's kick this off honest no bs what does it mean to humanize data strategy?
Tiankai Feng [00:05:41.276] Yes so um i think uh my starting point is that when when people talk about data strategies they always talk about people, process, and technologies right so the classic triangle And we all know that the technology part is very well discussed. And there's a lot of tools out there and solutions that can make things work. And the process part is, I think, by nature understood by everyone that you need to do processes. But the people part, I feel like is always on the surface discussed as important. But that's it. Like it doesn't go into more of the practical detail or what to actually do about it. They just say change management is important. Do it properly. Right. Communication is important. Do it properly. And there's no way to make it a bit more specific for people to actually apply it. So my whole attempt of this book is to actually make it a bit more practical and really make it something that people can resonate with to apply it in their own way, in their own organizations. So I break it down into the five C's, which are competency, collaboration, Communication, creativity and conscience, which basically is all going back to what makes us human. Right so we all have an intrinsic need to communicate we have language as compared to a lot of the animals out there i would say or other living creatures in the world right um communication um also collaboration right that need to work together and for human connection for example and even conscience which is a very human thing and although we are trying to apparently make ai feel conscience too i think it's something still very much inherently human to have that gut feeling that something is bad or good and what i'm trying to do is basically just to apply those more into the day-to-day of any data strategist.
Juan Sequeda [00:07:18.916] All right. So this is fascinating how you just jumped into this, the five Cs, communication, collaboration, creativity, consciousness.
Tiankai Feng [00:07:23.096] And competency.
Juan Sequeda [00:07:32.156] Competency. Okay. So let's dive into them.
Tim Gasper [00:07:39.929] Do you actually approach this in kind of a certain order?
Tiankai Feng [00:07:46.449] Yeah, I mean, I did start the first C in the book with competency because I feel like this is the one topic where it should be actually the foundation for the other four Cs to work. And competency, I think I do describe data literacy and data fluency as the concept that everyone is important. But I know also that you two actually love to talk about it, that I feel like only talking about data literacy is a little bit one-sided because you want data people also to get the business acumen and the business knowledge to actually be able to build that bridge, right? Just sitting back and hoping that our business stakeholders get more data literate is not going to work, but you need to build that bridge actually to also understand the other side more. And also within Competency, the idea is that we should encourage people to get up-skilled and to be able to apply what they learned and not only attend webinars and kind of go to conferences, but give them that space to actually apply and to try out things, to basically see if it works and how it can work in a different way. And lastly, also, it's basically about career steps and drop rotations as well, right? That if you give people the opportunity to grow and then you cannot reward them with like the next career step or jumping between between different departments because of what they learned and what they achieved, then very quickly, you kind of lose those talents again. And then all of the effort you made of building that data competency or like the business competency, you're losing again. So think about it more holistically, how you wanna basically encourage competency and then also reward it.
Juan Sequeda [00:09:17.409] This is really interesting on this competency, especially on the part of growth and education. And I wanna bring up, I was having a chat the other day with Omar Kajwala, also a former guest. And he was talking to me that, look, when you're doing all this training, you have to be very careful because you're like, oh, let's train. Let's give everybody education because that's a great thing. But then you're like, wait, if I'm giving people all this education, but they don't have a way to go try it. They're like, I just spent all this time. So what? And that can get frustrating around that. And then you're also not spending your resources wisely. So this is really interesting about bringing that up because I guess everybody's like, yes, literally, let's go educate. You should go learn this. But it's not just about learning. It's also about applying this to your day-to-day, which is helping you grow.
Tiankai Feng [00:10:05.889] Exactly. And I mean, this is even rooted in science, right? That if you learn something and you don't apply it practically, you're going to basically forget 90% of it because it only sits in kind of your conceptual mind and not actually in having experienced it practically. So just in the sense of the efficiency of sending people to trainings, even from that sense, letting them actually apply it practically should be in everyone's interest.
Tim Gasper [00:10:29.396] Mm hmm. That makes total sense. You know, when you think about competency. Right. I know sometimes people think about like, you know, there's more like the quote hard skills or the technical skills. Right. And or, you know, and then there's more like all the soft skills like communication and collaboration and, you know, ability to present and things like that. Right. You know, obviously, in your five C's, you have some of those C's that focus more on like the communication, the collaboration, things like that. Do you think of those as sort of these, quote, soft skills as being separate from competency? Or do you think of competency being the combination of the hard and the soft skills? And do you have better words than hard and soft? Because I've always hated those words.
Tiankai Feng [00:11:08.876] Yeah, I don't like talking about hard and soft skills either. But I do think that it's actually the combination that fits into competency. I also, for example, write about how you have to wear different hats as a day professional. So the idea is that even if your role is defined with specific responsibilities and specific tasks you have to do, it's still you have to act and behave differently in different situations, right? Like, for example, sometimes you need to be a detective and you do ask hard questions to get to the root cause of a certain data problem, right? Sometimes you have to be a negotiator because actually you have maybe some kind of a clash of priorities. So you need to find out how you actually can negotiate towards the one common goal with your stakeholders and do it. And some other times you need to be a therapist, right? Where actually you need to just listen very clearly to the other person's pain point, empathize with them, and then go into it and then thereby gain their trust. So in many ways, I think like competency of being able to wear those different hats and to know when to behave in certain ways to be the most successful data professional, so to say.
Tim Gasper [00:12:11.005] I like that. One last question on competency before we kind of move to some of the other Cs is that, you know, one thing that I find difficult as like a hiring manager is trying to determine the right competency through the hiring process. Do you have any thoughts and tips on like how to identify the right competency early versus find out maybe too late?
Tiankai Feng [00:12:33.685] Right. Yes. And I think the greatest challenge there is that many of the applicants and talents are probably already prepared for certain generic questions when it comes to this. Like, right. For example, tell me about a time when you solved a conflict. And then if people probably already made their homework, they're just going to tell you a story which might or might not be real. Right. And then you still don't know if that person is not in the conflict. thing. I actually realized that to be more specific and to catch those people a little bit more off guard to get the more authentic answer is to give them a scenario and to ask them how they would behave in a scenario like this that is very specific, because then they have to actually change their mindset towards that scenario and cannot give me this generic answer or something that fits like it, right? So, and that might be that, for example, somebody is reporting, two different teams are reporting on the same KPIs, but the dashboards show different numbers. Now the C-level people are really angry. You are working in data governance. What are you going to do about it? This is about financial P&L statements, and it's a really high stakes environment, right? And then basically see how they can actually adapt to it. Those people that actually have experienced it, they can pretty much transfer their existing experiences and skills to that situation. Those that only know it theoretically or haven't thought about it, they're going to struggle, right? They don't know where to start. They kind of just like say a lot of it depends and not sure and these kind of things. So yeah, I think it's more about the way of questioning and to make it more specific that you can kind of get a little bit under the skin of the people.
Tim Gasper [00:14:04.245] Yeah. Yeah, that's interesting. I love that. I found that to be useful in some cases as like, oh, hey, here's this scenario, or here's this situation, like, how would you respond to it? And then, and then you throw in some curveballs to see if they really kind of know the edge cases, right? Oh, that's a great suggestion.
Juan Sequeda [00:14:19.305] I have to say that an honest no BS thing right there is like, well, if you, warning signals is, you get a lot of, it depends, right? It's like, exactly. Okay, that's a good one right there. Okay, so after competency, what's the next one?
Tiankai Feng [00:14:36.065] I would say collaboration. Okay. Right and we can talk about this um and yeah maybe one of the key statements i'm trying to make is that we talk a lot about specific working models like service or self-service right now in the data space right so basically do you have a data team that is doing all of the work for their stakeholders so they're like a service function or you want to do the opposite and you kind of want to decentralize all of the skills enable a platform and then business teams are self-servicing their own things. In any of those setups, though, you still have people on both ends, right? We talk about data platforms, but you always have a data platform team and a data platform user team. So they're still peeping from each other. And what I'm trying to say is that no matter if you go for service or self-service, there's always going to be a mismatch of expectations, because it's always going to be wanting more from the other side, right? I want more of your service, or I want more self-service features. This is not enough. It's never going to be completely solved. And to basically soften that conflict and reframe it a little bit, I'm proposing to talk about co-creation instead, right? To say we do actually have responsibilities together to achieve something, and we both have clear contributions to it, but we have a shared goal. And it's not your goals versus my goals, but actually we do want to achieve the thing together. And by reframing that, you might think about also if a ticket system operationally completely makes sense, or maybe you should maybe start having more human contact again. It also impacts how you shape your quarterly objectives potentially and not just do it what you think is right, but you actually talk upfront and you commit together to things. And it also impacts how you then get credit for the things that you achieved with your stakeholder teams, right? And showing up as co-creators of something that worked and not just one party gets the credit and the other party doesn't. So I hope that this is kind of creating a ripple effect of just a better collaboration.
Juan Sequeda [00:16:30.525] So this aligns very much with how I do the work, how we do our work in our lab, which I call it as co-innovation. So for example, so it's, I mean, we are at Data.World, I lead our lab, and it's all about, yeah, we want to go do new, unique things, right? Push, put, I mean, push the barrier of what we do in our platform, in our industry, but not just by ourselves and kind of like in a corner, like we want to be always working with our, with our strategic customers to understand, Hey, what are you trying to go to? You tell us how we're going to, how we should push us. And I also want to push you. And we kind of find some middle ground. I'm saying like, this is actually valuable for you. We're doing something innovative and for the industry and for us and everything. So I always talk about the co-innovation. I really like this notion of co-creation. I think that's a good T-shirt there, Tim, going back to our... Nice. Like data should be like people teams or data people teams should be co...It's all about co-creation.
Tim Gasper [00:17:28.945] Yeah, people first is about co-creation, yeah. No, I like that. One other thing that I think is important, you know, obviously, you know, around, you know, you see, you mentioned the softening the conflict, you know, really working together on the collaboration that helps to manage the expectations better. One other area that I see be challenging with data teams, but it's also hard for product teams. It's hard for lots of other teams is the trade-offs, right? And when it's more of a service style organization, it's more like, hey, I open a ticket and I want that ticket to be served, right? Whereas, as we all know, there are limited resources, there's limited time, and so you have to make trade-offs about what's going to be prioritized or not. Do you find that the co-creation mentality helps with managing those kinds of trade-offs as well?
Tiankai Feng [00:18:23.584] I absolutely think so. I mean, I do get the benefit and I do appreciate the benefit of ticket systems because you need to be able to efficiently deal with all the tasks. You manage them, you kind of prioritize them, and you do it. But in many cases, it's seen as a replacement for the human contact that is around that and for putting the context around it, right? And unfortunately, those ticket systems then dehumanize actually things because stakeholders who created the ticket, they don't see any person behind the ticket. They just want that and quickly getting the results. So you don't basically talk about motivations or details or any contextual information anymore. More so basically um i would say that a ticket system is important but if you can create some human contact around it still and make sure that you understand that they are human being both sides who are actually working on things and that uh communication should still just do a ticket comment so to say then i think that can at least enable that right mindset about operational.
Tim Gasper [00:19:25.044] Well said well said, we're we're not just a service now or a jira monkey right exactly.
Juan Sequeda [00:19:34.084] I love this. You want to have that communication, but not just through comments, right?
Tiankai Feng [00:19:39.644] And you just act passive-aggressively the other side. Please still expect an answer.
Juan Sequeda [00:19:48.204] Yeah. All right. Collaboration. What's next? Communication is the next one.
Tiankai Feng [00:19:55.740] Communication is the next one. I mean, yes. So I would say for communication, and I put that really in the context of data people being often misunderstood by the business. What I'm basically suggesting is that although business impact and articulating what the impact on the business is, is important, right? Like gaining more revenue or saving costs or like avoiding risks. But in many cases, to actually get the proper buy-in, you need to address the intrinsic motivations of people as well. no matter if it's C-level or the operational people. And that means that besides the organizational framing of the value, that thinking about the personal reward as well of a value can also be really strong. So, for example, let's say that the business impact is higher profitability, right? And then the middle ground is more about, let's say, a higher productivity. activity. And for a personal reward, a few people have to spend less time manually cleaning up data once per week, for example, right? For them, what counts is that they are so tired of just cleaning up the data over and over again, because they don't know how to fix that problem sustainably, that they actually just want to spend their time doing something more exciting, right? And let's say now we save them four hours per week to actually not clean data anymore. For them, is more important than knowing how much money it saved on the overall business impact and how many dollars or euros is basically saved. And I feel like if you can frame it both, so you tailor your communication, not only towards business impact, which nobody can deny, but you also do it personally to certain audiences, then you actually get the intrinsic buy-in, not just the undeniable, I have to follow the business goal buy-in, and that should create a more sustainable buy-in, let's say, and commitment from people.
Juan Sequeda [00:21:42.273] That's uh very humanizing i like to say because um it's like yeah every we live i always say this right we live in a capitalist world we need to be able to like our our goals are twofold make money save money and then the third one i'll ask is like mitigate risk which is really going to be about how you like avoid making avoid so but like as a hard cold capitalist like there's still people right at the end of the day right let's remind ourselves that companies organizations are a human endeavor they're human beings who get together because they have a shared goal that they want to accomplish uh and so there's people who are part of that goal like so it's like and those goals are because they may align with the mission of the company too people will choose that but also they have their personal goals. And those personal goals maybe align somewhat with the company too, or they just have some other things. And I think that goes back to another point that you made up is when it comes to the competencies, what are your career progressions? What do you want to go do? How are we helping you there? Because if you're happy, then that's actually going to help the organization that you're part of to actually be more productive and so forth. So I find this very – part of the communication is communicate what you're doing from a business perspective, but also from a personal. I think that's really important. I like how you're, you're, you're calling that out.
Tim Gasper [00:23:05.393] Is there a difference in the advice you'd give around communication or the, or the guidance you would give when you're talking to a leader of a data organization, you know, a CDO or VP or something like that versus an individual, right? You're a data engineer or a data analyst or someone who's just trying to do your job, right? Does the advice around communication change or is it the same?
Tiankai Feng [00:23:26.589] Yes, I think the communication definitely changes, right? Because very simply put, a CEO cares about different things than, let's say, an operational junior data analyst, very simply said, right? And really knowing how these audiences work is really important. And I think this is also another part of what I'm writing about, right? So that communication always starts with understanding the audience, right? And I talked about that personal reward stuff. If you don't know the audience, then you will never know what the personal reward actually is for them. So I would suggest to find out first before you communicate. Because if you communicate in the wrong way and you look assumptions about them, and then they all feel misunderstood by you, you might damage things more than actually have bettered them. So then you have to repair the relationship again, which takes a whole more effort again afterwards. And so what I'm thinking is that with stakeholders, everybody does do stakeholder mappings. You usually differentiate between advocates and detractors. But I think there's even more to that. If you think more about them with different persona types, like, for example, I talk about thinkers versus doers. There are people that love to think, but then they don't want to do anything. Or they do a lot without thinking it through, and then it's too late to do it. And there's a lot of people in the middle, too. too, right? But thinking about the different audiences you have and stakeholders, how they are in that matter helps. And the other side is also skeptics versus believers, right? You have always people that from very beginning hate everything new first, and you have to convince them. Other people jump right on board. They love everything that's new, and then they might see it a bit more critically afterwards. But basically using different narratives for those different personas is important. And first understanding them and then sending your message, whatever you want to talk about, I think can make things really better.
Tim Gasper [00:25:10.469] That is fantastic advice. And I see two dimensions to kind of the advice you're giving here. One is understanding your audience from sort of a more structural level, right? Like what is their role? You know, even what is their personality? What is their kind of, you know, do they gravitate towards skepticism or, you know, enthusiasm, right? But then another is like, you know, another level is literally what are your goals? Like, what are you trying to accomplish for yourself for your role for the organization? And it's different than me. I think one thing that I think we don't do enough of as you know, both data professionals and just broader professionals in general, is, you know, we always feel like we need to go into meetings and things like that with an air of like, I have all the answers, like, you can be confident in me, I know everything, right. And, you know, the number of times that the person who called the meeting does more than 50% of the talking in the meeting, right? Versus starting a meeting with questions like, hello, before I go into anything, hey, Tiankai, I want to ask you some questions. What are your goals? What are you trying to do? Tell me more about you, right? Like, we need to do more of that.
Tiankai Feng [00:26:14.418] Absolutely. And I actually learned it the hard way myself because most of my career, the first step was being a data analyst, right? And as an analyst, everybody welcomes you because you bring insights to them where they're like, oh, wow, that person had analyzed the data and tells us what to do and what the recommendations are. And then I switched to the data governance side, and all of a sudden, all the doors were closed, and nobody reacted to me anymore. And if they reacted, they were really cautious around me first. So I had to learn how to start conversations in a better way. Because if I came in, I'm now leading product data governance, for example, and this is how I'm going to plan to do things, and this is how we set up guardrails and policies, then I'm scaring people immediately off, and they never want to talk to me again ever anymore. So I changed the narrative. I came in, I'm like, yeah, I started this new job, but I really want to understand how you are doing. What are your challenges with data? What are your business objectives? And then when they told me more and more and they were able to share, I could actually tailor what we're offering and what we're working on to solving their problems, which seems much more collaborative than just pushing my services on them or pushing my methods on them.
Juan Sequeda [00:27:20.158] This reminds me, a fantastic book I recommend is Fierce Conversations. And the author goes into, I just looked it up. She calls it an issue preparation form. So it goes, so you go, what the issue is, right? Spell it out. It's significant because, right? My ideal outcome is, fill in the blank. What is the relevant background information we should have? What have I done up to this point? And what are the options I'm considering? And then the help I want from this group is the following . And that really points to, like, this is like the clear communication that really you want to kind of, she calls it, you want to interrogate reality, right, around that. And I think just, I mean, her definition, after reading the book, Fierce Conversations, I'm like, our whole honest, no BS kind of mantra we have, we're like, that's exactly what it is. I just, honest, no BS for me means, like, I just want to have these fierce conversations with you. I want to get down. I want to understand what we're trying to do. Understand what motivates you. What's motivating the company. Why do you need to go do this? And sometimes like it just won't work and won't align and it's okay. We need to part ways. And that's better to know that upfront, right? Because part ways is going to make you happier and better and accomplish the stuff that you want to go do and make the company do better. Like that's nothing wrong with that, right?
Tiankai Feng [00:28:43.183] And that's a really good one because even with what you just said with filling in the blanks, not many people think that way, right? Because once they have an issue, they're actually more impulsive usually by nature. and they kind of just want to vent about it in a very emotional manner. But by asking those questions, what you just said, you actually kind of rationalize it more. And so you basically let them think more critically about it. And then you can both come to a rational agreement first and then basically emotionally solve it. Then I think that can be a really powerful move instead of just being really one emotion against the other emotion, right? So very cool.
Juan Sequeda [00:29:14.583] All right, next one. Okay. You got two left. Yes. Creativity or conscience?
Tiankai Feng [00:29:23.878] I would say creativity. Yeah. And creativity, basically, what I'm trying to say is that by nature, data jobs seem to be not creative, right? Because it sounds very analytical. And if you describe somebody's task in a data role, it doesn't really sound creative at all. But what I'm trying to say is that creativity really is a muscle that can be applied in different ways, right? Right. And no matter what your job is, there's always a way of being creative and bringing new ideas. And even more realistically, based on basically how the human beings are, creativity is the source for innovation. Right. Because it starts actually as personal ideas in someone's head. And then it turns into actually something technical or something for the society that actually innovates things. So telling ourselves that I work in an analytical job, so I do not get to be creative is really the wrong message. So instead, a data strategy should actually encourage people to be creative, give them that space of experimentation and to be able to fail without too many grave, basically, consequences and to encourage everyone to even co-create again, right? To bring in together their creativities and come up with even better things and really go with that. I mean, in my own style, as you know, I'm a musician, so I do make that comparison between music and data as well, right? So even music, you can be very analytical about it, right? So you have a scale of notes that is predefined, you have different rhythm types, you have instruments. And if everything is predefined, it's again, just how you mix and match the different things in music to actually turn it into art. So why should it be that different to data when you have data points and you have structures and everything? It's also art because you put something new together and you interpreted it. So the idea is basically to how to practically encourage more creativity also in data roles.
Tim Gasper [00:31:14.365] Interesting. I feel like that can be tough sometimes in the data world, right? Because there's a very strong, to your point about sort of the perceptions of the professions around data, that there's sort of like a right and wrong answer for everything. And certainly there are things that you can do that are pretty dumb and you shouldn't do those things, right? But like, you know, it's not completely open-ended, but, you know, in general, do data people kind of think too rigidly about things in their field?
Tiankai Feng [00:31:44.705] I do think so. I mean, I'm getting often the question, I want to be more creative, how can I get started, is usually when other data people ask me. And my answer is always think about something in your day to day where you feel like there's some optimization potential, right? It might be like a daily routine you have, it might be a certain SQL code that you're writing, anything. Anything and the moment you start reviewing it and you just bring in new ideas on how to make something better you are already creative you are already making the effort of thinking how you can bring new ideas to make things better and basically this being really practical can lead you to identify more opportunities where you can be creative right and over time ideally people realize that there's a lot of ways how you can be creative and how you can bring it more in and how it then actually converts into innovation that has then also an impact on the organization or on the society.
Juan Sequeda [00:32:42.305] I'm reading, well, listening to the book, The Rational Optimist. And I'm just kind of starting it right now. I'm on chapter two, three. And it's all about like kind of where innovation comes from. And innovation comes from, again, having so many different people coming together, right? Right. So it's like and there's like, why are these I mean, we've seen so many different type of like groups of people who are like in Australia. Right. And like, but they didn't interact with so many people. So when people came out, they realized they didn't have all the different types of tools that other people had already done in so many different places. So it's like that creativity, innovation fosters talk about co-creativity, right? right, co-creation, co-innovation, is because you have the people part involved in it. People are talking about that stuff, right? So you really need to encourage that. So I like this. This is, I'm really excited about the last one now, because this has been really, really, you have a really nice flow around this. And what I'm enjoying about this is that a lot of it is kind of like those quote unquote obvious things. They're like, oh, that makes sense. I never thought about it. But it's like somebody had to put it down and write it, put it down on paper and have this conversation.
Tim Gasper [00:33:55.469] So obvious, but easily forgotten.
Tiankai Feng [00:33:58.349] Yeah. Yeah, I actually literally, I think in the last chapter of my book, say that none of what I said is particularly groundbreaking, but I should be just holding up a mirror and maybe ideally show you what things you're doing well and what things you're not doing well too much and give you a little bit of a practical inspiration to do some things in a better way. That's basically what I'm saying. So I think I'm very conscious about this, that it should feel all natural because it's all based on basically human behavior, right? But it feels like people do need a reminder to basically beyond all the capitalism that you described, right? To also think about how to be a human being.
Tim Gasper [00:34:32.589] Yeah. If it's obvious, why everybody, why aren't you doing it? Right? Exactly. I think it's a call to action. So tell us about conscience. What's the role of that in all of this?
Tiankai Feng [00:34:41.849] Yeah, basically, conscience in a very nutshell, I would say, break it down into why optimism is important and how you have to use your human judgment and your critical thinking and how it's everyone's responsibility to apply that. And first, optimism is about that. If you really want to make better changes for the world in terms of data and AI, for example, then you need to believe that things can get better. So it's really like a foundational thing. And in psychology, you have that concept of actually active optimism versus passive optimism, which means that you actually make sure that it's the best possible scenario and you contribute to it becoming the optimistic outcome that you want to have. And this is what we all need to do when it comes to data and AI. And particularly because there's so much hype going on around AI, right, that we all have seen the public failures of certain big companies who have been doing bad things with it or things that were really embarrassing, right? That basically, if all of us, we might not be AI experts or AI developers ourselves, but we all have the opportunity and I would even say the mandate to give feedback if we see things are not going well, right? There's always some of our feedback mechanism. If we think something is wrong or something is going in a really bad direction, it's our mandate to actually tell the responsible people that something should not happen this way and it should be better. And it might sound counterintuitive, but that means also that right now we are in this phase where human oversight is still a bit more needed. Although we want to automate more and make AI more autonomous, but we cannot really afford for the human to be out of the loop right now. So to have the human in the loop with the right critical thinking and human judgment is really important.
Tim Gasper [00:36:29.228] I think that's super well stated. And two things come to mind is one, you know, I think when people think about human in the loop, they tend to think about it more from the context of, well, what if it's inaccurate, and I want the human there to help it with its accuracy, right? I think that's maybe the current paradigm, I'm sure we'll grow out of it. I happen to be a an optimist, maybe an active optimist around AI. But, you know, there's a whole nother dimension to the human in the loop, which is the human conscience to be part of the judgments that are happening with AI, right? And then the second thing that comes to mind is I think of Gandhi's quote, right? Be the change that you want to see in the world. So it sounds like you're kind of talking about that here with active optimism.
Juan Sequeda [00:37:13.388] Exactly. I just put this in bold. This is a T-shirt. We can't afford for the human to be out of the loop. That's a great quote right there. I think we always talk about it and so forth, but like right now is a big moment. And I think what you've really talked about all these different C's here kind of leads us to that. It's like humans are going to be involved in all these parts. And even though as a computer scientist, I remember I grew up being taught like, oh, computer science is all about humans. Automating right automations automations automations right uh how to do things faster and and and i i i grew up saying oh that that human side the computer human interaction that's some that's somewhere else though that's not real science i'm like but now i realize like that's the most important part i remember the the hci folks are the ones who are like who are telling me is like uh you know those computer scientists you computer scientists you like easier problems because you don't like to work with humans because that's complicated, right? But before we get to our lightning round, any final thoughts before we go to our lightning round? And also, please, how can people find your book?
Tiankai Feng [00:38:24.249] Yes, I mean, so my book is on Amazon on the Technics publication website. And I think everywhere else you get your book. At least in Germany, I know that people can go to our local bookstores and order it. I tried it myself just to see if it worked. So it did work. So I think everywhere where you get your books, you might have to order it specifically, but you can definitely get it online and offline.
Juan Sequeda [00:38:44.289] Perfect. Well, let's kick it off and let's go do some lightning round questions, which I haven't even seen yet. So Tim's been working on them. So here we go. Number one, people process technology. Do companies know they need to focus more on the people part of the equation, but just don't know how? Or they don't really even realize people need to have that focused?
Tiankai Feng [00:39:05.409] I think it's the first one. So they know it's important, but they don't know how. And it might be related to either not having the right people and experts in the leadership that know how to do it, or they are not sure, they're not investing enough in it because they don't know how to invest in it, basically, right? So is it actually people that more, do we have to restructure our organization? Like, what is it they need to do? So all of that panicking and uncertainty leads to just doing it more cautiously, but then that's too little, basically.
Tim Gasper [00:39:34.549] That's fair. That makes sense. Second question, who are more skeptical, data engineers or data governance people?
Tiankai Feng [00:39:44.849] Wow, that's a great question. Yeah. Those are probably the most skeptical data professionals. So I don't know who actually wins in that equation, to be honest. Because, yeah, I mean, both have probably experienced a lot of bad things. So I don't know who actually is more skeptical. I would maybe answer it the other way around, that I hope that we can turn the skepticism into just avoiding certain failures and problems in the future and use it to accelerate solutions to be more optimistic, so to say. Myself, having been in data governance, I think that I'm even a rarity that I'm still so positive and optimistic myself, having experienced all the resistance towards data governance that I've experienced. But it's really, yeah, it's important to believe that things can get better. Otherwise, you might as well give up, right?
Tim Gasper [00:40:39.502] I think that's great advice. And maybe we've got an episode that we need to do coming up, which is called The War of the Skeptics, Data Engineer versus Data Governance.
Juan Sequeda [00:40:49.862] That's a good one right there. All right, next one. Of your five Cs, which one is most lacking in organizations? The biggest missing piece. Is it conscious?
Tiankai Feng [00:40:59.842] I would say it's conscious, but not because people think they don't have it. I think it's not prioritized and cross-functional enough in many ways. So what I'm saying by that, right, that if you think about human oversight, there might be like a data product owner or something that is overseeing things that is one human in the loop, but that is not enough. In many cases, you might need somebody from legal, you might need somebody from human resources, you might need ethics experts to also go in there. And that, I think, has not been realized yet, to know that there's more implications than the technical implications. You need people that actually understand society implications and human implications too. And so it should be a bit more cross-functional than it currently is.
Tim Gasper [00:41:43.560] So people who understand societal and people aspects. I love that. All right. Final lightning round question. Does it help to humanize data strategy and work more generally if we're willing to drop our guard more and be more personal and raw in the workplace? Or is that pretty risky?
Tiankai Feng [00:42:07.380] I would say yes. It can make things better. But you cannot. Request people to do that, is what I would say, right? Itself saying you have to be vulnerable in the workplace is not a nice request of someone, right? But if people decide to role model it based on their own intrinsic motivations as a data leader, for example, and you show your team members that you can be vulnerable and you can be more open about things, then that is sending the right message. But you cannot basically enforce it from people to do that, right? So in a way, it's a little bit of a paradox, but I would wish that people would feel comfortable and safe enough in a workplace to be a bit more vulnerable and open.
Tim Gasper [00:42:50.620] Oh, very well stated. It's not something you can enforce. It's not a policy or process you can put in place. If you create an environment where people can be vulnerable and feel safe and you model it as a leader, that's the key. I love that. Exactly.
Juan Sequeda [00:43:08.290] All right. Takeaway time. Tim, take us away with takeaways. Kick us off.
Tim Gasper [00:43:13.330] Oh, my gosh. So many good takeaways. I'll try to keep it to just a few. So usually you said when Pete, when we started off, right, we asked you, like, honest, no BS. What's this whole humanizing data strategy thing? Right. And you said usually people talk about people, process and technology. And the tech's pretty well understood with all the vendors and the solutions that are out there. We love to talk about tech process, even though it doesn't get as much spotlight as tech also gets. That's, you know, it's decently understood. We know how to approach various different processes, right? We can always get better at that. But the part that is least discussed and the least effectively implemented is the people part of the overall equation. And you mentioned that people will say things like change management and, you know, communication and you got to do it properly and make sure you, you know, change management is like a sentence in itself. You could just say change management, period. And people go, oh yeah, change management, right? But like, well, how, how do we do that? You can't just say the phrase. So you outlined these five key areas and what makes it easy is they have a mnemonic, right? They all begin with C and it's competency, collaboration, communication, creativity, and conscience. And you started with competency, right? That's the foundation to really allow the others to work. And, you know, it's not just like data literacy or something like that. It's really the combination of the capabilities from the data side to the business side that make everyone effective. It's the hard and the soft skills, even though we don't love those words. Right. And it's the environment and the encouragement for people to upskill, to apply what they learn, to not just do things like attend conferences, but actually give them opportunities to apply these different skills. And they need people need to wear different hats and you need to look for that. We talked about how to hire for competency. You mentioned that a lot of candidates, you know, they're prepared for interviews now. They know how to tell the stories already. They have the data points already. So they're like they have their script. Right. But you can you can really evaluate for competency in a more by creating novel situations and asking them, how would you respond to these scenarios in these situations? And, you know, if you get a thorough answer where it's opinionated, it's thoughtful, it's clear they've demonstrated their learning and doesn't it's not just generic. And it depends, it depends, then, you know, you've got, you've got some good, good signs there. And, uh, you mentioned collaboration, right? Collaboration is about these different working models. Uh, sometimes it's more of a service model. Sometimes it's more of a self-service model, but regardless of that, you need to make sure that, uh, everyone understands there's people on both ends. This is not, you know, uh, you know, data as a service or just a platform, right? There are humans in all parts of the equation. Um, and there's always going to be some kind of a mismatch mismatch of expectations which you have to manage people will always want more there's nobody who says oh our our data at our company we have just the right number of people everything is perfect expectations are perfect nobody ever has any misunderstandings you know and that's not just data right that's business in general um but uh you know if you can soften the conflict through the art of co-creation that's where things can really be collaborative and creates a ripple effect that's positive throughout the organization so i thought that That was really great. Juan, I'll pass it over to you. What were your big takeaways?
Juan Sequeda [00:46:25.745] We'll continue. Communication. Data people being misunderstood by the business. That's one of the issues. But we really need to articulate the business impact is important. But in many cases, to get that buy-in, we need to understand that it's intrinsic motivation. Really think about the personal reward. That can be really strong. And tailored to the business impact and the personal impact, that's how we can make this sustainable. So we're like, yes, we need to go make more money and so forth. But then what are the steps? How do we tie things all the way? And then, oh, you say, oh, look, if we do all these things, then we're doing all this cleaning of data over and over again. We can stop doing that because that's just annoying for people and they're not being productive over there. They can now use that time for other things, like be able to go tie that. And people are just tired of doing that annoying work, right? Communication starts with understanding the audience. And without that, you won't understand their personal interest. We do stakeholder mappings. We talk about advocates and detractors, but there's much more about it. Who are the thinkers versus the doers? People like to think but don't do. People who actually do but jump in without thinking, right? Or skeptics versus the believers, right? They need to be convinced. Or people who just jump in headfirst too, right? On the creativity side, I like how you say it's a muscle that can be applied in so many different ways. It's a source of innovation. And it starts with the personal ideas in our heads. And strategy should really encourage to be more creative. And I love your analogy of music and data because it's like music is very defined by rules, but there's just so much in the art and how you can mix and match that. And then we close with the conscious. Why optimism is important. We need to use human judgment and critical thinking, and it's everyone's responsibility to apply that, right? I mean, we have this whole active versus passive optimism, right? And I think, like Katie said, the active optimism is contribute to the best possible outcome. Be that contributor right there. And there's always this feedback mechanism, and we need to have that responsibility to share if things are not going well, especially now with this kind of AI wave that we're in. And I think my favorite quote right now is that we can't afford for the human to be out of the loop. And at the end of the day, everything we've talked about, you kind of say it's kind of obvious, like this should feel natural. Tiankai, how did we do? Anything we missed?
Tiankai Feng [00:48:36.082] No, awesome. No, super good wrap up. Yeah, no comments. Everything's perfect. All right.
Juan Sequeda [00:48:40.362] So to wrap up, we got three final questions to you. What's your advice? Who should we invite next? And what resources do you follow?
Tiankai Feng [00:48:48.122] Yes, I think the advice is pretty much in line with everything we just discussed. But I would in a nutshell say, just try to be more empathetic in your day to day, right? No matter if it's at work or at home. But I realized also with two little children at home that I'm a role model for them. So how I behave in public and how I interact with other people, if I can show them how empathy is leading to the right things and to the better society, let's say, then this is I think, really nice. I feel like because of technology being thriving right now, that people hiding behind screens, it's taking a little bit away from empathy. But I would just encourage everyone to hopefully still imagine people behind screens that are actually humans to actually just relate to them a bit more.
Tim Gasper [00:49:30.162] Well stated. Even us behind this virtual box here and coming into your headphones, all three of us are humans too.
Tiankai Feng [00:49:38.822] Exactly. The second one was who should get invited, right? I would say actually, just representative of being in the German region here from Europe, I would have two people from Germany who might be good. One is Marco Goyer. He's a data strategy expert as well. And Jonas Rashedi. They are pretty, I would say, present in the German data management community. But I think they could deserve that space also on a more international basis.
Juan Sequeda [00:50:10.222] Thank you for calling them out. And then what resources do you follow?
Tiankai Feng [00:50:14.538] Yeah, I mean, I think I follow the usual conferences that there are at CDOIQ and the Dataversity ones, Enterprise Data World and DJIQ, et cetera, et cetera. I like the ones in Europe. We have some specific ones like Big Data London coming up or the Semantic Layer Symposium. All these kind of things are good. Overall, I like to read books. I feel like the two recent ones I read, one is AI and the Data Revolution by Laura Madsen, which is really cool. And the other one is called Microsoft Data Management Strategy from Alexis Plotnikos. And it's really an interesting one because he actually talks very in detail about how he basically built up data management from scratch for Microsoft within Microsoft. And seeing like such an honest and detailed depiction from within an organization that everyone knows is really new to me. because so far I've seen only many anonymized and I myself only anonymized my stories too, right? But seeing that actually in action was a really refreshing reading.
Tim Gasper [00:51:16.238] That is fascinating. I'll have to check that out. And then thirdly, everyone needs to read Humanizing Data Strategy. So grab your copy.
Juan Sequeda [00:51:25.338] Well, with that, Tiankai, I thank you so much. We really, really appreciate you being on the show. And as always, thanks to data.world who supports us and lets us keep doing this on year five now. And everybody, go get your copy of the book. thank you so much Tiankai thank you cheers.
Special guests
