[00:00:00] Tiankai: I think writing que creating dashboards [00:00:02] is definitely gonna be easier and less relevant for analysts, [00:00:06] But I think a lot of people underestimate is if you expect a business user to use a chat bot now to basically get metrics out of a database. the table doesn't on its own know where the technical, fields what the data is you're looking for. [00:00:19] What it requires is the semantic layer. And what it requires is all of the contextual information captured in the knowledge graph. And that knowledge graph and layer needs to be maintained and defined first by someone. a lot more analysts, they will become translators in a different way. [00:00:32] They're not translating data into insights anymore, but they're translating business language into table structures and semantic statements [00:00:39] ​ [00:01:06] In This Episode --- [00:01:06] Phil: What's up everyone? Today we have the pleasure of sitting down with Tiankai Feng data and AI strategy director at ThoughtWorks and the author of Humanizing Data Strategy, leading Data with the Head and the Heart. [00:01:17] He's also spent over six years at Adidas as senior director of product. Data governance and director of digital analytics. [00:01:23] In this episode, we explore the comeback of data quality, [00:01:27] how to organize data teams to improve marketing collaboration, [00:01:31] how to use shadowing to fix broken marketing alignment, [00:01:34] how NLP is changing the data analyst role [00:01:37] and how composable data management works in marketing. [00:01:41] All that, and a bunch more stuff after a super quick word from one of our awesome partners. [00:01:45] ​ [00:02:52] Phil: thanks so much for your time today. Really excited to have you on the show. [00:02:55] Tiankai: Thanks for having me. Glad to be here. [00:02:57] Phil: I read your book cover to cover. Can't [00:03:00] really say cover to cover because it was the ebook version. [00:03:02] I got a chance to search after I read it for specific questions here. So excited to talk to you about this overlap between marketing operations, marketing technology and the more data operations data professional side of the role. Um, I'm, I'm curious because like [00:03:18] 1. How Data and Marketing Create a Symbiotic Relationship --- [00:03:18] Phil: you actually started your background in marketing, right? [00:03:21] Like before your illustrious career in data management and governance. You spent a lot of time at a marketing and bi boutique agency. [00:03:29] Uh, like your LinkedIn said, digital strategy, market research, brand marketing, social media. Eventually you got into more performance management projects. We also had a short stint as a part-time head of marketing for the Germany chapter of the Data Management Association. Fun one, just to get us started here, like what makes data plus marketing one of the best matches out there? [00:03:49] Tiankai: Yeah, absolutely. I mean, I think, um, what I always, uh, felt like even growing up was, uh. I loved advertising and I loved to have the [00:04:00] fun with it, like all of the cliches looking at advertisement, but also understanding why it makes me wanna buy things, right? That that whole feeling. And I always felt like there needs to be some logic behind it. [00:04:09] Like even when I was growing up, I thought like, that has to be decoded somehow. And of course there's science behind it. Science is a way of data as well, of course, that you're collecting. But when I then chose to study, um, and chose a major, basically I decided for doing a mix of marketing and database systems as basically my segue into trying to do both. [00:04:30] And that really worked out and that made me actually start my career in marketing analytics and analyzing marketing data. And I always loved the two worlds, right of, on the one hand side marketing as being seen as this creative field and everything basically being there. To have ideas on how to convince people about being a better version of themselves, et cetera, et cetera. [00:04:51] Well, on the other hand, you have all these numbers and metrics and everything, and you try to make sense of it all, of how you can put even more, and basically balancing [00:05:00] that right side brain and left side brain right? That kind of always make me feel like this is the best way to do it. Right? You have a lot of thoughts. [00:05:06] You have a lot of gut feeling. What if you could actually rely on proof on one hand side to make better decisions, but build on it, then be creative and actually be great at storytelling and sell things, and so on and so on. So yeah, that for me has always been the best mix, I think. And I mean, I would even say nowadays working on data. [00:05:26] Um, a lot of the collaboration comes from marketing your team properly and marketing your function properly too, right? So even now I'm applying a lot of the marketing techniques I learned in the past to convince my stakeholders to work with me and have less resistance about trying to work with me. So all in all, it was a good investment and, um, I still really, yeah, by heart, I'm still, uh, love marketing, right? [00:05:49] I'm a marketeer. [00:05:51] Phil: Very cool. Hey, it's cool to hear that the, the creativity was one of the, the inspiring, uh, actions to get you to, to study marketing and, and dive more into that [00:06:00] space. [00:06:00] 2. If Data Governance Is the Jedi Council, Marketing Ops Is the Rebel Alliance --- [00:06:00] Phil: You obviously talk a lot about that in your book, like the whole creativity side [00:06:04] of being human and not just the, the other side of your brain and focusing on the data. Um. Creativity is a big part of your life. Obviously in art, you're, uh, a musician. You [00:06:13] love, uh, rapping and, and singing. I checked out a lot of your raps and, and the stuff that you've got on YouTube and, and some of the TED talks that, that you started on, on the piano. And one of your songs, you have a verse that says, if data is the force, data governance is the Jedi Council. [00:06:30] And I loved it and it made me think, all right, like what is. Marketing operations or marketers in that analogy, like who do you think is the marketing operations or the marketing team that's activating that data that's been governed and clean and, and like actually growing the business? Are we the Jedi Knights, the train specialists who wield the force in a practical way to accomplish missions or maybe rebel alliance, like the scrappy, resourceful team that builds systems and strategies to beat the empire? [00:06:58] What are your thoughts there? [00:06:59] Tiankai: Yeah, [00:07:00] I mean that, that's a great question. Um, I do, my, my initial intuitive answer would be both depending on the situation, right? Because sometimes you need to be the Jedi master to actually fight the good fight, but then sometimes you have to be against the mainstream too and do something that maybe not is the most popular way of doing it. [00:07:16] But you need that anyway because in your heart you believe it's right. So that kinda way I feel like, uh, both applies. Um, but nowadays, um, I think just given the maturity also of MarTech maybe as an industry and of being basically a field to work in, um, it's more jeon because there's more mastery in there now, right? [00:07:36] And there's more experience and more best practices to be shared. Whereas in the past it felt a little bit more rebellious, right? Like you were the underdog and it felt like not many people working in it. So everyone tried to make sense of it, but now with like years of experience, it just feels a bit more mature. [00:07:50] I think that's why I would. Tend towards more of a Jedi master at lg. You. [00:07:55] Phil: Very cool. It makes a fun, uh, way to create a, a header image for this section of, [00:08:00] of the interview folks watching on YouTube. Like there's like a, a video, uh, chapter, uh, image for each of them, and I use them for the blog post too. So, um, oftentimes I'm just like trying to think of creative ways and it's like a balance scale or it's like a flashlight or a loop. [00:08:16] And I think that, I don't think I've done a lot of, uh, Jedi [00:08:19] Tiankai: Yes. Now you have the one opportunity to do that. Go ahead. [00:08:24] Phil: Yeah. Love it. Um, I, [00:08:26] 3. How to Organize Data Teams and Improve Marketing Collaboration --- [00:08:26] Phil: I grew up in SAS startups where, you know, there wasn't a data team [00:08:31] yet, and this is going back like 13, 13, 14 years now. Data didn't really have an owner at that time. Like the technical tracking components of data, customer data, and like the product to marketing tool integrations were a mix of ownership between the engineers that were building the product. And the marketing team. I was, you know, using an IPAs tool or trying to dive into API documentations and make sense of like rest APIs to connect stuff together. [00:09:00] Things have changed dramatically now even in most startups. Like there is a data team or at least one data person. I [00:09:07] know you talk about that in your book a lot. [00:09:08] Like the one person did a team, [00:09:10] um. But I want to ask you about like people, process, tech and data. That's typically the mandate of marketing operations or like the revenue operations team within a company. The data component's really interesting there because in some companies I worked at marketing is, uh, like Marketing Data is owned by marketing operations. And a lot of other companies, bigger companies specifically, it's fully transitioned to a central data team. They own bi, they own all of analytics. If you're the marketer and you want to build a new dashboard or track a new metric, you have to log a ticket. But some teams have a decentralized approach where there's like a marketing analyst that's embedded and sits on a GTM team. What's the best format and curious your thoughts there. [00:09:52] Tiankai: Yeah, I mean, uh, I think that that is a good question, but there is unfortunately not a one size fits all answer to it. Right? It really [00:10:00] depends on our organization, on the maturity, on the culture of an organization, on what is the right setup. But normally I would say right, wherever I. The decisions are being made, try to be as closely as possible there, right? [00:10:13] And but you have to balance that proximity to the business decisions and the actions that you're taking to improve the business with consistency and managing equally all of the data efforts that are happening, right? And this is the usual trade off between centralizing or decentralizing your data people and data experts into either all being centralized and being consistent with data, but then having that hurdle to actually work with a business because they're so far away from them. [00:10:38] Or everyone sits in the different business functions, but they never talk to each other. So all data experts do whatever they want and it's chaos. So finding that balance is really hard. Right. And I think basically trying to then intentionally find the right balance is the right thing. However, I think the, the only right thing to consider, and I I talk about that in my book, right, is more about the mindset itself, rather than [00:11:00] actually where you sit and what the operating model is, right? [00:11:03] As long as you have shared goals. It really shouldn't matter in what function you're sitting or what team you're sitting, but if you have goals that are shared and everyone contributes to the same goals, then it's in everyone's interest to work together. Right. And then you don't have to use it as an excuse anymore, right. [00:11:19] To say I sit in a different team, or I sit in the same team, um, to justify uh, not collaborating or collaborating. You just want to collaborate anyway because it's the right thing to do. Right? You cannot do it alone anyway, so let's team up. We reached the goal together and we all benefit from it. So yeah, there's a lot to think about there regarding incentives, collaboration, uh, culture, uh, but that should be the ultimate goal, to actually have everyone work towards the same goals and not, uh, goals that are conflicting with each other. [00:11:49] Phil: Yeah. Yeah. Makes a lot of sense. What advice would you have for marketing operations folks? Or marketers when they collaborate with their data counterparts on the data team, [00:12:00] obviously they're, you know, the data team is a lot more technical. They're way more data literate, but they probably lack some of the business understanding that the marketers do, especially the marketing application, uh, understanding side of things. [00:12:13] I know you talk about this in your book, but I'm curious to get your, your advice there on like, empathy for, for the data team. [00:12:20] Tiankai: Absolutely. Um, I would say, let me first describe that a good data professional would ask you for the business context proactively, right? They would try to understand you first before they're trying to make you do things or I. Before they ask you for certain inputs that you don't know the context of, right? [00:12:39] Like, why are you asking me all these questions? Um, but if they don't do that right, then if you are the one who is marketing, uh, working in a marketing function, then you could also try proactively sharing that context with them, right? You can ask them if they are willing to understand you first before they ask all these weird questions or ask you to all these different things. [00:12:59] Um, you would [00:13:00] wish for them to understand you first, so then they know what you are going through and what your day-to-day looks like. So you speak at least a little bit the same language, right? And this is, uh, as you, uh, rightfully pointed out about empathy, right? It's not me against you and not having the intention of why are you against me and why doing all of these things to me, but rather. [00:13:19] Let me try to help you understand me, and I want to understand you too. And if we understand each other, then we can work together no matter what, because at least we get each other now. Right. But that's a much more human thing than anything. It doesn't, it's not even just work related. It's generally if we understand each other, we just work together better. [00:13:37] Right. Having, um, just shared passions for things, um, having a common understanding that's just very normal and it's basic human connection that we want to have. [00:13:47] Phil: Yeah, I, I think that's great advice. It makes me think of a couple of situations in, in my in-house career where, uh, it, it wasn't always the data team, but you know, marketers often needs to work with engineers [00:14:00] and the engineers will work on the data team, the engineering team, the product team, or sometimes even like IT [00:14:05] security systems and, and governance, right? [00:14:08] Um, and there's often a ton of like overlap when it comes to customer data and the more technical side of things like email deliverability, we talk about that on the show a lot. You have to have, you know, DNS records in there to do your DAM authentication and that overlaps with it and governance and security. And there's often like an understanding from a marketing application side of things that this is the best practice based on my experience, based on the research that I've done. We should have multiple subdomains for how we send emails. We'll separate the shady stuff that our sales team is doing from our corporate domain to, to save that reputation and not affect it. [00:14:47] So I need this subin that subdomain and they're like, [00:14:49] 4. Handling Healthy Data Conflicts Without Crushing Creativity --- [00:14:49] Phil: I've faced a lot of resistance sometimes with engineers when, you know, I entered the discussion and I am like. Trying to be empathetic about it and saying like, you guys don't have all the [00:15:00] application knowledge that I do in the marketing technical world. [00:15:03] Here's everything I know. Here's all the context. Here's the background on why I want to do this. But I've often faced like specific resistance from folks because it's a technical area and they think that they own it and they think they understand the world well enough that they have, you know, an opinion to share and, and we have like back and forth and they don't wanna agree with the decision. [00:15:24] Like, what advice do you have? Because like your first answer was just like, you know, just share the contacts and explain what you're trying to do. What advice do you have when after you share that context there's disagreement on where to go and, and what are the next steps? [00:15:39] Tiankai: It's a, it's a really interesting one. Um, because even if we just think about the general split between business and tech, right? Then of course, business people are the business experts and tech, other tech experts, right? But neither of them can work without the other [00:15:54] either, right? So like a tech, if it's not being used for business, it doesn't have any value. [00:15:59] When a [00:16:00] business person doesn't use the tech, they're gonna have a really bad time because they're not enabled anyway. Right? So it should be in their natural interest to work together. So basically disagreeing with each other is a little bit under. The underlying problem with that is that you are undermining each other's expertise. [00:16:16] It does feel a little bit like, what are you, who are you to tell me that I, I don't understand tech enough, [00:16:23] right? But in reality, I am a tech expert. You still need to use my technology. Actually, we should make it work, right? And if we haven't considered each other's point of views yet, then now is the time to consider it. [00:16:36] Because in the end, we need each other to be considerate enough to actually basically work together. So. That, that whole principle of always assuming positive intent, right, which is one of the principles of going through life, I would say, and being a positive person, um, is easily then mistaken for always assume bad intention, right? [00:16:56] So the moment you're saying something that's just human, you are interpreting it [00:17:00] as you are actually evaluating my expertise, whereas it's just a disagreement on, this is not how I do things. Can we please work together to make this [00:17:07] work for both sides? So if we can change it around, right. To say disagreement is to is a way to find common ground, and it's not about focusing on what makes us different, but it's focusing on what actually we have in common. [00:17:21] Then maybe that can lead to a better outcome. [00:17:24] Phil: Yeah, I feel like healthy disagreement is, is something that comes up a lot in, in your book. And actually you like checked out a lot of your posts on LinkedIn and, and what are your most famous posts? [00:17:35] You went like mega viral on that one. Uh, you shared, uh, an image from Marketist to kind of illustrate this data contradiction element. [00:17:44] You, you basically wrote about like how. Agreement doesn't always mean that you're on the right path. And you emphasize that like healthy disagreement and psychologically safe environments leads to better collective decisions. And, and, and like you when approach constructively obviously. [00:18:00] What about like when data reveals a narrative that contradicts a business team's creative vision marketing, maybe like how do you present. In that boardroom with a marketing team, like without crushing their creative enthusiasm. [00:18:17] Tiankai: Yeah. Um, I mean, that part actually is really about storytelling, I would say, right? And ironically, um, the ones who are actually expert in storytelling, while the creative people, right, they're the ones who need to be told a story now that makes them not explode or getting really upset about it. And all of a sudden you have to tell the right story to a bunch of expert storytellers. [00:18:37] So that makes it very tricky. But, um, all in all though, right, the, the thing is again, um, that the facts are contradicting. What they came up with is, uh, is the fact and what happened, but there's um, what you can communicate and how you communicate it. There's always a choice, right? And the what is probably already fixed because this is the message you need to tell, but you can [00:19:00] always choose how to tell them. [00:19:02] Again, this is where I feel like also having worked with marketers that, um, they of course want to be valued for their creativity and for their ideas and what they came up with, with the whole concept. If the data is now basically contradicting that, then the message shouldn't be, you were wrong, but the message should be, is there a way how we can actually tweak your creative? [00:19:26] So actually fits better to the findings that I have, right? My audience actually doesn't like this. They actually enjoy this. One more, maybe we could change the testimonial to this one, or my, the audience we found out is, uh, spending more time on this social media channel than the other one. Maybe we should not plan so much on this channel, but the other one, but giving actionable guidance, right? [00:19:46] That is not saying stretch the whole thing. It doesn't make any sense. But more like constructive criticism, right? As in the facts are saying that this and this. So our suggestion would be. Knowing how much work you put into it, could we tweak it in that direction [00:20:00] instead, and this has a higher likelihood that it should work. [00:20:02] So yeah, basically that takes conscious storytelling, right? You need to actually spend some time and tell our story, right? But I would always say, even if that takes a lot of time, because that is what people complain to me about, right? Why should I invest so much time in storytelling? It's just a table. [00:20:16] They can read it themselves, and now I'm creating all this power around it, right? But the thing is, if you don't. Tell the story and you just saw them on the table and they realize what's happening, then cleaning up and repairing that relationship might take a lot more time than the one you invested in the, in the beginning, in storytelling. [00:20:34] So I would say rather invest there and try to be proactive in, uh, maintaining the good relationship and being constructive and being a good collaborator than, um, actually having to fix it afterwards. [00:20:47] Phil: Yeah, that's great advice. I, I think one of the most important stories to tell, uh, as someone who's like reported and measured marketing activity throughout his whole career is. One of the most important to tell is it's okay [00:21:00] to have negative results. Like not every campaign [00:21:03] Tiankai: That is true. [00:21:04] Phil: win and sometimes. The results, the insights that we uncover from the losses and the misses end up being just as important because from those come a better campaign idea. And so it like we switched the narrative from it being like, your campaign really sucked to, here's what we learned about why this didn't necessarily resonate with folks, or this is the methodology that we use to measure it. [00:21:28] It's not perfect. Here are the holes in it like. We have a confidence interval on it. Like we don't think that multitouch attribution is the perfect way to measure this, but based on the trackable touch points, like, so there's a way to kinda spin it because Yeah, I, I think marketers are, are very sensitive when it comes to the, the creative side of, of the gig. [00:21:46] But yeah, I love your answer. [00:21:48] Tiankai: Absolutely. But I have to also say maybe to that note, right, there's a difference of when you use data in a campaign lifecycle, right? [00:21:55] Because if you use it earlier during the creation, conservation of [00:22:00] a campaign, and it's more about consumer insights, you are still in the planning phase. Then you can, you have a choice to listen to it, right? [00:22:06] It doesn't have to be bad. You haven't performed yet, right? You haven't run the campaign yet. So you can still cha cha change your mind and try to do things better and that looks good. But if you have run the whole campaign, you decided to not listen to any data or ignore data at all, and now you're performing badly, not even the data can save you, right? [00:22:22] That we cannot magically create better metrics for you just because you didn't listen up front. And this is why, right? As as the earlier you can actually work with data teams or data. In your campaign lifecycle, the better off you would be, uh, right, because then, um, ideally you worked on data informed decisions. [00:22:44] Right. And then they turn out to be good and you actually have better KPIs and metrics as well. Yeah. But it's not always the case. And this is where the tricky thing comes, uh, from, right To try to look good, but the data doesn't, uh, support it. And then you have this bad situation. [00:22:59] Phil: [00:23:00] Yeah. Yeah. We, we try to empathize like not hiding your wins and only showcasing or not hiding your losses and only showcasing the wins. But it's, it's easier said than done in, in [00:23:10] this economic climate, right? Like you don't wanna. Just showed about the losses, but, uh, it's, it's, a tricky one for sure. Uh, what comes to mind if I ask you like, what's the most valuable disagreement in a business context that you've lost, and how did it change your perspective or your approach to cross-functional collaboration? [00:23:30] I. [00:23:31] Tiankai: Yeah, that's a great question. I think the, um, yeah, I would say the most disagreements I had, um, without having to be specific about names and brands that I work with, uh, might get sensitive. I. We're about the balance between commercial success and doing the right thing societally or [00:23:53] uh, sustainably, right? [00:23:56] Um, whereas I would find consumer insights that are supporting that. People [00:24:00] want to be greener now and they want to be better and they want to support a certain minorities more, and they want to have a good cause now and things. Whereas on the other side, they decide to just put everything on discount and to still choose the cheapest way of producing things to sell them in the easiest way. [00:24:17] And those were always the disagreements kind of that I were facing. But I realized that there's always a moment, um, and a momentum, right, for these kinds of conversations. And I cannot keep on being just very stubborn about certain things that I'm trying to make. 'cause otherwise I come across as, um, yeah. [00:24:40] Hard to work with, right? Difficult to work with basically. Uh, because I cannot accept no. Right. And I just keep on pushing and pushing and pushing until it doesn't really end well. So I think the, the one thing I learned about it is that sometimes it's not about the decision being wrong, it's just not the right time or the right decision. [00:24:59] And [00:25:00] I learned that if I bring it up at a different time, when actually, um, things, people are in a different mood and, uh, timing wise, it's better that the right decision might be taken much more likely. And just because I am, um, like running out of time or I feel like I need to urge the decision now, it doesn't mean the rest of the world can follow me. [00:25:19] Right. So I think that that is the main learning I had from that. [00:25:23] 5. How to Use Shadowing to Fix Broken Marketing Alignment --- [00:25:23] Phil: In your book, you talk a lot about advocating for and spending time with other teams to, [00:25:28] to understand their world when there is conflict like this and, and when you do have disagreement because maybe you have this theory or this like innate idea and after you've actually spent time in their worlds. You've opened up to, like your idea maybe not being perfect and not having all the information to, to kinda like come up with, uh, with the basis of your idea. Um, you shared an example, uh, in your book. You also talked about this on LinkedIn also about shadowing a brand strategy team. Uh, [00:26:00] you did this one day per week, forget what company this was at. [00:26:02] But, um, talk to us about this idea and, and how you actually operationalized it. [00:26:07] Tiankai: Yeah, I mean, um, the premise is simple, right? For me to be able to understand, uh, my stakeholders or my collaborators, I need to see their day to day, but I cannot expect them to just block days of their own calendar just to teach me about [00:26:22] it, right? I. The easiest way probably is to just observe them. So shadowing them in that case is probably best way for me to just observe, capture, and have the takeaways that I need from observing what's going on to then know what they're going through in the day to day. [00:26:37] Uh, practically that means time commitment to right, because on those days that I'm shadowing someone else, I'm actually not doing anything else but observing them. So actually you need that commitment from your leadership, or you need at least the buy-in into the way of working that you choose to do to actually do that. [00:26:55] But again, right? I mentioned that before, the time you don't invest actually in understanding [00:27:00] them, uh, turns out to be much worse afterwards because there's a lot of more conflict that you need to deal with that is all costing a lot more time. So the one day that maybe you'll spend per week for let's say three months turns out to be preventing all of the issues afterwards because you understand each other now finally perfectly right? [00:27:18] That could be a good, um, balance, right, and a good, uh, return on investment in the end. But we need to think about it this way too, right? That creating intentional understanding for each other and to spend the time and the efforts to do that, um, that takes the right mindset. And culturally, I think we should just go towards it more. [00:27:39] The bigger organization is the more anonymous you can get, right? And then the bigger the organizations, the also you, the fee you feel. There's so much more you can do. And that creates actually the fatigue that we think we have a lot of choices. And then we focus on nothing, but we focus on everything, right? [00:27:55] We just like, I have to be for everyone there. And, [00:28:00] um, then you start everything, but don't finish anything. So if we can go back a little bit to focusing on specific stakeholders and calibrators, I think that might help. [00:28:08] Phil: Yeah, I, I think that's great advice. Uh, oftentimes marketing ops teams and marketing technologists are, are in a similarly collaborative role as, as the data folks, [00:28:17] Tiankai: Right, [00:28:18] Phil: We, we deal with like sales, we deal with data, we deal with products. And sometimes when we're doing like lifecycle stuff lower in the funnel, we deal with a bunch of other people too. And so like, there's a lot of context switching and kind of your, your answer there and it definitely empathize with that. [00:28:32] ​ [00:30:24] Phil: one thing that jumped out when I was researching your background and, and some of your experiences, the, the amount of time you spent with marketing teams with big names like Coca-Cola. Volkswagen and Adidas, you were six years at Adidas. Um, what's maybe like, like take us to, to some of those like discussions or like campaigns and the question I have for you, but like, feel free to, um, you know, just take us through that experience. Like what, what's the most counterintuitive data insight that you've defended that marketing team at some of these big companies, initially rejected, but later embraced maybe because of your, your [00:31:00] storytelling ability. [00:31:01] Tiankai: Right. I think I can answer that very easily. Um, the insight sometimes was you have too many campaigns going on at the [00:31:09] same time and you're all with the customer. And the reaction usually to that is no. But all of our products are important. We cannot stop any of the campaigns and because everyone, uh, the probably you mentioned before, everyone invested a lot of time on coming up with these campaigns and now they're ready to run it. [00:31:26] But you cannot run, let's say, 10 campaigns in one week because the customer will be completely overwhelmed by how many things you're marketing at the same time. Right. Um, this was counterintuitive for the marketers and it actually pointed to a much bigger problem because it wasn't really convincing individual marketers. [00:31:44] It was they having not aligned with each other and synchronized the calendars to actually have a proper sequence of, uh, campaigns, but rather just decide it to put it all in one. Using multiple tens of social media channels all over the place to all send [00:32:00] different messages. Um, that I had to basically use that inside multiple times. [00:32:04] Right. To try to explain. Now look, even the survey data says that the, um, respondents don't know what you're standing for because it's just all over the place. So is that kind of thing. That happened a few times. Um, I would say because on an individual level, then the, the better you get and the longer you work in that area, the more convincing you get. [00:32:24] And you know how to tell people data insights. Right? But, um, once it's actually about them not being aligned with each other across different departments, then it gets blown up. Right? All of a sudden it's a big collaboration problem. And that was an interesting one. Always. [00:32:39] Phil: Very cool. So almost like a frequency. Discussion there, like you mentioned survey results, like what other mechanisms came into play to, to being able to like surface that data to the marketing team? Say like, Hey, we're like inundating our potential customers with messages we're confusing them. Like how, how did you prove that to them?[00:33:00] [00:33:00] Tiankai: Yeah, I mean, uh, I think survey data was the most obvious one. A lot of time when I spent in that boutique agency that you mentioned beginning was actually combining traditional market research data with digital sources, like [00:33:11] social media data, social listening. I. Uh, web, web analytics and paid media research and these kind of things altogether. [00:33:18] So that's how we could combine the learnings into, uh, have proper consideration of a comprehensive consumer insights, right? Um, but also the search behavior. For example, Google, right? You could just check for competitors, how they are being searched for, specifically for the newest products, whereas for the brand that you're working with, all of a sudden it's like all equally little of different products. [00:33:42] Um, that are not all competing at all with a competitor in [00:33:46] terms of search volume. Right? And search volume, we all know is a big indicator for interest. And, um, you basically have distributed all of the interest into much smaller chance for different things. Um, instead of focusing on bigger [00:34:00] things that you stand for to then create that interest or hype or hotness for your brand, right? [00:34:05] That's. [00:34:06] Phil: Very cool. Yeah, that's a, it's always a balance between the, the qualitative stuff and the quantitative side of, of marketing measurements [00:34:14] Um, in your book, you actually talk a lot about this concept of. Being data informed versus data driven and making decisions there, and you highlight the balance between human expertise and analytical insights. What are your thoughts on like the hype with AI right now and how marketers and, and data teams determine which campaign decisions should remain human led versus machine led? Like, we're seeing a ton of that. It is been around forever, like ML'S been in marketing. For decades and like this whole like send time optimization, um, you know, figuring out channel optimization, the right time to send a message. [00:34:53] Like that tech's been around for a long time, but now it's getting a bit easier. Like AI agent orchestration is [00:35:00] kind of like blowing up the hype cycle and a lot of folks are asking themselves like, what needs to remain human led and what can we just kinda like give to the machine? What are your thoughts [00:35:09] Tiankai: Yep. I mean, I, I think I still stand to the point that I think we live right now in the time where a lot of the tedious, routine and recurring tasks can be automated and very well automated by ai. Right? And even with AI agents, the whole premises that, um, you predefined patterns and that based on certain rules and patterns and a agent will do exactly the same stuff over and over again, um, depending on what you're asking it to do. [00:35:36] Right? But I always think that that means that we can actually shift our tedious work and recurring manual work to AI more that should give us, uh, theoretically more space to think, make better decisions and be more creative on a human way. Right? And that is, I hope, where it goes to also in marketing for example, right? [00:35:58] Where if we [00:36:00] don't have to spend. Time manually filling in paid media, uh, uh, cost overviews anymore. I mean, that's a bad example, but you know what I mean. Right. Then we can actually think about what are generally is what we should do with paid media. Maybe there's some innovation kind of idea that we have on what to do with, uh, the paid media channels, et cetera, et cetera, to do it in a better way. [00:36:21] Right? And so jobs will change. Certain jobs that were only operational before might, uh, move and being replaced by ai. But that makes us all rely then hopefully more on our brain again and our heart. Right. And we should then all again, put more time into the thinking and making the right decisions. [00:36:42] Phil: Very cool. I love that [00:36:44] 6. The Comeback of Data Quality --- [00:36:44] Phil: At, at your current company, your role is focused on both AI and data governance and, and the strategy behind that. There's obviously a ton of overlap, like one doesn't happen without the other, [00:36:56] and like the empathy side of that for marketing [00:37:00] operations folks is that we. Are often like an underfunded team within the marketing teams, like 20 person marketing team and like one marketing ops person doing like all the MarTech implementation, supporting the people in the process. [00:37:14] And um, you know, a lot of folks just like look at the data and they're just like, why is this like data not fixed? Why is this data dirty? Why do we have duplicates everywhere? And it's like. I'm one person supporting 20 freaking humans and there's a ton of shit on my plate. And so AI now is kind of changing that discussion a little bit. [00:37:33] And marketing ops folks finally now have this like thing that's helping us and supporting us when all we want to do is clean the data. Like all we want to do is sanitize the data and the CRM and make sure none of it's duplicated now. Other people care about it. Like it was so hard to go and chat with like senior leaders and be like, this thing that we're pitching to do next quarter is cleaning our data. [00:37:56] And, uh, we're de-duplicating records and we're doing, [00:38:00] uh, you know, um, uh, ID resolution. And we're gonna combine all these sources. And like the senior teams are just like, why do we care about this? [00:38:07] Like, how much revenue is that gonna give us? Now we can finally say. It is gonna give us revenue because here are the AI applications that we can put on top of that clean data. What are your thoughts there? You see a lot of overlap with what you do day to day. [00:38:22] Tiankai: Absolutely. I think this is really the key. I, I also, um, this is, uh, I call it the comeback of data quality [00:38:28] because I. [00:38:29] Phil: Hmm. [00:38:29] Tiankai: Before, um, data quality was not sexy. Everyone was focusing on, everyone complained about it, but no one wanted to take care of it. Right? And now that AI is there and everyone realizes that, uh, AI with bad data is just bad ai, right? [00:38:42] All of a sudden the focus is much bigger. It's like we are bad. Data quality is hindering all of our innovation here. And, um, if we cannot compete with the rest of the world in ai, then what are we doing? This needs to be the highest priority all of a sudden. And that puts a lot of pressure on it. Right. [00:38:59] And, uh, going back to [00:39:00] the basics and actually focusing on air quality again, is the right move, I would say. Right. I have to say that, that also a lot of, especially in the gen AI space, right, there's a lot of pre-trained models. That only needs the context of your organizational data and doesn't need to be trained completely by your own organizational data. [00:39:16] So, um, there's um, always a certain impact that you need to deal with. But even then, still, right, you want to tailor it to your organization that that means you need to provide it in the high quality contextual data that you have for any AI model to probably work. So, yeah, absolutely. Um, it's, uh, now that everyone wants to do it, we all get to do data quality again. [00:39:35] So good for us. [00:39:37] Phil: Data quality's time to shine again. [00:39:40] We're finally getting our time in the sun. [00:39:42] Tiankai: exactly, exactly [00:39:44] Phil: It's funny that like most. It's obviously like more than one team, especially in bigger companies that owns data quality. Um, but like, you know, marketing operations folks, the acronym is like mop often like MOP and the icon [00:39:59] for a lot of folks [00:40:00] is a mop bucket because we're cleaning data and we're in the background and folks don't really care about what we do. [00:40:05] Now all of a sudden it's like, oh shit, this is important work. Like [00:40:08] our AI isn't great without. Marketing also focusing on, on the marketing data side, but you obviously have a bigger mandate of this for data governance, for, for the whole company. Um, curious to ask you your thoughts on like, just the overlap between like, customer data obviously is a big part of marketing and activating that stuff. [00:40:27] Figuring out what's working and. When you're talking about data governance and AI strategy for the company, you're talking about the AI strategy, the data strategy. Hopefully these things are kind of like being combined there a little bit, but yeah, chat with us about like that, that overlap. What do marketing ops folks need to know about what is important in the data governance world in the next like five, 10 years? [00:40:49] Tiankai: Yeah, absolutely. I think customer data is a really great keyword because the problem with, or the challenge with customer data is that the creation of it is still highly manual, [00:40:58] right? Either the [00:41:00] customers, uh, for the first time signing up with their own data, that every human error is possible, right? You can mistype your own name or your address, whatever, right? [00:41:08] Or a salesperson that is creating leads is also typing in the data manually into somewhere, and that creates a lot of data. And if. The first record ever of customer data is already wrong, that you don't really have a comparison towards anything to say what is correct, and without that reference, data quality management is very difficult because you cannot even define what's correct, looks like. [00:41:30] Imagine just your last name being now typed in the wrong way. And I don't have any reference about you else. How would I know what the correct last name of you is? Right? And, um, that is where a lot of the issues come from. So to avoid that human error and that then cascades down across the whole data life cycle, wherever it ends up in BI and whatever reportings, um, actually human uh, data entry needs to be better. [00:41:53] And that can have two sides. One is the typical training people to be more diligent, right? And, um, giving them [00:42:00] more guidance to do the right thing and just be more diligent about all of it. For the second one, technology enabling in to do it right with guardrails, right? To have some type of a, a check as in like, can you just check, this is an unusual name. [00:42:14] For example, are you sure that this is the right one? Or your address might be in the wrong format. Could you check if this is actually what you mean? Like some kind of technological guardrails to actually help you. Right? And then data quality management can be much easier because you start already with beta data from the very beginning. [00:42:30] And then it cascades down to the other things. So, uh, to any marketing operations person, right? Um, the more you can help the people that enter the data for the first time to already put in the right data, the less you have to deal with all of the rest and getting the requests to clean up data again afterwards because it was done already from the beginning in the right way. [00:42:53] So, yeah, don't be the mop bucket. Try to, uh, give other people the mop first to already do it clean, right? You know what I mean?[00:43:00] [00:43:00] Phil: Very cool, yeah no. Good, great point there. And that's not even like you cover the manual data entry part there. There's a whole other bucket of customer data infrastructure and [00:43:09] how we're tracking that data a bit more automated, but there's a lot of errors [00:43:13] Tiankai: That is also [00:43:14] Phil: there. And we unbox the whole like privacy discussion there. And so yeah, we, we don't need to unpack that there [00:43:20] 7. How Natural Language BI Tools Change Data Analyst Work --- [00:43:20] Phil: because [00:43:20] I want to ask you about. Um, like this whole like AI replacement of, of jobs. Like a lot of folks, you know, there's like this opportunity and excitement with ai, but there's a ton of fomo and like fear around shit. Like, am I falling behind on stuff? [00:43:36] Like, is my job gonna be replaced? Like, am I really adding value? Should I be doing this or should I be doing that? One of the roles that's really interesting with this holy topic of replacement is analysts and specifically folks that are focused on like BI and reporting. A lot of folks think that like some of the jobs most at risk for AI disruption in like tech, I. A mix of reporting and, and BI role. [00:44:00] And I think that like the front end dashboard and vis part of this is already a lot easier and that this space has changed a ton over the last couple of years. Uh, NLP Natural Language Processing is allowing, uh, folks to like, interact with their data, like have a chat experience. Non SQL experts can finally like, ask clean English questions to their data and build pretty quick dashboards. Um. There's a whole other category of like recommendations from pattern detection. Uh, there's like predictive engines. Now, I'm less convinced that some of these tools are gonna be able to like, help too much with the data prep and the modeling, the pipelining side of things. [00:44:37] But what are your thoughts on, like everyone's saying that, you know, analysts are one of the jobs most at risk of, of being replaced by ai. [00:44:46] Tiankai: Um, yeah, I think, uh, that, that's for sure. It's too much of a generalization, [00:44:51] right? Because. I feel like that is, uh, describing as if analysts are only writing SQL statements and getting tables on and sending into [00:45:00] stakeholders where in reality, of course, there's a lot more there. Right. And we all know. It's like even just tell me what the business problem is you're trying to solve. [00:45:07] I. Um, coming up with the query itself takes a lot of thinking work and, uh, communication work to even get there because you need to first know what you're even looking for, right? And then find the right tables and then generate it, and so on and on. Right? Um, I think writing que creating dashboards might get easier in the, uh, in the, uh, future, right? [00:45:25] And we talk about vibe coding nowadays, right? A lot too, where you don't even have to just, uh, check the coding, where you just kind of do it with AI and then you'll focus on the end and the overall bigger picture, right? Um, who knows where this is going, but, um, that part is definitely gonna be easier and less relevant for analysts, I would say. [00:45:45] But then that puts more actually focus on enabling business users to use that conversational BI that you just mentioned. Right. But I think a lot of people underestimate is if you expect a business user to use a chat bot now [00:46:00] to basically get metrics out of a database. Right. Then it doesn't, the table doesn't on its own know where the technical, um, fields what the data is you're looking for. [00:46:11] What it requires is the semantic layer. And what it requires is all of the contextual information captured in the knowledge graph. Right? And that knowledge graph and layer needs to be maintained and defined first by someone. Right? And this is also where my prediction kind of goes. I feel like that a lot more people of analysts, they will become translators in a different way. [00:46:29] They're not translating data into insights anymore, but they're translating business language into table structures and semantic statements to enable then the business users to use chat bots more in the future. Um, so all of that is just shifting a little bit of responsibilities, but it requires a similar mindset and a similar skillset still. [00:46:48] Phil: Very cool. Yeah. Uh, you know [00:46:50] 8. How Composable Data Management Works in Marketing --- [00:46:50] Phil: , MarTech is hard enough to keep up with these days, but you mentioned a ton of different, uh, technology there with the, the semantic layer and, and, and data management and. The data landscape is moving probably faster than, than MarTech from data warehouses to data lakes, to data match to data fabric. [00:47:08] And you mentioned semantic layer there, realtime dashboards are, are popping up and uh, a lot of folks are talking about data observability and the interlock with, with marketing there. Um, can you help us like bring clarity to where we've been with data management and where we're going with customer data management? [00:47:26] Is it fair to say that the trend is moving away from. Monolithic architectures that are like platform based all in one tools, more towards a composable distributed system with a lot stronger governance, more control, more automation, more self-serve capabilities with point solutions. A lot of it works Are implementing like a hybrid of the two and from multiple paradigms, like what? [00:47:50] What are your thoughts there? [00:47:51] Tiankai: Uh, absolutely. I think that is a very valid, um, uh, assessment of the current trend. Um, I mean we started all with like, uh, [00:48:00] data warehouses being very central, right? And everyone had to basically get data from there. And you had a central data team that were servicing everyone to do it. And then it shifted into more of a data lake, which meant that it's still centralized, but unstructured. [00:48:14] So people to get it, they needed to structure it for their own reasons. You started decentralizing a little bit. Now with data mesh or data fabric or whatever, you kind of start decentralizing even more, right? We can talk about self-service, meaning that all of data is just there somewhere, but everyone needs to create their own efforts, make their own efforts to actually make it usable for their own purposes. [00:48:34] And the, the main driving force behind all of that is because, um, it, it became more and more of a bottleneck, right? The centralization was good at the beginning to have the expert there. But now that all of the business functions want to use data quicker and for more timely decisions, they cannot always wait for a central data team to do it all for them, right? [00:48:54] Um, that makes them, um, become empowered and they need to do it, be [00:49:00] able to do it on their own, and balancing basically the scalability of everyone can using it and being able to use it quickly and for the right things. Versus then having guardrails on there to not do anything illegal or to do it in a way that's consistent. [00:49:15] Right? That is the key here. Um, and it's not only the data world that is like this, right? In the software world, you have microservices that need to be governed in a certain way as well. Um, and even in AI now, right? The same, uh, paradigm will come again. We probably have a central AI innovation team right now in many places. [00:49:34] Then all of a sudden, all the business functions have AI experts and want to do their own thing because it's quicker that way. So we kind of have to go through the same evolution again, maybe a little bit faster since we learned from data now how should work. Uh, but this is kind of the direction where it takes, um, yeah. [00:49:49] And that actually, and uh, in return means that being a data expert alone without the domain association is not good enough anymore. [00:49:58] Right. You need to be a data expert. [00:50:00] And a business expert as well, to be able to actually support any business function in the right way. Because chances are you're not gonna sit in a central data team anymore in the future. [00:50:08] You're gonna sit in a domain or a business function, and you need to be able to keep up with them actually. [00:50:14] Phil: Very cool. I, I feel like the, the role of the marketing operations person or the person like deploying some of this AI tech with all the changes in data management, it's so important to collaborate with the data governance [00:50:27] team. Um, that. Oftentimes, like we'll sit at different parts of, of the org, but I, I love that like you're combining the two functions together because something we chatted a lot about on the show around like AI agents and orchestration is like, what is gonna be the future layer of AI orchestration, like every single MarTech vendor is. Recalibrating their value proposition right now to say we have an AI agent [00:50:53] turn on our AI agent, use our AI agent, and in a couple of years we're seeing it already. Every single tool in your stack is gonna have an [00:51:00] AI agent, and so the marketing operations person is essentially gonna become like an AI referee and [00:51:05] decide. We're gonna be turning it on for this tool, not for that tool. Our IPAs is gonna be our central orchestration layer, or maybe it's gonna be something else. How can like the data governance team help marketing ops folks figuring out this mess? Especially in enterprises when like all the marketers are gonna be like, we want to turn on the agent for this tool and for that tool like. Who's gonna be owning, like customer engagement, AI agents, when the sales tool's gonna have it, the customer management tool's gonna have it, the marketing automation, blah. Like what? Talk to us about your advice for [00:51:39] Tiankai: Yeah, I mean, um, it's interesting because you ask about data governance, and I think the, the short answer to it would be, of course, that, um, any gov governance team right, would ensure that you're minimizing the risks, right? [00:51:51] So whatever you're launching, you have the right guard rails on it. If it's dealing with personal information, make sure it's not linkable or there cannot be any bad things [00:52:00] happening to the data, right? [00:52:01] That they're dealing with. But it's actually much more a service design problem or like a user experience problem even, right? Because if you think about a user coming and they has to face 10 different AI agents, all of a sudden, which should be maybe one seamless interaction with you as a customer, then you are, again, back to what I said before, with 10 campaigns in one week, you are overwhelmed, right? [00:52:23] So it should be actually much more driven by is that the right thing for our business and is that actually helping our customers to be loyal? Buyers of our product or services, then thinking how can we launch more AI agents, right? Um, if we rooted back to what is helping us to be successful in a business and not just what is exciting to do and what is quick to launch, I. [00:52:45] Then I think the whole mindset can shift more towards end-to-end thinking, value thinking, design thinking, all the things to actually make it better. And, uh, that does not, not, doesn't necessarily mean it's a governance problem, right? It's actually more of a [00:53:00] how do we even manage overall our business processes and value chain with AI in the future, [00:53:05] you [00:53:05] Phil: Yeah, so such a great point. Like thinking about the. Customer experience first versus asking yourself, do we need to turn on the AI for this agent or that agent? Like start with the customer first, work backwards and, and [00:53:17] that'll help you decide whether to turn it on in that tool or that tool. And yeah. [00:53:22] Great point on, on, uh, the P-I-I-P-H-I side of things that, that data governance and coming in and help marketers with. It's been a super fun conversation [00:53:30] 9. How to Use Authentic Communication to Build Influence in Marketing Ops --- [00:53:30] Phil: . [00:53:30] Uh, I've got one, uh, one more fun question for you and then, uh, we'll ask our, our last question. Um, you kind of talk about the one person data team, [00:53:40] uh, a few times in your book. [00:53:42] Um, I'm, I don't know if you've worn that high, your yourself in, in your career, but, uh, there's, there's a ton of overlap with the one person marketing [00:53:50] operation, uh, team. And many of the folks listening have been that sole marketing technologist. You're a big proponent of like balancing technical work with [00:54:00] self-promotion [00:54:00] and communication, and you recommend at least 30% of your time should be spent on communication and building a relationship with your counterparts. What are some of the best tips and strategies that you've come across to make the most of that 30% comms and relationship? Is it like quick coffees with random people you dedicate like your Mondays to do it? Like what does that look like in terms of practical tips? [00:54:24] Tiankai: Yeah. Um, I think there's many ways of doing that, but the most important part I realized that is, uh, to pay attention that you're still authentic, right? And by authentic, I mean, um, be out of your comfort zone, but do not try to be a different person. Just try to be out of your comfort zone. What I mean by that is, for example, if you're an introvert and you don't, you hate being a public speaker for a lot of people at once. [00:54:49] Don't put as a target for yourself to just send newsletters to thousands of people, or go to team meetings and town halls all the time to present there. Then do it the other way around and do one [00:55:00] one-on-ones, and try to network with more specific individuals that have certain decision making power or influence in our organization to then make it work. [00:55:08] So whatever you prefer as your communication style. Identify that herniation style and then double down on it because that makes you at least feel like it's not a chore, but it's something you enjoy doing. Right? And then it's about how you communicate. And that's a lot more about mindset, because [00:55:24] if. Uh, for you to be able to communicate naturally the successes you have in your achievements, you need to acknowledge them as achievements and successes too. [00:55:33] Right. And oftentimes, especially in operational jobs, it feels like a lot of reactive firefighting, so you cannot actually describe what you're doing every day. It just feel like, oh, I'm just doing it with a bunch of different things and I don't know what the hell I'm doing every day. Right. But that is not something you can do in your networking, right? [00:55:48] You cannot just go to a random person and say, I don't know what the hell I'm doing every day. Right. So trying to. Fall it down again or see the patterns, see what you're actually doing. Maybe the firefighting itself [00:56:00] has a pattern or a a above the firefighting. You identified a way how to quickly solve problems. [00:56:06] Even that is an achievement, right? So see the achievements, talk about it in your own authentic way. That is the way how you can actually make meaningful connections with people. You. [00:56:17] Phil: Such a great answer. I, yeah, I love so many things about that answer. I feel like sometimes you, you look at your day like, oh, what did I do today? What did I accomplish? And it's like, well, I fought a bunch of fires. I helped a bunch of people do something, but I didn't move the needle on any of the projects that I wanted to do. I closed three tickets and I made a couple people happy. But yeah, I, I closed fires [00:56:40] 10. Happiness --- [00:56:40] Phil: . [00:56:40] Uh, this has been super fun. Uh, really appreciate your time. Uh, we'll link out to your book, um, but I got one last question for you. Uh, you're a data leader, obviously AI strategy director, a team leader. You're an accomplished author and speaker. [00:56:53] You're also a board member, um, but you're also a musician and a rapper and avid sci-fi book [00:57:00] enthusiast and a silly dad of two as well. One question we ask everyone on the show is, how do you remain happy and successful in your career, and how do you find balance between all the stuff you're working on while staying happy? [00:57:12] Tiankai: Yeah. Um, I think, uh, there's two sides to it. One is, um, I try to be less negative and that means, um, I am conscious about what I have under my span of control and what I don't. Right. If they bad things are happening to me and I cannot control it. Then I shouldn't be bothered by, right? I cannot control it or change it anyway. [00:57:30] But if I can change it, I do everything in my power to make it better, and then at least I try it no matter what the outcome is, right? And so turning it into solution, uh, basically mode rather than problem mode is one key how I stay, uh, or how key negativity out of my life and the positivity, I think. Is actually I feel like, um, a habit that I created for myself. [00:57:53] And those are small things, like every day in the morning, look at my calendar and see what I'm looking forward to. Right. Today [00:58:00] in this case, I was recording this podcast episode with you, right? So I had the whole hours before this conversation to be happy because I was looking forward to something. [00:58:08] Right? [00:58:09] And on the biggest scale, you look forward to vacations, for example. You look forward to your kids having milestones in their lives potentially, or. Everything else in your life, right? But having something to look forward to is, I think, a big driver to be positive, because then everything else matters less because you're focusing on something that is a little bit further away. [00:58:28] So no matter what happens, at least you have this, right, the good thing that you're looking forward to. [00:58:33] So that's how I stay fun. [00:58:35] Phil: Love it. It, it sometimes, uh, is the thing that you're looking forward to, uh, opening back the, the sci-fi time travel [00:58:42] book that, that [00:58:43] Tiankai: absolutely. Yeah. Yeah. Right on. [00:58:47] Phil: any, uh, any, any cool books that you're reading right now or, or, uh, sci-fi books you wanna recommend to folks? I. [00:58:53] Tiankai: Uh, not, I mean, uh, not right now, but I mean, we talked about here now and then, right? That is the book that we [00:59:00] recently just, uh, chatted about. Um, we mentioned also Dark as a Netflix series. It's really good about time travel, I would say. Um, my. What I actually want to do is, um, I, I'm a big back to the Future fan right of, of the movie series itself. [00:59:14] And I know it launched on Broadway as a musical, but I haven't actually watched that [00:59:18] musical yet. And I really want to see, because I hear their original songs in there that were obviously not part of the movie, but I haven't really heard them or seen them live yet. So I kind of wanna explore that and experience that because as you mentioned, I'm a musician too, right. [00:59:33] And if I could bring time travel and music together, that sounds amazing to me. [00:59:39] Phil: Is there a future book in store for us where you're combining data and time travel? [00:59:45] Tiankai: Oh, I would love that. I mean, the easy answer is something like a backup solution sounds already a cover isn't like what MacBooks do [00:59:54] to back up the data already. [00:59:58] Phil: yeah. I love [00:59:58] that. [00:59:59] Tiankai: a good point. [01:00:00] Yeah. Let me think about that. That's a great idea. [01:00:02] Phil: Yeah. Um. [01:00:03] Tiankai: maybe even something with AI agents in there, right? [01:00:06] So an agent, an AI agent that travels back in time and [01:00:10] Phil: Definitely, I've actually joked about this, uh, on, on other podcasts that like one of my dreams is to write a sci-fi novel about time travel, and the main protagonist is a marketing operations person. I. And they go back in time and like in the past, like there is no MarTech, like [01:00:26] they're working like admin ages and then they go to this year and they're discovering automation for the first time. [01:00:32] And maybe in the future there's a couple tools that like still hang around and the whole discussion about like what is gonna be MarTech in 10, 15 years and. I don't know. May, maybe, maybe there's something there, but plug, plug your book. Uh, junky. 'cause I read it. I loved it. [01:00:47] As someone that interacts a lot with data professionals, I got a ton of value from it. [01:00:52] Learned a lot from like, the collaboration side of things. Obviously like the, the human part of working with data teams. Uh, but yeah, plug [01:01:00] the book. Why should folks read it? [01:01:01] Tiankai: Yeah, absolutely. So, uh, my book is called Humanizing Data Strategy. Um, and basically the, the premise or, or the main reason why I wrote. It's because I realized most of data issues in organizations are rooted in human issues, right? A lack of communication, a lack of collaboration, a lack of competence. [01:01:19] Right. And this is what I'm trying to counteract with by introducing a framework of the five Cs, which are competence, collaboration, communication, creativity, and conscience. And to each of them, there's a list of, or a bunch of different experiences that I share or different practical advice nuggets I call them, right, where basically people can maybe see if that works for them and can apply it better. [01:01:42] The, the whole idea is just to have a more people centric data management, a data strategy in place, and not do technology first, but actually people first. Because at the end it's the people that are using the data and that are generating value with it, which that should really be the key. Plus the, uh, the cover looks [01:02:00] really awesome. [01:02:01] I would just say that is the main reason why I should buy it. [01:02:03] It's very colorful and cool. So that's the main reason you should buy it. [01:02:07] Phil: The cover is awesome. Definitely stands out. Uh, I, I saw your recent post, uh, where you probably used, uh, Chad g pt, [01:02:15] uh, to do like the, [01:02:16] Tiankai: The [01:02:16] Phil: version. They're all in there. Yeah. But yeah, I love the book 'cause it is like, I. Very much in tune with the mission of our podcast, the Humans Behind MarTech, the [01:02:28] Humans Behind Data Teams. [01:02:29] And so, uh, thank you for writing it. Thank you for, uh, being with your, with us today. It was super fun. Appreciate your time. [01:02:36] Tiankai: Thank you. [01:02:38] Phil: Boom.