Humanizing data strategy: The foundations of data strategy

Transcript

[00:00:01] Markus Buhmann: Welcome to Pragmatism in Practice, a podcast from Thoughtworks where we share stories of practical approaches to becoming a modern digital business. I'm Markus Buhmann, your guest host for the first episode of our Humanizing Data Strategy podcast takeover, along with my colleague and the author of Humanizing Data Strategy, Tiankai Feng, our head of data strategy and governance here at Thoughtworks. Hello, Tiankai.

[00:00:22] Tiankai: Hello, Markus. Nice to be here.

[00:00:24] Markus: Thanks for being here. Today, we're joined by a special guest, Asmaa Hechenberger, data engineering lead at WWK Life Insurance, one of the largest life insurers in Germany, based in Munich. Hello, Asmaa. How are you?

[00:00:38] Asmaa: Hi, Markus. I'm fine. Thanks to be here.

[00:00:40] Markus: It's lovely to have you here. In this episode, we are going to be talking about the foundations of a strong data strategy. Asmaa, thank you for being here today. Can you tell us a bit about yourself and your role at your company, please? Tell us a bit about some of your experiences in formulating and implementing data strategies.

[00:01:00] Asmaa: [unintelligible 00:01:00] Markus. I'm Asmaa Hechenberger, and as you notice, as my name is combination between Egyptian name and German as well. Yes, I'm coming from Egypt and I've live in Germany since 2006. Yes, I come from the tourism management. I already studied in Egypt and then I decided to travel to Germany. My dream was to learn languages and another reason for my journey and it was football, so I was really, and still, as a football fan. I couldn't wait to see the football World Cup in Germany live. It was actually the reason why I'm in Germany. Now let's start with data.

Since 2006, I started to make more interest to data and I started to study different programming language. I started to work in one company in Munich and it was only for data warehousing for seven and a half year. After that, I changed the fields, another company. It was for e-commerce and working with controlling finance data, CRM marketing. Now recently, this time one year, I'm working in insurance field in WWK in Munich. It's really a new direction for me and it's really interesting and exciting as well. My focus is, you can say, all around data, business intelligence, AI, I was building AI solutions and data management and also data and integration platform. That's all about me.

[00:02:47] Markus: Amazing. Thank you. Turning to Tiankai. The reason we're taking over this podcast is Tiankai has released an excellent book, Humanizing Data Strategy, available at good booksellers everywhere. Tiankai, can you tell us a bit about humanizing data strategy and the five Cs model that sits at the heart of it? I think it's really, really important conversation to be having. We all know about all of the frameworks and everything, but this takes on a slightly different, I think, more important view.

[00:03:20] Tiankai: Absolutely. Happy to talk about it and very excited that we are able to do this takeover of the Thoughtworks podcast. Maybe first to the motivation of writing the book. I actually did write the book about the human side of data because I felt like, comparably to the technology side of data, there's very little practical literature about how to deal with the human side of data. Although everyone, in theory, agrees on the human side of data being the most critical and the most difficult one, it's not really reflected in practical advice that we see anywhere.

Although there are few authors that actually doing a great job writing about it, in contrast to all the rest that exists in literature, it's very little. I decided to give it my attempt to write a bit more practical advice about the human side of data. Basically, for me, the human side of data and data strategies boils down to the five Cs, which are competence, collaboration, communication, creativity, and conscience. All of this have different elements along the journey of data professionals and how you plan out different data initiatives as well but also, ultimately, they're supposed to lead to increasing value of data by motivating people to do the right things with data in the best way possible.

I actually had to chuckle about the passion of Asmaa about football because I use football often as an analogy for data because it has all of it in there. You have the teamwork aspect of it that the whole team has to play together to shoot a goal. You have the accountability aspect of it all because everyone has their specific role, either you're in defense or in offense. You also have the referee, which is a little bit like the governance aspect of it all, to make sure everyone sticks to the rules. I feel like it's a great sports analogy that we directly threw into the podcast for this conversation.

[00:05:06] Markus: Thank you, Tiankai. I think, today, we were going to focus mostly on the collaboration piece of those five Cs and how that's really foundational to a good data strategy. As much as some of our former colleagues and some of our customers would like to think, they can't do it all themselves and the data team. I guess, Asmaa, that's something that you'd be acutely aware of with your initial work experience in the hospitality industry. You can't run a kitchen or a restaurant on your own. It's just not possible.

Do you want to tell us about how some of your experiences in your previous professional life have played into your data career? That's a career journey that I haven't really, really seen. Yes, please.

[00:05:57] Asmaa: To be honest, collaboration is really important. It doesn't matter in which field in our life. When we're talking about data, so data teams, this is one team, but actually, there's many team members that is coming from different areas. For example, data scientists, data engineering, from the business side, for a stakeholder, and so on. Because of that, we need really a good cooperation. We have to work together. Otherwise, we can't achieve our goal.

We can say, "Yes, what's happening in football or what's happening in tourism is the same." We have to work together as a team. Otherwise, we will not have the good service at the end as an output. Collaboration, as I told, this is really important. It's not only at work with the data, in the data field, but also in my private life. For example, there's a lot of communities that we need also to collect or to gather each other. It's also a kind of collaboration.

It's not only, as I mentioned, at work. If you ask anyone who's working with data, we know how this is complex. This is one of the data projects and how many tasks can be behind one task, only one data product? If we work all together, we have an established and solid collaboration and valid collaboration, so we'll really achieve our outputs and our goals. The keyword is work together. We have to have a shared or one goal for the whole team members, and also for the whole other departments and everyone's involved in one data product or data project or something as [unintelligible 00:07:43]. That's just my opinion about collaboration because I can see it's really important.

[00:07:50] Markus: We 100% agree with you there. I think the thing is that it is very much a team sport if we want to keep going with our analogy. I think there's something that I want to pick on, Asmaa, that you said there that was really, really interesting to me, the shared goal that you need to bring people together across various teams within the business, within your data team, within the technical function, to be able to implement an outcome from a data strategy.

Can you speak to a time around how big, from your perspective, do you feel, is it a bigger ambitious goal that fires up or do you work best with a smaller goal along the way, or do you work from a-- one of the things that we talk about a lot at Thoughtworks is this idea of a thin slice where we go all-- a very narrowly and tightly-defined piece of work as part of a strategy outcome just to prove that you can do the thing, and then to flush out any issues or gotchas that nobody's seen, has anticipated. How do you work, in your role, within your organizations along those sorts of spaces? What goal is energizing and practical for you to deliver?

[00:09:04] Asmaa: If we talk about the practicals, of course, we have to have a clear communication. This is the first point. Also, we have to share our vision and goals. This is also the second part that we have to talk about. We also need to have a rule clarity, to define the rules. For example, each team member-- I'm not talking about only the data engineering team or tech team, but I mean everyone's involved in one data product. We need to define clear rules.

One further point is the trust and respect in this group. It should be we trust each other, the business side trusts the work and the efforts from the data engineering team, how they will manage a topic or issue. It doesn't matter what is it but this trust, we should have it. That's really important.

How can we make it in practice? Is it a regular meeting, like bi-weekly or stand-up meeting, or a regular meeting to discuss and talking about the next steps, challenges, and keep everyone aligned and accountable? This is really important. Also, we can use collaboration tools. There is a lot of tools that I don't need now to mention the names, but there is a lot of, for example, boards, et cetera.

We can use it also as a cooperative tools. What else? Workshop, brainstorming, this is also one of the important point. In my experience, I always used pilot projects. I tried to find, I guess, as a start small, as a really important project, start small, so a use case, how can we solve it? How can we start at first? Just the first step. In my opinion, this is really the important part. Just to have a frame rule, how can we follow it and just how can we manage it together, and it will work.

[00:11:18] Markus: I just want to pick up on something there around the use cases. Now, Tiankai talks in his book quite vividly about following the pain. Do you want to talk about following the pain a bit, Tiankai?

[00:11:27] Tiankai: Absolutely. Following the pain is basically just a principle to say, to show value in your data efforts, you need to fix problems. You need to solve problems. It basically goes with the assumption that any organization has some kind of pain point existing at any given time around data. You should just find those people that are hurting because of the data pain points they have, and you help fix it for them, and you have an immediate value.

I think, with following the pain, it basically means don't just pick any random use case to do, but pick a use case that is actually solving a pain point for someone. Not only does it have a value at the end that comes out of it, you actually gain word of mouth as well. We gain advocacy from it. Because those stakeholders that actually get their problem solved are finally relieved by having someone have solved the problem and they're going to spread the word.

They will say, "Oh, finally, we work with the data team and they actually solved a problem. They are great. Everyone go to them and they're going to fix your data problem." Creating not only it for a value point of view to get buy-in, but also to actually gain that trust that we just talked about is actually really key. The more you have other people talk on your behalf about you positively, the more authentic actually the trust is because it's not just you bragging about how good you are, but other people can vouch for it, how good you are as a data team as well.

[00:12:48] Markus: Asmaa, I see you nodding enthusiastically there and with a broad smile on your face. It seems that this is something that you do when you identify your pilot projects and use cases. Do you want to talk to that at all?

[00:13:01] Asmaa: I confirm what Tiankai said. It's absolutely correct. Yes, it was a smile on my face because I'm a data engineer, and when I think about data engineering, we're just thinking about or we are really excited to start something new as a new project, a new idea, as something new trend, and so on. When we're talking about fixing some issues or bugs, we have to double-think about it. We have to start really with this point to follow the pain as Tiankai mentioned now.

It's really important to see what's the problem, what we have. We need to have a stable data platform or a stable integration platform. It doesn't matter what I'm talking about. When we have a stable system, so we can really make it bigger and bigger and bigger. At first, check our issue and we try to fix it, try to optimize everything, but it shouldn't be 100%, but at least make this balance. Work in something new, but even at the same time, also try to solve our problem. Of course, this makes the business side really happy.

If you already have a problem, it's always talking about it, always frustrated, and then, oh, it's now fixed. It will have a good atmosphere, and really from all sides, for business side and also tech side. It will help us to improve the whole process. The only thing also I would like to mention here I'm using my daily business. For example, if there's someone from the business side comes with the issues or problem, so I said, "Okay, it's fine. Okay, I get it." We write a ticket together and the requirement and try to define everything, and clarify everything.

Then I'm always try to involve this part as the one who is already complain with us in this solving process. In the beginning, it was something like a service, "I have a problem, and you have to fix it," and it's fine. I said, "Okay, it's fine. We accept it, but please come in board and help us we finish it, me test it, me try it. What's your opinion about it? Then 'It's this,' we change everything, we change the rule." Instead of just as a request and answer, it will be we work as a one team. At the end is everyone is really feel good, and it's really-- and it's also cooperation, you can see it like this.

[00:15:49] Markus: It's building that trust that we were talking about before, isn't it? Insurance is a very staid-- well, it's perceived to be quite a conservative sort of industry. It's mostly conservative. They're very structured, very, dare I say, old-fashioned, probably not fair on them. Some insurance listeners might be screaming at me right now. I do apologize. Taking that co-creation approach is quite an exciting and different-- it's becoming more and more prevalent now. How are your colleagues at WWK taking that on board? How are you finding that landing?

Some people really take to that, because you bring them along on the journey with you, they get a sense of ownership, all of those good things. Tell us about how that's landing in your new role.

[00:16:43] Asmaa: It's a little bit difficult, to be honest, because insurance is another product as e-commerce. This is really the difference. I'm just start recently since seven months or eight months, so that means although I'm still new in this direction, but I can notice that, of course, it's different. It's a different product, it's a bit serious, and even the kind of the data and also the amount of data is not big data and so on, what I already have before as well last year, at least, or last couple of years, but it does matter.

We are talking now about human side. Yes, at the end, we are all human. Still, all of us comes in the same table, and it doesn't matter which field is it. We are talking about data, is we're really good in data because we have one data language. It doesn't matter in which field. We discuss it together, we try to find our SharePoint, how can we do it? How can we solve it? To be honest, I can't find really a big difference for my other rules or other fields for e-commerce or other controlling or something like that. It's actually-- as I mentioned now, we still talking about data as we have one language.

[00:18:01] Tiankai: Actually, since we talk about different industries, it's really interesting one that certain industries, they actually originate in being data-driven already. If we think about insurances, not just life insurances in general, the whole business model evolves around risk calculations, and evolves around basically understanding the customers and what are the right products they can buy from an insurance. Since data is then already very much top of mind, and there's the common ground between everyone that the core business evolves around data, in certain cases, it's a little easier to actually break the ice.

You can actually connect it to the core business processes that are already related to data, to then create awareness and interest maybe for other data use cases. I think in other cases, in other industries, it can be really quite hard. If you think about, let's say, like a restaurant, for example, since we brought that example before. The maximum data you have might be how many seats you have, and what reservations you have and all of the financials, but as a waiter, for example, you maybe not care a lot or too much about all the data points that exist in the back end.

It's really different stories, and depending on actually where the, let's say, common employee in an organization then comes from, you have to make different types of efforts to actually onboard them towards speaking the language of data, as you mentioned, Asmaa, to actually be on common ground. Like, what is the right way for us to understand each other so we can actually work together? If we speak completely different languages, even, or just use different words for the same things, there's a lot more work to do, for sure.

[00:19:39] Markus: I think that common language, that ties back to Asmaa's earlier point around a common purpose, and a common set of goals to really unite a team. I want to talk about something that you mentioned earlier in your personal life, Asmaa, around communities and community building. How do you see data, in your view, data playing a role and building those communities beyond just, say, identifying potential members and so on? I'm really interested to hear your views on that.

[00:20:10] Asmaa: Community. What's the purpose behind community? Many of people, they already have a similar interest, challenges, hobby, doesn't matter, but they always have a shared thing between each other. This is the idea behind community. Yes, in IT, also we have this kind of communities for engineering, for example, data engineering, software engineering, and cloud. There's really several types of communities. If we can build a kind of communities for data, it's really great. Why? Because everyone, we are human, we still talking about it, and everyone also have a specific experience.

For example, someone is really good in a specific program, language, Pythons, other one, and scripting, other scripting, and so on. We can really help each other and we don't feel this feeling that we are alone. That's mean we work together as a team, we work as a community, we share the information, and support each other. At the end, it's really a good feeling when you have the feeling that you belong to communities, a group of people, all of them have the same interest.

You can imagine, if you have a community in your organization, how it will be helpful, how it helps. No one will have the same plan that you can't plan me, that you can't do it or do it or you are not good in it, or not because all of us, we can support each other, and this is a really good feeling. Community is really important.

[00:22:00] Markus: Tiankai, do you have any reflections on communities and data?

[00:22:05] Tiankai: Yes, for sure. I think, maybe from the other point of view, and fully agree with the human needs and the human benefits of a community, what Asmaa mentioned. Another side where communities are established already are in tech, where you have communities of practice. That is not really a new concept in tech, where basically they have the shared interest, as you mentioned, of doing something technically, and either they are all data scientists, or they are software engineers, but there are these communities of practice.

In a way, these communities of practice are not only there for the fulfillment of human needs of exchanging about best practices. I noticed also that, especially when you create a data community of practice, it also can become almost like a self-help group of saying, "We can actually complain together about things that are not working." We can give each other some courage to say, "You're not alone in this, we are all facing the same problems. We can get through this." As cheesy as it sounds, curing that loneliness of feeling that you're alone in all of this, and you are facing, as the only person in the world, this problem, the moment you know other people are facing it too, you just feel better.

It's a pattern I see at conferences, too, when data leaders come together, and they all present about similar problems. When they then talk to each other, after drinks, they all are like, "Oh, this feels so good that I can talk to someone about my problems," that this is, I think, also what's the human side of it. In data, we deal with enough problems. If we can feel a little better about them in our day-to-day through communities, then that's pretty good.

[00:23:43] Markus: I've got to say something, Tiankai. I used to work with a lady called Caroline Carruthers, who you might know from [crosstalk].

[00:23:48] Tiankai: Of course, yes.

[00:23:50] Markus: She would allow space in her workshops to have what she called a therapy session, and it was a safe space where she would allow people to get all of these things off their chests. It was really, really useful because people bond in adversity, is what you're alluding to, but it also gives that validation that you're not plowing this furrow by yourself.

The thing is that you can complain and complain, but what comes out of the end of it is that there's a little bit of, "We're all in this together. It's not us just by ourselves." What Caroline was very, very good at was showing a way to move beyond that into something that was more aspirational, helping them discover their common goal, and giving them ways to work towards that. It's really, really interesting that you hear these things over and over and over again.

One thing, Asmaa, given you're new in your role, and when anyone senior goes into a new role, there's a big part of networking, if you want to call it that, but let's call it community building. Showing that you're there to help people, you know what you're doing, you want to build this collective vision of how data can solve problems, meet your regulatory requirements, which is particularly important in the financial services space, build a competitive advantage, et cetera, et cetera. How do you go out and find those individual people that you can collaborate with, going forward, and how is that going in your new role?

[00:25:31] Asmaa: It's not easy to just ask people, "Come and let's build a community." Of course, it comes from the situation of the use case. If we have already a problem or issue, it's coming automatically. There's a couple of people gathered together and try to solve it or to start a discussion between us. Here starts the first communication or the first connect to the others. How I do it? The answer, I don't have a specific plan or poster that I follow to do it, but this comes automatically as just I see which one or which people already have the same interest and would really have this encourage, this power to do it so we can do it.

What I would like to say that I can't go for the other people and ask them to do it. It should be the same interests together. For 80%, this is the case, and we can actually find each other directly. I don't need to make a specific plan to follow it, to be honest. Of course, for something like that, I know as a leader or as a leadership position, you have to act as a role model. That means something like that should start from my side at first. I have to show something for the others, and then they will be able to say, "Okay, let's try it and let's do it." I always make it like that when I have a problem or I would like to build a kind of community in specific area or expertise.

I start as the first one. I said, "Okay, people, I will do it." I will start the first bi-weekly, in two weeks. If there's anyone would like to share or join, just come at this open meeting. I just send the link and I see who is interested and the people is coming, so it will be the start. It's the first member in my group. I don't like to force the people. I do it. I have to start at first, and the others try to do the same.

[00:27:49] Markus: At the risk of doing consultant thing where we repeat what you say back to you, but you want to create an inclusive space for people who are interested in data, talking about their data problems, and solving data problems. Would that be fair to say?

[00:28:04] Asmaa: Yes.

[00:28:05] Tiankai: I think that repeated what I understood very well. That's exactly how I see it, too. I really like this, also what Asmaa said, that you cannot force other people to collaborate with you. Basically, have to give them intrinsic motivation to do it. They have to understand that it's important to do so, so they want to work with you and not like their boss told them to work with you and they do it. That only works once or twice. Then they'll probably never talk to you ever again. I think that approach of finding common ground first before assigning tasks is really important, especially when you're a new joiner and you start building a network.

We shouldn't do the mistakes of just starting to talk about specific tasks in the very first meeting when we meet new people. That is not really helpful. Only after you understand the culture and the way of working together in the organization, you find ways to naturally come to the conclusion that you need to work together and have certain tasks distributed among each other.

[00:29:03] Markus: I think that's a really, really positive way to end our discussion today. Collaboration and inclusion, not just in the data space, but even outside, as Asmaa has quite eloquently pointed out, are so important right now in the world. I think that's something that when we talk about humanizing data strategy, that connection, that inclusion, that collaboration is right at the heart of that message. I want to thank you for joining us on this episode of Pragmatism in Practice. That leaves me to thank Tiankai. Thank you, Tiankai.

[00:29:42] Tiankai: Thanks for having me.

[00:29:44] Markus: Asmaa, thank you so much.

[00:29:46] Asmaa: Thanks, Markus.

[00:29:47] Markus: If you enjoyed this episode or have any questions, please feel free to reach out to us on LinkedIn. I'd urge you to tune into our next episode where we'll dive into the human side of data, the communication and creativity—yes, you can be creative with data and data management—with Sebastian Kaus, the head of data governance at Vattenfall. If you'd like to listen to similar podcasts, please visit us at thoughworks.com podcasts, or if you enjoyed the show, help spread the word by rating us on your preferred podcast platform. Thank you very much.

[music]

[00:30:31] [END OF AUDIO]