According to the NewVantage Partners’ 2021 Big Data and AI Executive Survey, almost 100% of organizations report investments in AI and data initiatives. Yet, only 24% of respondents claimed that their organization is “data-driven”, a marked decrease from last year’s 38%. In the same survey, the most significant impediment towards becoming data-driven is culture, with 92% of respondents stating that lack of data culture is the biggest impediment towards becoming data-driven.
In this webinar, Sudaman Thoppan Mohanchandralal, Regional Chief Data and Analytics Officer at Allianz Benelux, will deep dive into the ins and outs of building sustainable data cultures. Sudaman will bring his strong track record creating data cultures and will outline how and why data culture is a variable in business strategy execution, reflect on the meaning of data culture and how it’s an extension of organizational habits, how to demonstrate the tangle business impact of data culture, and execution plan for building data culture utilizing the D3M framework from Allianz Benelux.
Understand how and why data culture is a variable in the business strategy execution equation
A reflection on the meaning of data culture for organizations trying to attain it
A discussion on how data culture is an extension of habits, and the importance of the science of habit
The ability to show tangible business impact owing to the attribution of data culture
An execution pattern for building data culture: D3M framework from Allianz Benelux
I divided this talk into two parts, one, I would like to call it a theory of building data cultures, the second a case study on how we built data culture at Allianz. Even though it is a theory, it's something that's going to be a lot more interesting than the case study itself. Because the theory of culture could be misconstrued as something that is a lot more boring than it leads to believe.
Why build a data culture?
So getting into the theory, part of it. Let's imagine what is the first thing that keeps happening. That's change, right? When anything changes, in business, what do we do? We respond. There is another name to this business response, and that's called strategy, right? Anytime you say strategy, what this kind of does is basically nothing, but you are trying to express that you are responding to a change. Now, what would this totally entail? It’s essentially looking for the opportunity to generate value, and that comes in the form of the question “where's the money?”. But you need something to get that money, and that's the people they are, the people who would help you to get that money. Now, you have these people, you know, where the money is, but it is exactly what competitors will be able to see as well, they're also looking for the money, and they also have the people.
One thing that you also need it’s what we call an opportunity, right? So there was a change and then you said, I have a strategy which you call a business response, and then you've said: “Ok, fine, in my response I'm gonna look for the money, I have the people to get the money.” Now it's very important that you also know, where's the opportunity so that you get the money that you very well deserve.
Now, when you have all of this in place, one of the things that you really want to do is to ensure that it is sustainable as well and it is continuously bringing you the money, and that's exactly why you will have to design culture. The particular opportunity presented itself because of the change, and if you want to capitalize on that, you need to ensure that the culture that you have in your organization before fits into the culture that is required for this new opportunity as well. Hence, you will have to think about doing a bit of culture design.
If you do such a thing, then you will realize that this particular opportunity, that you want to capitalize around, basically everyone will be able to contribute to it, and then you could maximize the returns out of it. Now, when talking about culture more precisely, we’re talking about data culture. Of course, we crossed the days of megabytes and gigabytes, now we are basically talking in Yottabytes, right? So there's so much data. All of the companies today are basically saying, how can I capitalize on this opportunity that this data creates, which means, how am I going to not only capitalize but sustain it? This entails asking, how am I going to design my culture? And that's exactly what we're going to see today. Very much what we call building this data culture.
What does it mean to build a data culture?
What does it mean to build a data culture, right? Let’s break it down into components:
It's not just a buzzword. For everything and anything we say that we're going to build a culture, and then we are talking about, “that's basically a buzzword”. No, it is not just a buzzword.
When you create this culture, it leads to creating data-driven decisions, right? You've always been making decisions before, but now, you're going to put some spotlight now that the data is available, and we're going to capitalize on the opportunity that the data creates. How can we ensure that this decision is data-driven so that you take the advantage of the opportunity that data provides thanks to the change? Hence, data-driven decision-making is essential.
Remember, any kind of a change to process it into something that, which is concrete, and constant, and then sustainable, requires time. It requires a certain level of discipline, and then, it requires a certain bit of individual change as well.
That is exactly what we call our habits. What is culture? Culture is nothing but an aggregated way of looking at habits that a particular organization goes through. This requires a habit change, building habits if that did not exist before.
You're going to do the decision-making through the data that we have, and we understand that it's going to take time, that it's going to be disciplined, and we will have to build some habits if those habits were not there in our organization before.
Build data culture starts with the end-user
Now that we know that we have to build something, anything that you build, irrespective of what it is, you will also always have to have the person from whom you're building for in mind. This is what I call the foundational aspect, which we should take into account when we are interested in the job of building data culture. Do we know our business experts? That's the most important thing that you should think of before you start building a data culture.
Business experts are passionate and follow intuition and instincts. I don't know if you've seen this movie, Field of Dreams, a pretty old movie. If you're not seen it, you should see it. In this movie, the lead character dreams and actually builds a baseball field in this agricultural land, bringing his dream baseball players and legends like for me, legend players to play, hoping that people from different parts of the world will come and watch them play. That's as passionate as it sounds, but there is absolutely no data or evidence that people will come, then pay money to watch the game, etc... But still, he goes ahead and builds it. The whole movie stands for something called hope. Hope basically has no data. If you could think deeply about it, hope doesn't have data in it, but it is good intentions, well-wishing. The reason why I call Business Experts passionate is that they have the energy to follow their intuitions again, and again, and hence they are super passionate about what they're doing and you will realize that they always follow intuition and instincts.
Business experts are also artists in a sense and they consider what they do as art. Now, I don't know if you've watched this movie called Mona Lisa Smile, and indeed, in this scene, the lead character projects certain paintings and basically asks students to rate these paintings. They basically rate it according to standards, technique, composition, and color, etc... Suddenly, she projects her painting, which she gifted her mother long ago. Everyone also starts rating it as one of the best paintings. She states: “Hey, that's the painting that I gave as a gift to my mother”, and that's when it occurs to them. It's not about the painting itself, but it's about the way it makes you feel. So, business experts have a lot of passion, they sometimes struggle to explain their decision-making that is backed by intuition and passion. If you ask them a question about, “Hey, how did you make that decision?” They might say, “Hey, I cannot explain, it's my intuition, it's my expertise”, and then that's true. They're not trying to hide anything, but that's true. They cannot call it a standard. They cannot build a standard out of it, They can not think about a methodology or a framework simply because they consider it as an art.
If you think about business experts, they also believe a lot in their experience. Right? Experience is very non-linear in nature, and if you try to explain their non-linearity you won’t be able to succeed. The simple reason being you cannot score the decision made by a business, without having to look at the results from the decision. Imagine, you take away the results part, and you only look at the decisions taken, you will never be able to score it, just like a poem.
So, how do you build a data culture for someone who believes, who is passionate, believes completely in intuition, and follows their experience? How do you bring in data thinking? So what you need is to inspire them to adopt this culture. So how do you inspire them?
First of all, figure out a way to explain to them that their work should be changed. In fact, before doing that, why should they change? Then, you should explain the nature of the change to them, you need to show the opportunity in change. That's exactly what you're going to see. So, why change?
The nature of change
Why should we change? Look at the financial industry, just as an example. All industries are undergoing change. In the financial services industry, the needs have changed. We have moved away from mass-market products to personalized services. These days, we don't even buy a product unless there's a personalized service attached to it. This need has changed. Our relationships with customers have evolved from and have been accentuated by data to stay hyper-relevant to them. So there is a considerable amount of potential which is untapped to customize services for customers with data. There's also a considerable amount of resources, that, which we know are available, and, indeed, that is something that, which has to be tapped as well.
So, this gives very much a good segue, for something that I call a historical change, as well. So, long before, and even now, there is this concept of working directly with the customer, which we call B2C. From there we shifted to another way of working where we have businesses, which support businesses that directly serve the needs of the customer and that is B2B2C.
What we are doing today? It’s what I call BPLEX-2-C. So, in the case of Allianz, we have to work with other businesses with the customer in the center. The customer really doesn't care, as long as the service is enjoyed by the customer. So, basically, an ecosystem is supporting the customer, as long as the customer needs are met. This BPLEX-2-C is one of the most important things that was also driving change. Remember the initial discussion: we are changing the business response. Where's the money? Where do the people ready, see opportunity? And because of all this change, we need to change the culture as well because our culture, before the change, did not fit with this kind of requirement.
We also said that we now understand why the change, and we should understand the nature of the team, right? I'm from the insurance industry. So basically I put this in front of you. If you're not from the insurance industry, it's a little bit difficult to understand. I'll help you with understanding this graph which looks at driving score and risk exposure.
What I intend to say is that with the availability of data, you're able to have a driving score for all of your drivers who have a policy with few motor policies with you. When you are able to basically overlay the driving score levels and look at the risk exposure, you are able to see that on the left-hand side, you have a relatively still driving score of 14.6. While on the right extreme, you will also see the need is 0.2.
Now, this gives you a great opportunity to either overprice or underprice, based on the overlay of this driving score. Basically what we're seeing is that the risk premium model is now completely different, with enhanced data that is outside of all traditional factors. So we bring in modern practice to basically look at it. This gives you an opportunity to compete very differently in the market. Hence, the nature of change is totally changing the dynamics of the market, and the way you compete in it as well.
The opportunity in change
What is the opportunity here? The opportunity is not just one chart. It is going to be long-term, as well. As data availability increases, in the long term, the sophistication of the solutions is also going to increase. I give you an example of a medical world where we see a plethora of use-cases emerging across time horizons. This is an extraordinary value that we create for the ecosystem. Indeed, we know why we have to change. We know that it's definitely the nature of change that is fueling the dynamism of the market now, there's even an opportunity there, which is long term. So it all adds up to why the culture has to be designed differently.
Data culture is the basis of capturing opportunity
Having said all of these things, we know that the data culture that you're trying to build is the basis of capturing any opportunity, right? What do we mean when we keep saying change? We need to facilitate a cultural change. So the behavior working with our colleagues, in our organization, in our company has to change to adapt to this environment. So what are the components of this change?
Breaking down silos: We come from a timeframe where we created data and data was not an asset. So, we didn't really care to organize it the way it should be organized. So it is scattered all around. So, we should be aware that it was basically scattered all around.
Specialization vs Generalization: We also come from a timeframe where we give importance to specialization. Now we are entering into a timeframe where specialization is all right, but generalization is required as well.
Stakeholder engagement and new ways of working: Several stakeholders need to be engaged in the data value chain, and this will enable us to come up with something, what we call radically different methods of ways of working, and new working models, start appearing.
Using data as an asset: Finally, it’s the recognition that data as a stand-alone asset will be futile without systematically extracting value from it.
So data culture it’s not just an option, it becomes business-critical. Because you are in a landscape where you are competing with entirely different kinds of models and ways of work that are radically different from what you have been looking at before. So I cannot preach more than that. Without wide-scale adoption of data analytics, it will definitely not pay off. Think of it as getting an Apple Watch, where it will just show you different health measures and metrics, but if you don't want to take that data and make decisions from it, it has no value.
The two hurdles you need to overcome
What are the hurdles of accepting this data culture? In my opinion, there are two hurdles. We basically come from a world where we have the guts and the instincts, that which was driving it, and we now want to go to a world where we will be able to use data.
The first one is prioritization. How do you prioritize your move towards data culture? Of course, the results are going to come, but it will take some time, and you need patience, especially as conflicting priorities come up.
The second is sustained investment: The second is how do you sustain the investments that you're putting in the culture change? Remember, culture change necessarily means you're going to change the habits of not one person, but the whole company, which means the culture is an aggregated way of looking at habits of each and every employee within the company. So the investment needs to be sustained.
When to invest in data culture?
When should we invest in this data culture? I think there is no real answer to it. You should invest in it at some point in time throughout your data journey. I don’t think you should begin investing in data culture at the beginning of your data investments, nor should you do it after your data investments. All I can say is you need to look into your company, the culture of your company, and then you need to find the gap. Then you need to find a way to say that, Ok, I'm going to start investing data exactly at this point in time.
For example, if you see that your company is going to take time to even adjust to the fact that they need to make data-driven decisions. Then you need to start with data collection. You need to find a way to convince them that data is usable, and create trust.
So, first of all, you need to define why we exist as a business for our customers, and why customers choose to do business with us. Then, you need to see, what kind of data do you have that aligns with this definition of value that helps me to serve the interests of a customer better? Then you need to intervene in different processes and habits and inject data habits there. This is a kind of a rule of thumb for you to know when you need to invest your data culture, and this, of course, is not a single step, that once you do 1, 2, and 3, everything is done. No, you need to continuously iterate and test. You need to choose exactly when, how, and where.
The successful ingredients of a data culture program
So what are the successful ingredients of a data culture program? How do you ensure it succeeds?
Know your impact: You should design a culture program that has the goal of generating impact and value for customers. So you need to know how you’re creating it now, and what exactly are the moves that you need to make in your culture program to scale this value.
Know your default behavior: Which also means that for all this, while you have been doing this, there is a default behavior in our organization, that you can understand what the default behavior is, and this is the default behavior you want to reprogram so that this particular behavior changes into a behavior that puts into the future.
Prioritize energy: That also gives you an idea of how to prioritize energy, and where to prioritize it. This will enable you to develop some project lifecycle, kind of a framework where you don't have to iterate all the while, but you need to have a controlled way to iterate so that you don't end up in an analysis paralysis. In the end, you are able to track progress and recalibrate and monitor your return of investments.
Summarizing the theory behind building data cultures
So starting from the impact that you create on your customers, all the way until recalibrating to be hyper-relevant, this would mean that your data culture program is going in a very good direction. To summarize the first part, at least. I think what is important to know is that you should allow the organization to:
Discover, develop, and deepen: Discover the data world and get their feet well. Also help them to develop their skills, that's what we call, data literacy or data awareness. Then you need to also help deepen their skills of decision making.
Nature vs Nurture: What is meant by nature here is original habits and attitudes, remember that business experts are passionate, guts-driven, and they also follow their experience. There is also another part, which is nurture, and that's what these three-Ds you see above. that you saw: Discover, Develop, and Deepen. You need nature and nurture and need to balance these two things and know exactly when you're doing.
Redefining Culture: Remember that you are redefining culture, which means habits are changing. To be aware that you need to sustain the investment, you need to prioritize those hurdles.
Remember that, in the end, don't ever forget that whatever change can happen, but the people's agenda has never been more important.
So, this is the theory of building data culture. This theory we developed, as we built it on a larger scale, at Allianz Benelux. With this, I will go into the case study itself.
Case study: Building a data culture at Allianz Benelux
So how did we do this in Allianz Benelux? So, we started, just like everyone else, to build data solutions in 2018. Then we started to look at habits and routines, and we said that, hey, you know what? We need to change the habit of the business for whom we are building the data solutions. The results of which are data solutions that can be used by the business to generate income. So we discovered there was a link that was missing between the solutions we built and their adoption. So we realized we need to do something about it. We realized that habits that define the culture and data-related habits, will transform our culture into a data culture and that is clear, but how do we get there?
We started to read a lot. Of course, we found this book, The Power of Habit, by Charles Duhigg. His idea was very interesting when we adopted it in fact, and customized it to our context, and we started to use it. This one very clearly says that, there is a routine, and that's what we call a habit, and this routine is a habit because there is a reward. You will notice that this routine has a queue that triggers it.
This is really interesting simply because it kind of fits into generally everything that we do. If you are a smoker you know that there is a trigger for you, which makes you think that you have to smoke. So you go out, you smoke, the reward is that you get out of it, of course, the feeling that emerges after you smoke. But indeed this continues, you know, and if you want to stop smoking, there is something that you have to do and that's exactly what you're going to see.
Right, so, if you want to stop smoking, of course, that's the routine that you want to stop. You cannot touch the reward. You cannot touch the trigger, but you can change the routine. If you do change the routine, then you will have a reward. Of course, you need to go for the routine. So, basically, what we are saying is that, don't change the reward, don't touch it at all. Because reward is the reason why you basically do anything. Don't change the trigger, because whatever routine you are going to change, has to be triggered. The golden rule of habit change is to keep the queue, provide the same reward, but insert a new routine, right?
Enabling elements of a data culture
We will come to that part and I will explain to you the power of habit in action. But before, what are the enabling elements of this data culture?
Common Language: We decided that we will have to have a common language. And that means marrying data literacy with subject matter expertise.
Data Governance: Then we needed a very strong data governance setup. Obviously, giving the ownership of data to the business.
Continuous Measuring: Indeed, we then had to basically figure out, how are we going to continuously measure that data literacy and data governance.
That's why we set up a kind of an approach to design an experiment and also measure data from there. So we started to do two things, surveys, and interviews. We needed to know what to measure. The D3M FRAMEWORK
We said that if you know that you have to do a survey or to basically go for an interview. We said that we would have seven dimensions or seven operationalization areas for 14 subjects. These seven dimensions are kind of universal in nature. So, you can choose seven operationalization areas that fit the best for your company, and then all the 14 subjects are applied to them. What are these seven dimensions?
On the seven dimensions, we wanted to measure the impact of data culture. So what are we trying to do? We said that we want to change the culture. We said that we're going to change the culture in such a way that we will be able to use our data products and solutions for the business, but how is this culture adopting these data products? Basically, what is the impact? We then chose the seven operationalization areas such as which are:
Tools that are available, which makes the data findable and understandable?
Data literacy: How do we increase that? How do we increase the data literacy in the company?
Embeddedness: What do we think about data solutions embedded in the company?
Customer Centricity: Are we thinking of customers every time we're thinking about using data?
Innovation and experimentation
The 14 subjects that we wanted to measure were all of the following:
Data Availability: Is the data available?
Data Quality: What’s the quality of the data and is it accurate?
Vision & Strategy: What’s our vision and strategy?
Role: Are there roles specific for certain data workflows?
Modeling: How can we model data?
Communication and Support set in place to help in the culture program
Relevance of the culture program
Readiness: Do we have the mindset to work with data?
Ability: Our skills and behaviors
Training & Support: Do we have learning resources to provide experts?
We call the combination of this experiment set up the D3M Framework. And we started essentially measuring dimensional, operationalization, and granular subject scores across the organization to measure our culture program progress.
Even the data culture program is now data-driven. This was very important for us since we want to measure the progress of the culture program. It gave us the tools which are required to measure the impact that the data culture was creating.
Making data culture programs personalized
So now that the culture program is data-driven, how do we make it more personalized? We basically used surveys in a way where if you take a survey, it will clearly tell you what you need to do in order to improve your data literacy and subjects automatically.
It will tell you which training you need to take. This was very useful for all the company employees since they are able to figure out, where should I get help? How do I get this help?
Allianz in partnership with DataCamp
I kept talking about data literacy and it’s supercritical. Here at Allianz, at a global level, the Group Data Analytics basically has a Data Analytics Academy. We also partner with DataCamp since April 1st, 2021. While very recent, we already have 1000 learners using courses and custom tracks that we have created.
We have some custom projects as well. At the group level, even at Allianz Benelux, we have something called Accelerate Academy where we do Bronze, Silver, Gold Data Literacy Training. All of these things kind of incentivizes our employees towards the world of data.
Indeed, now that we have the framework, we have a score that comes out of it. We know that the survey results become the input. The scores, which basically tell us what the data maturity is, become the output So we hope that this feedback loop positively reinforces itself, slowly pushing the organization's habit loop into something that's just more data-driven. We end up measuring outputs and inputs via a D3M dashboard that provides this impact.
Once you have all of these things, remember that you need to industrialize and operationalize the surveys. You need to provide supplemental data. You need to have all the personalized learning right next to you. Then, with this data-driven decision-making dashboard, you are able to also connect all of the dots to the business impact that you're creating. This is a data-driven way to measure culture change and we wanted it this way. We wanted to make it as personalized as possible. And we also wanted to take the medicine we are pushing to others. We’re not using gut instincts, but we’re using data to measure our culture program.
An example of that, two years back, we did a survey, exactly with the Belgian sales and distribution team. We wanted to look at the decision-making habits they have. The way we went is that we listed all the decisions that they make. We ordered them by business impact. And then, we started to see whether data is really involved in the decision-making? Indeed, we realized that in certain, high business impact decisions, for example, broker steering, data was not as heavily used as it should be.
It deserves to be used. In this situation, the routine was steering the broker. The reward was, of course, the business impact. The trigger was, of course, the time frame and the seasonality So we retained the queue. We retained the reward, and, in fact, even improved the reward, by making the broker steering data-driven. Remember what you do, you should think of the outcome, not the output. You should stay hyper-relevant, and then any program will succeed.