 Hey, welcome back everybody. Jeff Frick here with theCUBE. We're in the Palo Alto studios today to talk really about the customer journey. We're excited to have our guest today who flew in all the way from Istanbul, Turkey, which is a very long flight. It's Attila Bayrak. He's the chief analytics officer for ActBank. Welcome. Hi. So first of all, I hope you get some time to catch up on your sleep before you turn around and fly all the way back. Yeah, it's a little bit quick to speak about finance and banking, but it's good to be here. We're glad you made the trip. And so before we jump in, for people that aren't familiar, give us a little bit about ActBank and the history of the bank. Yeah, sure, sure. ActBank is one of the leading private bank in Turkey. And it's almost 70 years old and we have nearly 14,000 employees and with the 850 branches, around 4,000 ATMs and probably half a million merchant point of sales. We can say that we have a good footprint in Turkey and also we are keen on to be leading a digital bank in Turkey. And just a brief information about Turkey. The Turkish market is quite young and the 50% of the population is under the age of 29. So the- 50% is under the age of 29, okay. It's huge and the total population is around 80 million. Okay. So Turkish economy is quite performing very well for the last 10, nine years. And so that's why being a digital leader is quite a crucial issue for us. So with these numbers, we are performing around probably the best or the second in many KPIs. And we can say that we nominated, many times as the best bank in Turkey with the bank in Europe from some of the companies. Okay. And how long have you been there? So I've been there in 11 years. 11 years. But you said before that you're at some other banks. You've been in the banking industry for a while. Yes. I've been banking industry for almost 20 years. So I used to work two other competitors of Akbank. Okay. So I'm curious, with that large percentage of younger people, how many of those people ever come into a branch or go to an ATM as opposed to using their phone? So they should prefer doing business in phone because it's quicker, faster and easy. And the experience is quite much more under control in the phone. And we can say that we have 80, 85% of younger people are preferring the digital business rather than the classical ways. It's just fascinating to me, especially in banking, because in banking, you know, it was that trusted facility on the corner, right? In every town that you knew it was stable and it was always there and you went into the branch and you knew some of the people that worked there. And now almost the entire experience between the bank and its customers is a digital interaction, especially for the young people. They've never been to a branch. They don't hardly ever go to an ATM. In fact, the whole concept of cash is kind of funny to them. It's a very different world. So digital transformation and banking is so, so important. Yes, they're going in hand and hand. You know, the millennials are living in the digital world. So, and after the millennials, they're born in the digital world. So it's obvious that the business are transformed itself into the digital way and to deliver the products and needs in a way of doing things with the digital processes. Right. As chief analytics officer, with that move and the millennials, of course there's always regulation and other things that are driving your KPIs. But how has that migration to younger people interacting in a digital way impacted your job and what you measure, what you have to do every day? They directly impacted my job. I used to lead the customer relationship management initiative for 10 years, which covers the sales and marketing automation and the analytics and the design of the processes in the sales. And a year ago, one and a half year ago, we transformed the role into the analytics office and we are keen on the deep dive in the customer behavior and define what are the needs of the customer and how it's evolving in the digital era. And we are trying to position the banks' products and the communication skills in the digital world with the customers. So it is similar in the old days in the subjects, but it's really different in details. So the story begins to understand the customer and the segmenting the customer for sure, for the last probably more than 30, 50 years. But in the digital world, the footprint of the customer and the digital footprint is quite diversifying the thoughts in the corporate side. So we have around 15 million customers and nearly 90% is the retail ones in the new ages. So we need to optimize the banking, let's say, the cost structure of the bank and for sure the digital business gives us the enablement of the optimizing the customer service. So the segmenting the customers, not for the value basis, the behavioral and the other perspectives and creating very well-defined segments is the initial step and we are redefining ourselves in serving in this era. So I'm just curious, 20 years ago, we'll go back to 30, but 20 years ago, how many segments did you use to segment your customers? I mean, how many kind of classes and how has that changed today? Well, 20 years ago we have three to five segments. Three to five segments, that's my thought. So it's like the big ones and the small ones and if you have the analytic capability, you have the middle ones. Right, right. For nowadays we have 80, 85 different perspectives for the customers and so we created that platform to enhance the segmentation capability to serve our specified problems of the bank and I mean problems with the missions of the marketing, let's say. So we are considering now the life stage, lifestyle and some spending behaviors and some investment behaviors, some credit risk behaviors also as well and the potentials of the economic size. Right. And we can say that now we have more than hundreds but the optimal point of the segmentation is so there is no meaning to create some segments that you do not take some actions. Right, right. The actionability of the segment is quite coming forward in this topic. So we created the platform to enhance the capability to create dynamic segments and dynamic targets to each marketing event. Right. And I was going to say and hand in hand with that and you just mentioned a bunch of different variables. How many variables fed that segmentation before versus how many variables today feed that segmentation analysis? So it increases probably 100, 100 times. So we used to, I don't know, analyze a couple of hundreds of dimensions and variables in older days. It's more than 10,000 today. More than 10,000 variables to segment into hundreds of classifications and customers. Yeah, why not? Yeah, sure. Well, there's a good opportunity for an analytics of executive. Yeah. So how you address that challenge, right? So obviously you're here as a data mirror customer. How did you do it in the past? What were the things you couldn't do and what forced you to go with kind of a new platform, a new approach? So we can say that we have a quite well-defined analytics architecture in the Egg Bank and we are using different types of technologies in different types of solution areas. Data Mirror is positioned in the measuring of our marketing campaigns. And as we mentioned, we have more than millions of customers and we have quite, we can say that in a given period of time we have more than hundreds of campaigns. So we need to speed up the measurement of the campaigns and the results in a business perspective. And once we come across with the data mirror and the capabilities of the technology is much more related with the Hadoop structure and is the integration of different data sources in one place, so we think that we can optimize our ETL type of measurement data load technologies transformed into the Hadoop structure. And it seems it worked, so we reduced the time to transform the data into a single platform from diversified places. And we create an easy to use measurement platform to give some feedbacks before the things that happen. Right, right. Because there's a lot of elements to it. Just on the data side, there's the ingest, as you said, now you have many, many variables so you got to pull from multiple sources. You got to get it into a single place. You got to get it into kind of a single format that then you can drive the analytics on it. Then you got to enable more people to have the power and I'm curious how that piece of your business has evolved where before probably very few people had access to the data, very few people had access to the tools and the training to use them. But to really get the power out of this effort you need to let a lot of people have access to that data, access to the tools to design these hundreds of campaigns. So how's that evolved over time? To be frankly speaking, there are thousands of variables are related to the predictive part of the analytics. But the other critical point is, so the results are how are things are going on in the business side. So is there banking, let's say culture of bank? So we are keen on to put the business value on the front and then think with that mind and design each and every process in that way. So that's another perspective to get support, to change the classical data, load and upload and transform the data and analyze the data to see the results, that's the old way. And we were good to be frankly, but we transformed that into a much more dynamic structure and the knowledge, as you mentioned is a critical point in the team. So the easy to use, the usage of easy to use or of the technology, quite another critical point to create that type of thing into the place. So at the end of the day, you're measuring hundreds of marketing actions just in a single month and if there's something happening that doesn't plant, so you need some time to rethink on this issue and redesign it. So we think that we are at the door of this stage. So we can say that we can use the output of the predictive analytics much more in an efficient way by understanding the results in much more frequently and speedily, let's say. Right, right. And would you say this effort has really been offensive in terms of you trying to get ahead of the competition to be aggressive or has it been defensive and if you're not playing this game, you're not really in the game anymore? So it depends on the prior subject. If it's retention action, it can be defensive. It seems like defensive, but if it's, let's say, up-selection, it can be offensive. So there is no chance to choose one of them because we have a variety of products and a variety of businesses in Turkey that we are operating and at the end of the day, we need to serve each and every action. And I think it was very insightful too that you said you don't do it just for the sake of doing it and because you can do it, that if there's no action that can come from it or it's not actionable, what's the point? It's a wasted effort. Yeah, sure. At the end of today, we are doing banking business so we are not doing the analytics business. Right, right. So that's the point. Right, yeah, exactly. So as you look back, what has been, if the high level result of this effort, if you're reporting to your boss or the board of using this type of approach and then secondly, where do you go next? We're almost at the end of 2017. What are some of your objectives and kind of priorities for 2018? So we are creating, we are now just, nowadays, seeing the results of the new system and we can say that in some actions, we've started to increase the results 10 to 15 percent. 10 to 15 percent? Yes, it's in the result phase and it gives us some courage to design new use cases. Right. So the new use cases are much more related with the visualizing of the results in real time, these type of things. Right. And basically I can say that we are trying to get everything in real time. Right. The modeling in real time, measuring in real time, visualizing in real time. So we are trying to push each and every action in the analytics to the closer. We do not want to work in the offline phase. Right, in the past. Yeah, it's fascinating to me that to think that we used to make decisions based on a sampling of things that happened in the past, now we want to make decisions on all the data that's happening. Now it's a very different approach. Yeah. All right, great. Well, Attila, thank you for stopping by and sharing your insights. It's a pleasure to share. All right, absolutely. All right, so he's Attila Bayrack. I'm Jeff Frick. You're watching theCUBE. Thanks for watching. We'll see you next time.