Silvija Seres: Hello and welcome to Lørn. My name is Silvija Seres and this is a podcast in collaboration with Telenor. Today our topic is going to be big data and artificial intelligence applied to customers inside. My guest is Liv Elise Saue Tøftum. Director for analytics and CLN for Telenor Norge. Welcome. 

Liv Elise Saue Tøftum: Thank you. 

Silvija: You’ve a complicated title. You have to translate CNL and tell us about Telenor Norge. 

Liv: CNL stands for customer lifecycle management. We’re responsible for creating good customer journeys and good personal communication with our customers. Making sure they’re taking care of the whole customer lifecycle. 

Silvija: What’s a customer's lifecycle?

Liv: It’s from when you hear from Telenor and consider to buy or subscribe to it to when you start using is and maybe you want to use more products, and then you eventually you’ll liked to leave us, hopefully not, but it happens and it’s important that they follow the customers through the whole cycle and make sure they stay loyal. Keep you happy. 

Silvija: Telenor Norge vs. Telenor? 

Liv: Telenor Norge is Telenor for Norway and where we started. We’re the main connotative provider in Norway, and the largest. We stand to significantly share the value creation on top line and bottomline of the Telenor group.

Silvija: The biggest part in terms of PNL. For me as a customer of Telenor I’m basically a customer of Telenor Norge. 

Liv: Yes, I’m working for the mobile division. We serve the customer market in Norway. It’s my job to keep you happy with our mobile product and our digital offers. 

Silvija: Is it a different job in Norway vs. Bangladesh? 

Liv: Yes and no. Core connectivity is basically the same, but the business logic is different. The consumers using patterns are also different. In Asia there's still a lot of prepaid in the market, while in Norway almost all our products are subscription based. 

Silvija: Data more than minutes?

Liv: Much more data, but it’s the same across the countries. Data uses are the core and the minutes come for free. 

Silvija: Mobile users for different geographies have very different expectations of what mobile service is and how they purchase it. 

Liv: Not with the mobile service but how they purchase it. It comes with the financial situation and how much flexibility you have, how much money you can spend. 

Silvija: Could you tell us about yourself?

Liv: I’m born and raised in Vestlandet at Karmøy. I studied a master of science in industrial economics and technology management at NTNU in Trondheim. I worked ten years as a consultant for the Boston consulting group  and then I started at Telenor. 

Silvija: You’ve three kids, 6 and twins at 3?

Liv: They keep me busy. 

Silvija: Boston consulting group was consulting focusing on strategy and then you started at Telenor. Do work in technology, strategy, data science, how would you describe your job now?

Liv: All of them. That’s what fantasizes about my job. I get the chance to work in the most interesting cross section between understanding the market dynamics in a sales marketing division, but with the technology and you’ve to understand technology development. Also the data and inside part. 

Silvija:  Boston consulting group are brilliant in understanding strategy and advising on how to adapt to change with new strategies. It’s an adaptation to change or driving the change that’s the challenge now. Old change creeps up on us as users because technology opens up for completely new services without us understanding or realising. Somebody needs to help us to understand what’s possible and efficient. 

Liv: That’s our job. It’s our job to make the customer understand the services and how they can use them, and what’s relevant for them. Alongs comes the key with personalisation because we need to make sure we don’t overflow our customers with information. Today we get a lot of information in every channel, overloaded and difficult to get through to the attention of the customers. We have to make sure whenever we speak or send a message to them, that it’s relevant, meeting the context and interest. 

Silvija: Relevance in communication and services. You do this by gathering data from customers, how do you do your job? 

Liv: We start with the data. To collect the data we need to get the data from the customer in one place. It’s about finding the insight of the data and making sure it comes out to the whole organisation so we can work with the customers' entry way. We have to understand them, talk to them and how to further develop our products and what’s important. We have to build it from product development to marketing in our sales division and into the full customer journey both digital and physical. 

Silvija: How do you get the data and how do you balance personalisation vs. privacy? By understanding my patterns of usage with my phone, and also my geographical movement given that you know my location, you probably have the best insight on how I live my psychical life than anyone else in Norway. I would rather Telenor have the information than an international monopoly. What do you think about it? 

Liv: It’s one of our main values in Telenor. We want the customer to trust us and we have a trusted brand which makes our position unique. It gives us an advantage towards the global players that know about us from a digital brand. We do collect data from our customers but keep a straight line on the regulation and what we think the customer would be wanting us to know and us. It’s important to keep the balance. We could know a lot about you, but not be allowed to track you without having your consent. We make sure that what you don’t want us to know to be anonymous before we can use it in analysis etc. We do collect data for how use you mobile, data usage and which interaction you have with us. We’re not allowed to track your web behaviour. We can use a cookie to see what you do on our website but we can’t follow you on the internet. 

Silvija: So if I buy a bikini on XXL you won’t know. 

Liv: Exactly. 

Silvija: I think people hear about data, but can’t visualise what data. What is relevant data?

Liv: It's a lot. This is where it starts to get complicated and you need to navigate. There’s a lot of relevance and it's not easy to know what’s relevant before you can gather it. That’s when machine learning and artificial intelligence comes into play. We can’t see all the connections ourselves. 

Silvija: So you gather data and look for patterns?

Liv: Yes, and try to track the customer's journey so all the interactions we have with you, we can see you’ve been in Telenor to solve a problem, did you manage to solve it? Did you call customer service afterwards etc. It’s important to understand. Not only that you did it, but that we see patterns so we can make the customers' journeys simpler in the future. 

Silvija: What’s the most important pattern that defines us? I’m a woman about to turn 49, I have kids, can you see from our patterns who we are? 

Liv: To a large extent we can. What we’re seeing for what’s relevant for you is that it might be that you’re a woman in your age and number of kids, but it can be something different. We try to see the context of the different services. It’s different if you’re a young, single student, and we try to tailor to those needs. We’re building the product so it fits with the customers needs. 

Silvija: You said the simplest is often the best when it comes to data. Sometimes we over architect the gathering and analysis process? 

Liv: It’s easy to think that AI is the solution to everything. Sometimes it’s the basic statistics that provide the best insight. Before you can look for insight you need to understand what the problem is. To understand the business, business logic and how the customers behave in a certain way. Then you need to collect the data and it’s extremely important. It’s not necessarily the most advantaged algorithm that’s providing the best insight for the organisation to provide great customer service. It’s knowing the question and finding the right technology to solve it. 

Silvija: Can you tell us about your projects? One is democratisation of data, your new, big data platform at Mjøsa and the idea of causality vs. correlation machine learning. 

Liv: Democratization of data is important. It’s about scaling the insight and get it into all sides of the business. It’s about gathering the data in one place and having one system for self service, if it’s product development, marketing or sales channels, they know they’re problems but can find the data there. 

Silvija: This is important. Too many software companies work with this kind of data insight in a centralised way. They’re a few high priests of data tuning and insight that know the heart of advantaged alorightm. You’re dependent on them giving you the consequences. If your getting the whole organisation to grow up in the dataworld, then maybe the questions you’ll asked will get better. 

Liv: Yes, we’re noticing that. We can make better decisions where the decisions should be make. Out in the organisation where the experts work. 

Silvija: Do you’ve to educate them on data science?

Liv: We do have to train them in the tools that they’ll use. It has taken time and we have invested time in it, but it’s paying off. One of the beauties of it is that before we had a lot of requests about the insight analysis that people wanted to know about and then they had to wait for weeks because it took time to produce them and got a reply back with “this is interesting, but I have a new question”. Now they can do it themselves and use our data scientist to focus on the more complex problem like the causality analysis that you talked about. Causality i.e is we have a product called “family bonus”. Customers who have more products get free data as a bonus. The customers that have family bonuses turn in less than others. We don’t know if that’s because they're more loyal or it’s the family bonus. We can only observe it. Then we use the machine learning algorithm to help to find a difference between what is causality and correlation. Causality is what the effect family bonus has. We can scientifically prove that it has a positive effect and see that turn is reduced for the customers using family bonus. The effect is largest in the beginning and then it decreases. That helps us to work further on with the product to develop it and keep it relevant. 

Silvija: How we buy, use and leave. But in a way you understand, not correlation but causality. Why do we confuse those two?

Liv: It’s easy to see and it makes sense. We generalize, simplify and sometimes you over simplify. 

Silvija: You worked mainly based in Telenor Norway. Can you expand it internationally?

Liv: Yes. To make sure it’s scalable you need to look at the total technologies stack to make sure it’s connected. That is one of our successes. We managed to build it so we have the data, analytics and put the models right into production and automise it so it’s real time making sure you get relevant information and communication for us when you need it. 

Silvija: You can also scale it in other countries so their data about the customers could potentially be a part of the same calculation system. 

Liv: It's something we could share and they could use the same infrastructure. You have to take into consideration who you are today. If you have a legacy which we have in Telenor, by being a traditional company with a history, you need a starting point before you continue. You need to have a target and build step-by-step. You start with the software companies, they have the advantages of starting from scratch. Is easy compared to the complexity when you build up the silos over time. 

Silvija: It’s digitizing your old infrastructure by putting the intelligence data on top of old mobile information. From frequencies and licenses into a world of routes. It’s a different game but it can be translated. What is Mjøsa? 

Liv: Mjøsa is our big data lake. We have to call it Mjøsa since that’s Norway’s biggest lake. The data center is located close to Mjøsa. It took a long time to build a data lake for Telenor Norway. We have strict regulations being the national telecommunication operator. 

Silvija: What’s a data lake?

Liv: Is one place to store all the data you need to make sure you have a source of data and collect them to build algorithms. When it comes to AI and machine learning the key is to have sufficient data. 

Silvija: Why called a lake?

Liv: Because it’s big and a place where you put the data. It’s a collection where the data flows around and picks the data you need. 

Silvija: It has to do with the structure. What do you lack in terms of skills whether it is accessible people for your team or in general to work with this kind of work? 

Liv: One of the key success factors of our team is that we’re able to put our data managers, data engineers and data scientists teams close to the business side. You work in sales and marketing, not IT. That's the core. It’s not just to find educated minds, but to find the right thinkers. This is also about understanding the complexity of the infrastructure we have and how you can utilise it, and understand the dataset we have. It’s more complex than it sound. It takes a lot of years to educate a good data manager or data scientist. The campaign managers and the ones executing our activities towards our customers. 

Silvija: Diversity of background or ability to combine two different background 

Liv: Cross functionality is important. We see more of the need for UX-resources, user experience resources. Who can make the journeys better and content developers, the ones who're writing the content and making what we present to our customers. They have to be linked into our team. 

Silvija: A talke to Ola Jo about women in technology. Tricky to find women? 

Liv: We do find women, but when you look at data scientists and data managers the vast majority of the people and applicants of our team are men. But for a while in my management team we were 80 per cent women. We just had a chance to we’re down to 60 per cent, but that’s brilliant. 

Silvija: If you could encourage young women to study to be like you, what would you tell them?

Liv: They should study technology because that’s where change is happening. It’s defining and building our societies. In Telenor we have the privilege of affecting our customers in their everyday life. It’s important to understand how to best use the technology in the best way for our customers. 

Silvija: The game is more fun and more efficient if you have some technology background as well. Then you can’t be fooled. Having confidence by being trained as a technologist, but then applying it for the stuff you're passionate about is a nice combination. What would you like us to learn more about?

Liv: We’re trying to train the organisation to always think about data from the start of a process. Think about how we should collect the data and make sure we get it into our data storage, so we can use it in a unified way. Very often we plan and develop products, then we start using it, and we think “we should have insight on how the product is working”, if you have it developed in a silo it takes a lot of time to incorporate the data. Data first, get the data collected, that’s 80 per cent of the work. The next thing is that you need to understand the business. Understand the problem and what kind of insight can the business benefit from. How we make good decisions based on insight is crucial. The fancy algorithms and machine learning is the frosting on the cake. 

Silvija: You said you read a lot of books to learn more, but you learned the most from talking to people in your organisation. I like the answer. It’s the first step in defining a learning organisation, people are sharing the knowledge in a relevant way. How did you do that in Telenor?

Liv: In my team we have a culture of being curious and we ask. We share and people are enthusiastic about what they’re doing, so if someone ask them they share everything they know. It’s about getting together the cross functional teams because the people understand each other. 

Silvija: Do you have a quote? 

Liv: I think things will change all the time. It’s an Indian quote that says “if you sit still the grass will rot underneath you”. It’s a way to show you’ve to keep on moving, but also when it comes to moving you should have a strategy which is not too firm, but a direction. You need an organisation that’s ready to change and learn. That’s key to analytics and insight. We have to circle the feedback when we do and test things towards the customers we’ve to understand what’s working and not, and continue to improve. 

Silvija: I agree. You have to be knowledgeable and brave about the direction and efficient on adjustments as you go along. If we’ve to remember one or three things?

Liv: Understand the problem you want to solve. Always start with the data and make sure you build an infrastructure that’s making it possible for you to scale and automise. 

Silvija: Liv Elise Saue Tøftum, thank you for coming to Lørn and teaching us about data insight into the new customers. 

Liv: Thank you. 

Silvija: Thank you for listening.