Olof Bartling is since September Head of Group Data Analytics.
“The common thread in my career has always been data and analytics combined with commercial responsibility. My role has been to use data and analytics to generate business results,” he says.
Group Data Analytics
The Group Data Analytics unit, which consists of about 15 employees, has three focus areas.
First, there is a team of data scientists under the direction of Chief Data Scientist Ala Tarighati. They work as internal consultants and use their expertise to help the business divisions implement prioritised data-driven projects.
The second area involves strengthening the bank’s overall ability to work in a data-driven manner by creating tools, routines and governance models for machine learning, AI and data analytics.
The third area of Group Data Analytics is coupled to research, innovation and new technology.
“Development in the area is progressing very rapidly, and it’s essential to constantly stay abreast of what is happening in data analytics, AI and such, to ensure we don’t lag behind.”
It’s often said that data is the new gold. Is that true?
“Data itself is a passive object, sort of like books in a library. It’s when you read books and obtain knowledge that value is created. It’s the same with data. It is when data is used to create models that can make decisions, make SEB more efficient, and create benefit for customers that value is created. Then data turns to gold.”
What does creating a data-driven organisation entail?
“When you go from a traditional, expert-driven model that builds upon experience-based decision-making to a data-driven model, there are certain things that change. In a data-driven world, you allow data to show what is right and wrong, which can sometimes contradict old truths. I have experienced this myself many times.”
Working in a data-driven way therefore entails testing, collecting data, and continuously developing.
“The first model is a starting point. It may be simple and rudimentary, but by testing different variants, you will get feedback that shows if you should move in one direction or the other. Over time the model will be optimised. The more tests you can run simultaneously and the faster the cycles are, the faster you can learn. Everything is a continuous learning process.
“It is therefore important to work in an agile way – to increase the speed of implementation and get data to flow so that it creates a fast feedback loop.”
Are there risks from a human perspective in allowing data to decide?
“There is always a risk that things will not work as thought. This can be due to a number of different reasons, such as poor data quality or human error. This is why it is so important that we have employees with the right expertise and that we create simple and clear work processes, to make it easy to do the right things and help us detect if something is wrong.”
So of course there are risks, but the opportunities outweigh these, according to Olof Bartling. Even human decision-making has its own inherent risks.
“The advantage of allowing models and machines to make decisions is that they can be extremely consistent and can make an incredible number of decisions incredibly fast. We know that as people we are not always consistent. We have opinions and previous experiences that lead us to interpret situations and information differently, which can result in us making different decisions in the same situation with the same information. With the help of machine learning we can make sure that we make decisions only on the basis of the parameters that we have set.”
What is the relationship between short-term sales and long-term customer relationships in a data-driven operation?
“There is no difference. We are the same bank, with the same strategy and focus on long-term relationships in a data-driven world. However, a data-driven operation has a better ability to balance the relationship between short- and long-term results since it can measure and more quickly influence results in another way.”