Home / Fintech / AI, machine learning, data and Nordic banks

AI, machine learning, data and Nordic banks

Last week we had our first meetings of Nordic Finance Innovation, the largest networking group in the Nordic region for people involved in finance and technology which I happen to chair. The meeting focused upon data: data usage, data analytics, data leverage and data privacy. It was an interesting meet-up and Danielle Neben, editor at Nordic Finance Innovation and Marketing Director for ePassi, provides a nice summary of the events as they happened.

It’s all about the data

By Danielle Neben, Editor at Nordic Finance Innovation and Marketing Director for ePassi in Iceland

Nordic Finance Innovation hosted its first 2019 meetings in Oslo and Stockholm on February 6th and 7th. We had a range of business leaders from banks, insurance companies, retailers, consultancies, big tech and innovation centres. A full list of the presenters is available at the end of this article as well as upcoming events.

Many of the topics followed the themes that Chris Skinner is now approaching in his new book “Doing Digital” – lessons from leaders from 5 international banks who have embarked on this journey. Becoming a fintech bank, where Tech is on an equal level of importance as Finance.

The topics that were covered during these two days include:

  • Platforms and big tech as catalysts

  • Being obsessed with the customer

  • mobility on demandBurning sense for change and communication

  • AI bringing exponential development – how to commercialize

  • GDPR and Ethics

  • Finding new ways to collaborate, attract and retain talent

Platforms and big tech as catalysts

All businesses are seeing the big tech giants as catalysts. The music and movies industries were first impacted. Companies like Nokia and Blackberry did not take notice of the smartphone launched by iPhone and within 3 years they were bankrupt. For banks, this is probably 7 years.

This leads to the conversation around the bank branch model. Ten years ago, they were seen as dinosaurs. But today they are marketing showcases. And having a physical presence gives trust.

We see this with the major tech platforms, where they are combining physical stores to complement their online businesses. For example, Amazon has direct investments in Whole Foods and Amazon Fresh, while engaging in alliances such as Morrisons. Google has alliances with Tesco, Target and Carrefour. Walmart has partnered with JD.com with a new shopping application available through WeChat to skip checkout counters. And Alibaba’s Hema stores are changing the supermarket experience.

At Bertel O. Steen, the dealership model has been turned on its head. Previously, customers would visit 4 times before making a decision, collecting information provided by the dealer. Today, they have this information online. They will only visit a dealership once to make a decision – the dealer has only 1 chance to make the sale, now with a focus on service, not product.

But do the tech giants really get data? Google Maps will recognize your home, but not your cabin. Netflix lose customers 20 minutes after watching a movie they are not interested in – this is a content model, not an AI one.

Being obsessed with the customer – mobility on demand

The customer is core to the business. In Chris Skinner’s interview with the DBS CFO, he focusses on customer statistics, not ROI or the shareholder. Because if they have happy customers, this will deliver good ROI for the shareholders.

At Bertel O. Steen, the customer now demands mobility. They no longer want to buy cars, they want to use them when they need them.

Irmelin Stolen’s challenge, from Rema 1000, is to bring her grandfather’s customer centric model into the present. He knew everyone’s name, their shopping preferences, price sensitivities and even where their house key was to make home deliveries. The challenge is keeping customer focus – not just collecting data, to help with better decisions and customer experiences.

“The battle of the shopping list is very hard to beat”, as per Truls Fjeldstad of Norgesgruppen. They have taken hundreds of discarded lists to analyse what people write. They do not write brand, they write dinner, milk. The big question is how to digitise this approach and tailor services to the client. Challenges in Norway are that the language is not available on Google for voice recognition. Truls estimates they have around 3-4 years before the Googles and Amazons will adapt to Norway.

Holistic view of the customers

Traditional business models have been built around product lines and not the customer view. Kim Hyldig at Worldline challenges business and whether full CRM 360 is realistic, where individuals receive appropriate services according to their needs, and not what products the businesses want to sell.

Impatient customers

Siri Børsum at Google asked “Do you know how long it takes to load a page in your company app? 53% of visits are abandoned if a mobile site takes longer than 3 seconds to load. And for every second it takes; conversation rates drop by 20%. Speed is about a first impression.

Burning sense for change and communication

How can a traditional business with good returns actual make the move to digital? Creating that burning sense for change and communication for buy-in are required to change the direction.

Platforms, the perfect storm

Girts Berzins uses simple messaging to motivate this change at Swedbank. First he shows that the tech giants are “Platforms, the perfect storm”. And he uses two key phrases to drive this understanding: “API or die” and “AI or die”.

Every project to be 10X

At Nordea, the focus is on 10x projects – not 10% incremental. How can service be fundamentally changed in relationship to the customer.

Having clear and urgent communication, with relevant KPIs for the entire business (not just the tech labs) help align the organisation to change this direction for the future.

AI bringing exponential development – how to commercialize

In the last two years we have created more data than in the history of mankind. Spotify and online grocery shopping are driving customer expectations. Over 30% of Google searches are done through voice. On singles’ day, Alibaba sold $15 voice units, Tencent has put voice units into hotels.

But how is AI being applied in the Nordics?

Folksam and estimating how much it costs to repair a damaged car

Every day, there are over 500 crashed cars with pictures taken of all of them. The challenge is using deep neural networks to estimate the cost to repair a damaged car. This case was a good fit to use data to establish models. The pictures alone were not enough to come up with conclusive estimates. Key factors in the model brought better results, including car model (Mazda), position of the damaged parts (front door, bumper) combined with the pictures.

One of the key takeaways at Folksam is that more data is not the solution; successful AI experiments require domain knowledge. Short experiment cycles and demo driven workshops will pivot their perspectives to find new use cases.

Worldline and Payment Fraud Detection

Worldline is applying expert rules, AI and machine learning to detect fraud. Objectives are to increase accuracy, have less rules (which takes machine power), adapt faster to threats and avoid bottlenecks in the system. The data set is very large and can be more efficient. So far, accuracy rates have improved by 20%.

High level recommendations by Kim at Worldline on AI:

  1. Find projects where you have 100% control of data

  2. Ensure you have the needed insight, capabilities in the projects – Business/IT/data-analytics

  3. Unless you can put a price tag on your learnings or increased insight in the capabilities of the AI technology, make your initial BC qualitative (budget for experimentation). The price tag should be on learning in the organisation with AI

GDPR and Ethics

“Today, clients want to be recognised as individuals, the only way is using data and AI. People want powered assistants, not clicks. Clients will have shared meetings with Google, Amazon and banking at the same time. Shared meetings will hugely save customers’ time. Transactions are embedded. Human activity generates data – you need consent to obtain this data (a function of trust), then put the AI and application. This is the storm and you need consent”, say Girts Berzins from Swedbank

Santander has 10 years of saved conversations with data with clients, but the key question is how can they use this data in line with GDPR.

How to be responsible with AI and ethical dilemmas? How do you do digital data marketing without being creepy? Catharina Nes at Datatilseynet gave an excellent overview on ethics in AI.

To have AI, you need data. China has access to vast amounts of data, yet they do not have the privacy legislation in Europe. They have much wider opportunities to exploit this data and build AI systems. There is a recent article: “Big data meets Big Brother as China moves to rate its citizens”.

In Europe, the response is GDPR, a framework for trust. French president Emmanuel Macron has said that badly implemented AI could “jeopardize democracy”. The EU has recently published draft “Ethics Guidelines for Trustworthy AI”.

Norwegian banks have a high level of trust. In a recent survey by Datatilsynet 77% of Norwegians trust their banks, but this drops to 31% in European banks, 25% of US tech companies like Apple or Facebook and 9% in Chinese tech such as Alipay. As such, trust is a major competitive advantage and going forward, it is important for banks to consider when they build new business model and keeping the trust.

Catharina gave practical recommendations when purchasing and using AI-based systems:

  • Carry out a risk assessment before you start to use a system

  • Demand the system satisfies requirements for privacy by design (GDPR, art 25)

  • Test the system regularly for bias to avoid discriminatory treatment

  • Ensure you have good systems to protect the rights of data subjects, such as right to information, to access, to deletion and to explanation

  • Functionality must include consent, which can be withdrawn

  • Establish code of conducts, ethical guidelines or a data protection panel

Finding new ways to collaborate, attract and retain talent

Bringing innovation into companies and cultural change is central to a successful digital transformation strategy. We heard a variety of initiatives that are helping with this change.

External partnerships are a major theme. Worldline has a University collaboration program on data having invested EUR 100 million for innovation. One of the objectives is to tap talent and bring this knowledge in house for development.

Epicenter in Oslo is a digital innovation house for corporates and scaleups. They aim to share competence, matchmaking between companies big and small, and go to market strategies. Rema 1000 and Bertel O. Steen are two corporates with teams located there to leverage these opportunities and create an environment for innovation.

Amazon has created the “two-pizza rule for micro-teams”, where every internal team should be small enough that it can be fed with two pizzas. These small teams are based on agile: discuss, develop and deploy. Developers, designers, auditors, data scientists, data engineers and business people work together. They are moving away from yearly cycles to a week-end.

At DNB, an isolated innovation team is not the answer. Creating engagement with objectives and shared agendas is hard work in a matrix organisation. DNB is putting resources as close to business problems with the central organisation as the conductor for the shared agenda and communication to make the invisible, visible in its digital transformation.

Attracting new talent is key. Bertel O. Steen recently had an award-winning campaign for recruitment, which translates to “The last person to take a driver’s license is already born.”

In addition to attracting new talent, upskilling their people who understand the business is the challenge. For example, at Bertel O Steen – they would like to invest in their mechanics to learn about big data. Chris Skinner spoke of how JP Morgan traders are learning Python and China Merchants Bank have a staff program to learn to code 2 hours a week over an 8 week period.

Focus on partnerships, talent and upskilling as part of digital strategy is core to its success.

Presenters at the NFI meetings in Oslo (6 Feb19) and Stockholm (7 Feb19)

In Oslo

  • Chris Skinner, Chairman of Nordic Finance Innovation

  • Iren Tranvåg, Chief Executive Officer of Nordic Finance Innovation

  • Irmelin Stølen, COO, Rema 1000

  • Are Knutsen, Head of New Services and Innovation, Bertel O. Steen

  • Truls Fjeldstad, Director of Marketing and Business Intelligence, NorgesGruppen

  • Aidan Millar, Chief Digital Officer, DNB

  • Siri Børsum, Leader of Digital Transformation, Google Norway

  • Catharina Nes, Specialist director, Research & Analytics, Datatilsynet

  • Kim Hyldig, Business Development Director, Worldline

  • Morten Krogh-Moe, CEO, Sannsyn

  • Øyvind Ertzaas, Chief Risk Officer Norway, Santander

In Stockholm

  • Chris Skinner, Chairrman of Nordic Finance Innovation

  • Iren Tranvåg, Chief Executive Officer of Nordic Finance Innovation

  • Girts Berzins, Head of Digital Innovation & Strategy, Swedbank

  • Kim Hyldig, Business Development Director, Worldline

  • Lars Engvall, Head of FutureLab, Folksam

  • Mattias Fras, Group Head of AI Strategy & Acceleration, Nordea

  • Rune Rasmussen, IBM Global Services

  • Johan Jerresand and Alexander Fritsch, Partner and Director, PwC Consulting

  • Aage Reerslev, CEO and Co-founder at Wrapp

We would like to give special thanks to our two sponsors for these events: Rema 1000 and PwC

Upcoming events

Do follow us on Facebook, LinkedIn and Twitter to hear of upcoming events.

Below is the planned schedule of events for 2019:

  • Madrid (3 April): United Europe enabled by Technologies – how NFI members and visionary partners can collaborate to drive the innovation agenda in Europe

  • Copenhagen (29 May) and Helsinki (30 May): Sustainability and Banking

  • Dublin (30 August): NFI Women visit to Facebook

  • Amsterdam (26 September): NFI Summit

  • South America (October): NFI Executive Tour

About Chris M Skinner

Chris M Skinner
Chris Skinner is best known as an independent commentator on the financial markets through his blog, TheFinanser.com, as author of the bestselling book Digital Bank, and Chair of the European networking forum the Financial Services Club. He has been voted one of the most influential people in banking by The Financial Brand (as well as one of the best blogs), a FinTech Titan (Next Bank), one of the Fintech Leaders you need to follow (City AM, Deluxe and Jax Finance), as well as one of the Top 40 most influential people in financial technology by the Wall Street Journal’s Financial News. To learn more click here...

Check Also

The Finanser’s Week: 10th June – 16th June 2019

The main blog headlines are … Challenger banks don’t matter as they’re not a primary …