How data science can save traditional banking and insurance from GAFA
December 12, 2016: In 2016, there are very few industries that big tech players (Google, Apple, Facebook and Amazon – or GAFA) don’t have their hands in. The finance industry is no exception; after having recovered from the global economic crisis of the early 2000s, banks and insurance companies now face a new threat from these large tech companies as well as small, nimble, well-funded startups.
Silicon Valley poses a real threat to traditional banking and insurance industries; but companies in those industries still have the opportunity to fight back.
Big Data: Big Tech’s Weapon
The largest advantage that financial startups and GAFA have over traditional players in the industry is data science. Data science and algorithms are GAFA’s bread and butter – they’ve built their businesses based on complex algorithms and have made huge investments in data science. Similarly, startups in the space are arising out of a mobile, digital era, so they are naturally in tune with big data, having been born out of that era.
Another big differentiator is that GAFA and startups don’t view data science just as a job title for one person at the company; data science is core to the business and driving it forward. It’s the collaborative discipline (the combination of people, data, tools and processes) used to transform raw data into actionable insights and business innovation.
As a result of data science, cutting-edge, technical companies are able to make strides in the financial space by leveraging data mining and predictive modelling to do things like:
- Personalise offers
- Reduce risk
- Create disruptive new products
- Expand markets
- Minimise operating expenses
- Automate traditionally manually processes
And this is just the beginning. With the rise of the internet of things and advancements in machine learning, there will be much, much more.
Big Data in Traditional Banks and Insurance Companies
Thus far, unfortunately, traditional banks and insurance companies have been slower to embrace data science and use advanced technology like predictive modelling and machine learning to improve business. However, those that have started to use big data for meaningful insights have done powerful things, including:
- Enabling personalised policies and premiums in the auto insurance market. Allianz offers a car insurance policy leveraging IoT that allows data tracking through a mobile app and a GPS-equipped dongle. The company can now deliver personalised pricing models, better understand customers, decrease fraud, and encourage positive driving behaviour.
- Proactively preventing customer defections in banking. Bank of America used transaction history to develop new behaviour models of mortgage customers at risk of switching. Data scientists built models based on these recommendations and pushed retention offers to at-risk clients.
- Discovering new customer segments. A large retail banking group mined large archives of transactional data to develop new models of customer behaviour. From these models, marketing and sales identified new offers and promotional campaigns with a threefold improvement in conversion rates.
- Automating life event marketing. A mid-sized insurance company combined customer relationship management (CRM) data, contracts data, weblogs, and social media data to analyse and develop predictive models about when life events were likely to occur. The marketing team was able to successfully develop new event-themed, personalised campaigns automatically triggered in real time.
- Efficiently and accurately detecting fraudulent claims. A large provider of supplemental insurance combined siloed data sets and created fraud detection algorithms, which routed claims real time based on their likely validity. This proved to be three times more effective at fraud detection than the legacy approach.
When competing with the likes of nimble startups and GAFA, data science is clearly the path to levelling the playing field. And in some ways, it even allows traditional banks and insurance companies to get ahead because these businesses have some additional advantages over startups and GAFA. For example, the traditional finance industry already has deep troves of historical (and largely untapped) customer data, physical touchpoints (or branches), a high level of consumer trust, and a deep reservoir of professionals with extensive domain expertise and advanced quantitative skills.
Once traditional finance companies realise the value of using big data to drive business decisions and implement innovative machine learning techniques to get ahead, the next step is actually putting a data lab in place.
Again, banks and insurance companies have the advantage of a professional staff with deep experience in the industry coupled with quantitative skills. These are perfect tools for assembling a data lab; it’s not necessarily about hiring new staff, but assembling a team of the right staff from the business and IT side that will be able to effectively communicate, collaborate, and execute on projects.
Of course, once assembled, this team also needs access to data and an effective tool with which to qualify and prepare that data as well as easily deploy data projects to production. But already equipped with the right staff to make up a team, this is the easy part.
Ultimately, equipped with the right people, processes, and tools, traditional banks and insurance companies can avoid the fate of becoming the back-end plumbing for GAFA and other challengers. They can appropriate the advantages of these newcomers and merge them with their own to become the new marketplace innovators of the 21st century.
Florian Douetteau is CEO and co-founder of Dataiku