Anadolu Sigorta brings “everyone onboard” across the business to become truly data driven

Identifying and investigating fraudulent claims is extremely time consuming for many insurance companies. In 2017, there were approximately €13bn of detected and undetected fraudulent claims in Europe. These claims take valuable time away from insurers who need to treat genuine claims and serve customers quickly and efficiently. However, many insurance companies are stuck with disparate data sources, legacy systems and many employees working manually, making the ability to identify and weed out fraudulent claims extremely difficult.

Anadolu Sigorta was founded in 1925 under the leadership of Isbank, Turkey's first national bank. Today it is Turkey’s first national insurance company, employing over 1,200 staff and offering a wide range of non-life policies to customers throughout the country.

Anadolu Sigorta is on a mission to implement a customer-focused approach to service, deliver new products and create a more efficient way to spot fraudulent claims - using data. In short, they have put data at the core of their innovation strategy.

Here, Vedat Gunes, BI & Analytical Applications Manager at Anadolu Sigorta, shares how the company has transformed into a fully data-driven organisation, predicated on a powerful data culture where employees are developing business strategies and making decisions, based on data.

Data has always played a big role in the insurance industry, for the underwriting process of policies and assessment of customers, but now it’s at the very heart of the company, driving everything we do. All of our new business strategies and plans are now built on data, while prediction-based applications are being used to open new doors all the time from an operational perspective. In addition, we developed a social network platform to identify organized fraud cases and served it to our claim department usage and also we replaced our central fraud platform with inhouse developed fraud application and integrated to organized fraud detection application.

We’re creating a more accurate 360° customer view, which in turn enables us to optimise pricing, develop new product offerings and put more meaningful KPIs in place like Customer Lifetime Value (CLV), health anomaly metrics, and campaign predictions parameters.

The main challenge is the efficient preparation and modelling of data. As a 95-year-old company, we had many legacy systems and applications throughout the business, creating a fragmented ecosystem that was extremely hard to work with. In many cases, we had two or three different production environments for the same process, which we were combine together in our unified data model on our data warehouse before attempting to extract insights. This led to problems with data quality and reliability, as well as effective governance, all of which needed to be resolved in order to achieve our data transformation goals which are shortening data preparation step for analytics and visualization, faster time to market, and simpler integration process to operational systems

We also needed to overcome some cultural challenges. Many of our employees and external distributors were accustomed to working with list-based reports, so were resistant to change when it came to modern data visualization trends.

Firstly, we’re much more comfortable sharing insights with business users now, using highly visual, interactive Tableau dashboards, which enables them to explore their data with ease. Rather than trying to input everything into our data warehouse first, we can combine multiple data sources directly through Tableau and visualise it all together in a single dashboard, greatly improving the data quality and reliability in the process. Employees and customers know they can trust the data when using it to inform key decisions. Furthermore, Tableau’s extensive suite of security tools gives us tight control over the data from a governance perspective.

Once our employees and customers started to use Tableau, they quickly fell in love with it, leading to a rapid cultural shift that we thought would take a lot longer to achieve. We see an increasing number of business units adopting Tableau to improve their own business processes from business development to fraud prevention. Now all of our distribution channels use Tableau over legacy dashboard platform, which was Qlickview.

We also use Tableau embedded analytics to distribute key data to over 4,500 insurance agents around Turkey. Tableau’s automated report functionality means they have round-the-clock access to the latest customer details, claims information and product lists, all updated on a daily basis. Using the agency portal, greatly reducing the demand on our central BI and analytics team here.

We are serving all visuals related with central fraud platform to our internal users with Tableau dashboards. We can say Tableau is the GUI of our fraud application. Users can consume all fraud related information via Tableau platform.

Key data sets such as claims made, frequency of claims, transactional data and provisions can all be combined and analysed together in a single Tableau dashboard. Rather than relying on tables, employees can ask questions of the data and explore the insights they are uncovering. This makes identifying anomalies and marking them for further investigation much more efficient, which has dramatically cut fraudulent claims going unchallenged.

Identify and use one platform to consolidate all your data! Also, make sure you consider all business users and their needs. Getting input from a wide range of people at an early stage will help avoid resistance further down the line. The IT department isn’t doing data transformation for itself, but the chances of success will be much higher if everyone is on board from the start.