Freo

Freo: higher online conversion thanks to artificial intelligence

Consumer lending is a fast-paced world. People are impatient and want to know immediately if their application will be approved so they can order that new kitchen or car right away. Lenders that provide quick answers are capturing the market. Rabobank subsidiary Freo wants in on this and is succeeding thanks to artificial intelligence (AI), from a data warehouse in the Azure cloud, integrated into the website.

As an online lender, how do you find the balance between speed and diligence? You want to serve customers quickly, but also protect them from irresponsible charges. Let alone that you yourself want to be damaged by non-payment. Business intelligence (BI) helps to create risk profiles and test applications for risk.

Challenge: doing more with BI and AI

Freo is playing this game to good effect. To stay ahead of encroaching competition, however, the company must go the extra mile to keep its information supply at the highest level. Near real-time is the watchword and cloud data warehousing is the chosen direction. After a successful proof of concept (POC), Valid was engaged to bring Freo's data warehouse to the Azure cloud, and to introduce AI and Machine Learning.

Valid - IT
Valid - workplace 8
Solution: data warehouse in Azure

"The advantages of data warehousing in the Azure cloud are the additional functionality, for example for self-service BI, but also the scalability and lower management costs. Valid has shown that it can be done, but also - at least as important - that they can do it. That they fit with us." So says Tom Jans, Manager Business and Risk Portfolio Analyses Rabobank and 'owner' of the project. His team will work with reports and self-service BI, in order to gain new insights from the data that will help online marketing and sales.

This fits with Freo's efforts to become data-driven. "To do more with data and make data accessible to colleagues requires a platform that the entire organization can use. With our existing facilities, we were insufficiently able to get that done."

He explains further, "Server capacity is always a bottleneck in a dynamic environment. In the cloud you can scale up and down easier and faster, on-premises you can't get that done."

Result: better customer experience

However, the project goes far beyond innovation in internal information delivery. The new data warehouse is also crucial to improving the customer experience on the website. "Freo scores very well when it comes to customer satisfaction, but there is always room for improvement. It has to, because we expect new entrants to the market who can act from a greenfield situation," says Sander Janssen. He is involved in the project as Head of Digital Rabobank.

Specific reports, preferably customized and as needed, are necessary for his department. In addition, increasing the conversion rate on the site is a key objective: going from request to contract as quickly as possible. Visitors to the site are offered a "Next Best Action" based on information about their online search behavior.

"We want to use predictive analytics to make the customer journey as short and easy as possible," Janssen says. "Marketing is also in transition with us. You used to decide a lot more based on your gut feeling. Now we are shifting to measuring effectiveness, including on social media. We are bridging the gap between the power of Google for online marketing and analytics, and internal data on customer value."

Freo's back office needs to be included in this process. It must be possible to take out a loan in a few minutes, instead of days. The new information infrastructure makes it possible to make an automated assessment of an applicant's creditworthiness in seconds, based on data. Machine learning technology embedded in the website by Valid, ensures that the predictability and effectiveness of the Next Best Action is continuously optimized.

"We are working on an innovation that is new in the financial world. There are few parties in the market that have this expertise."

Sander Janssen Head of Digital Freo
IT infrastructure consulting
Valid - man 3
Cultural Fit

Guido Ritzen, Business Intelligence Team Manager at Valid, explains what he found: "The on-premise data warehouse contained six terabytes of data and was at the end of its life cycle. The performance was below par."

Ritzen and his people built a data model, loaded the data, and developed reports. In doing so, Valid proved that what Freo wanted was technically possible. That and the cultural fit between the two companies' teams led to the assignment going to Valid.

Janssen: "There are few parties on the market that have this expertise. We are working on innovation that affects our entire company. Then you need a reliable partner, with whom you can work well, and who contributes great knowledge and experience in a domain that is new to us."

Approach

Experts from Valid and Freo set to work as a close-knit team. Janssen: "A short time-to-market was important: showing results quickly. In agile sprints of three weeks, we went through the process of development, testing and delivery of functionality."

Valid held workshops in part to help business users understand the possibilities for their work. Microsoft followed the project closely and provided good support. Because of the AVG, careful consideration was given with Rabobank to how to deal with personal data and how to set up risk management. Privacy-sensitive data will not be transferred to the cloud or will be hashed.

The cloud data warehousing project went through several phases. First, Freo and Valid jointly designed the architecture and set up the data warehouse. Then the main source systems were connected and Power BI was set up for self-service. In addition, Valid and Freo built reports and dashboards. The existing data warehouse was phased out.

In the future, the number of connected data sources can increase effortlessly, as can the number of standard reports. Jans: "Data will take center stage. Data warehousing was always the private domain of power users and business analysts. Now it is available to the entire organization."

Freo
  • Part of Rabobank since 2007
  • Online loan specialist
  • In the top 3 low-cost lenders for years
  • Maximum 5 stars from Moneyview for 7 years in a row
Freo
  • Higher conversion
  • AVG-compliant
  • Strong improvement in data quality
  • Greater confidence in the data
  • Short time-to-market
  • Latest Azure technology deployed
  • Scalable
  • Pay-per-use
Added value Valid
  • Result-oriented
  • Teamwork
  • Proven expertise Artificial Intelligence, Advanced Analytics and Data Science
  • Proven expertise Azure Cloud