Hypofund

Abstract

This is a project report on the research, implementation and empirical test of an online system for mortgage advice that have been completed in the context of our bachelor project. The project objective was to build a platform for online mortgage advice, which could substitute a real-life, human adviser. The use of such a platform costs far less than hiring a human adviser, but a major problem to solve is that a website usually cannot give personal treatment to the user. In the research phase, we looked at how to apply the fields of affective and adaptive interfaces to our problem domain, as well as how A/B testing is commonly done for software products and how the results are statistically analyzed. With the knowledge we gained from the research, we came to an agreement of what to build with the client. The result of this agreement is the specification of the ’hybrid workflow’ interface (tailored to the philosophy of the client’s vision), a personalisation feature and an empirical test. We implemented said specification and conducted a proof-of-concept A/B test with 47 users. After statistical analysis, we concluded that the results of these tests were not significant: there was no significant difference between the users that got a variant of the platform with the personalisation feature and the users that got the variant without our feature. In the discussion segments, we discuss the results and make further recommendations.