Case Study · Insurance
CosmosDirekt: more completed insurance applications
CosmosDirekt sells insurance without brokers, directly online and by phone. Its lead product is term life insurance. The online application flow is a checkout with more required fields and more need for explanation, and that is exactly where there is plenty of room for clean A/B tests.
I'll show you where your shop loses buyers.
A checkout with weight
People take out term life insurance to protect their family. It is a weighty decision, and the flow asks for a lot, from personal details to health questions. Every field is a moment where the applicant feels reassured or unsure. This is exactly where every good hypothesis pays off.
Some steps are set by regulation and stay as they are. Others have grown over time, because at some point someone added a field or a note. From the outside both look the same. In a large company, every change to the flow also needs internal sign-off. A testing concept that plans for those sign-offs from the start gets into production quickly.
Hypotheses that survive sign-off
The basis is an analysis of the application flows. Where do applicants drop out, and which of those points can be changed? We classified every step as either set by regulation or grown over time. That classification is the basis for all prioritisation. The steps that grew over time are the material for the A/B tests, and there is pleasingly much of it.
The real skill lies in writing the hypotheses. We built them so they get through internal sign-off and still deliver a clear signal. A variant needs enough distance from the control for the result to be usable, and enough legal cover to go live. We found that balance together with the internal team and set up A/B tests in the online application flows from it, with the lead product as the focus.
What regulated providers can take away
In our experience, the most room in regulated flows lies in the steps that were added at some point and that nobody questions today. Separate them cleanly from the mandatory steps and you suddenly have plenty of test material. Just as important is turning them into hypotheses a legal department understands and can approve. A good testing concept in a large company plans for sign-off from the start. That costs a little time up front and brings pace to every test round after.
Let’s start with your shop.
In the Dropoff Analysis you see where your shop loses buyers. Afterwards you know where the first step pays off.