The uncertain element 191
not started to use them. In fact, some of our informants had purchased a
behaviour- based policy but did not remember this before being reminded of
it in the focus group. Finally, MT also interviewed people who did not have
life insurance policies from the companies but were seen as potential customers by the market research panels through which they were recruited. This
group of informants acted as a comparison group for the insurance clients.
The data was collected in collaboration with the insurance companies as
part of a larger research project. We promised to report customer insights
that emerged in the focus groups to the insurance companies in order to
obtain access to the field, and, especially, establish contact with policyholders, a group that is otherwise difficult, if not impossible, to reach. Because
of legal restrictions, we were not allowed to recruit the customers ourselves.
Instead, they were contacted by the insurance companies. This could have
been a problem in terms of our results’ validity if the insurers had determined the ‘right’ informants for us. However, as recruitment proved to be
difficult, the selection of participants ended up being quite random.
The collaborative research design required MT to balance the roles of
independent scholar and collaborator. For instance, she needed to emphasise in the focus groups that she did not represent the insurance company.
This was generally clear to the customers, but on a few occasions, MT was
still addressed as a company representative.
The preliminary analysis of the transcribed focus group discussions
was conducted by MT. With the help of automated coding, MT searched
for extracts which entailed the concept ‘data.’ After this phase of research
was complete, MT carefully read the interviews and checked the selected
extracts, adding or removing excerpts when needed. The selected extracts
were imported into an Excel spreadsheet which MT used to conduct more
precise thematic coding by hand. Through reading, comparing, and rereading, MT classified the extracts into different thematic categories that represented experiences with personal data and behaviour-based insurance.
These codes included ‘interest,’ ‘suspicion,’ ‘imaginary,’ ‘privacy,’ ‘reliability/
trust,’ and ‘user experience.’ This coded data was discussed and analysed by
the authors in a joint data session. The initial analysis was drafted by MT
based on the data session outcomes, and the final analysis was developed
jointly by all authors through rounds of writing and rewriting.
Findings
Adopting the policy
Although behaviour-based insurance policies have previously been discussed in a variety of studies (for a review, see Tanninen 2020), these have
typically overlooked the policyholder’s perspective. Specifically, why do
people opt into these new policies and make the crucial choice of purchasing the technology? In our focus groups, people answered this question
not started to use them. In fact, some of our informants had purchased a
behaviour- based policy but did not remember this before being reminded of
it in the focus group. Finally, MT also interviewed people who did not have
life insurance policies from the companies but were seen as potential customers by the market research panels through which they were recruited. This
group of informants acted as a comparison group for the insurance clients.
The data was collected in collaboration with the insurance companies as
part of a larger research project. We promised to report customer insights
that emerged in the focus groups to the insurance companies in order to
obtain access to the field, and, especially, establish contact with policyholders, a group that is otherwise difficult, if not impossible, to reach. Because
of legal restrictions, we were not allowed to recruit the customers ourselves.
Instead, they were contacted by the insurance companies. This could have
been a problem in terms of our results’ validity if the insurers had determined the ‘right’ informants for us. However, as recruitment proved to be
difficult, the selection of participants ended up being quite random.
The collaborative research design required MT to balance the roles of
independent scholar and collaborator. For instance, she needed to emphasise in the focus groups that she did not represent the insurance company.
This was generally clear to the customers, but on a few occasions, MT was
still addressed as a company representative.
The preliminary analysis of the transcribed focus group discussions
was conducted by MT. With the help of automated coding, MT searched
for extracts which entailed the concept ‘data.’ After this phase of research
was complete, MT carefully read the interviews and checked the selected
extracts, adding or removing excerpts when needed. The selected extracts
were imported into an Excel spreadsheet which MT used to conduct more
precise thematic coding by hand. Through reading, comparing, and rereading, MT classified the extracts into different thematic categories that represented experiences with personal data and behaviour-based insurance.
These codes included ‘interest,’ ‘suspicion,’ ‘imaginary,’ ‘privacy,’ ‘reliability/
trust,’ and ‘user experience.’ This coded data was discussed and analysed by
the authors in a joint data session. The initial analysis was drafted by MT
based on the data session outcomes, and the final analysis was developed
jointly by all authors through rounds of writing and rewriting.
Findings
Adopting the policy
Although behaviour-based insurance policies have previously been discussed in a variety of studies (for a review, see Tanninen 2020), these have
typically overlooked the policyholder’s perspective. Specifically, why do
people opt into these new policies and make the crucial choice of purchasing the technology? In our focus groups, people answered this question