Digital Product ResearchWe’re a team at Co-op Digital and this is the archive of our experiments… Read more

Protecting your stuff

A lack of trust has broken insurance

People make fraudulent claims, which means insurers create policies that make it harder to claim, which means bona fide customers can’t make a claim when they need to.

Insurers don’t trust customers. Customers don’t trust insurers. Insurers collude to raise prices and don’t trust each another... And the regulators don’t trust the insurers.

There’s space for Co-op to take a radically new approach to insurance: local, mutual, community focused and owned by members.

In collaboration with colleagues in Co-op insurance, we explored this alternative model built on trust. If people with similar risks could form groups and insure each other, they might be more likely to trust one another.

How might we create an insurance model where people trust each other, and where members feel confident making claims?

Small groups with something in common

We created a prototype of Protect Together. Small groups of people with something in common (living in the same area in the same type of accommodation) could form a group and insure each other. Members would pay a regular contribution to a fund, and claims under a certain amount would come out of that fund. The group would decide what to do with any money left over (split between members or invest in their community to reduce the group’s risk), or what action to take if the fund ran low. We showed the prototype to six people who fit the requirements for our example group.

A woman tests out the Protct Together prototype on a laptop guided by one of our team
Testing Protect Together with a research participant

Could we explain the concept?

A new model of insurance needs a new way to explain it

The Protect Together home page did a poor job of explaining what the service was at a glance. While all respondents recognised it was something to do with insurance, none were entirely clear how it worked.

I don’t know if it’s insurance for a group or if you’re putting money aside in case you need it - it’s a bit ambiguous.

- R, female, 26-35 year old, owns a laptop, tablet, phone, jewellery, TV

Having started to sign up and clicked through a few of the screens, respondents then understood more clearly what Protect Together was. They all felt it aligned with the Co-op’s overall values.

I think it’s a lot cheaper, it’s nice that your neighbours are going to be involved, and I’d trust it more being Co-op.

- A, female, 26-35 year old, owns a phone

Respondents saw the difference from traditional insurance.

It’s along the lines of a group insurance scheme. But the money you’re investing you can get back which is completely different to any other kind of insurance scheme.

- P, male, 36-45 year old, owns a phone

Could we create trust? How do we define communities and groups?

Living close to someone isn’t a strong enough reason to trust them

Despite living in the same area, none of the respondents felt a particular affinity with our example group

Despite living in the same area, none of the respondents felt a particular affinity with our example group (people renting flats in the Green Quarter, Manchester). They either felt that they didn’t know their neighbours, or that the age group wasn’t right for them.

I wouldn’t trust my neighbours because I don’t know them.

- R, female, 26-35 year old, owns a laptop, tablet, phone, jewellery, TV

I’m too old - my son is 21! If there was an age group 33+ I’d have much more in common. Some 21 year olds are very focused, but some are not!

- C, female, 36-45 year old, owns a laptop, tablet, bike, jewellery, TV

Is there a need for this type of insurance?

Renters don’t feel that they have a lot that’s worth insuring

We focussed on people renting property for the test because of the complexities involved with building insurance claims but we learnt that several of the people we interviewed did not see a need to insure their contents in a furnished flat.

TVs, laptops, I don’t insure anything like that, I’ve no contents insurance either.

- A, female, 26-35 year old, owns a phone

Should members of the group be visible to each other? Is transparency better than anonymity?

People felt they’d need to talk to the group

When presented with the emails explaining a vote or action was required, four respondents weren’t comfortable with voting. They indicated they’d want to sound out the group in advance, or talk to people before making a decision. They seemed to assume the group was about 10 - 15 people strong.

You’d almost want to get the email and then have a group meeting, everyone could vote independently but before that have a discussion.

- C, female, 36-45 year old, owns a laptop, tablet, bike, jewellery, TV

The perceived effort of participating in a group can be off putting

One participant was concerned that there might be a significant overhead in managing a group…

One participant was concerned that there might be a significant overhead in managing a group, and that it might involve meeting up or other commitments.

For me to receive a report, it sounds like you’re committing to something you don’t have time to do. This seems like a community or a steering group for your own bills, not something I’ve seen before. It feels like you might have to meet up, that’s a time commitment.

- C, female, 18-25 year old, owns a phone, TV

Things we’d like to try next

Experiment with other ways of communicating the concept. A short 1 minute animation explaining Protect Together might be more effective at communicating the various parts to it. For a good example, see

Explore different types of groups that people might associate with. We should experiment with whether individuals associate and trust other people based on age, salary, job, interests. We could also look at whether it makes a difference if the group already know each other in another context or come together specifically to take part in the scheme.

Explore different sizes of groups that people might join. We didn’t explicitly show the size of the Green Quarter group. A future design sprint might present respondents with different size groups to better understand how they feel about this and how they might communicate.

How we ran this experiment

Talking to experts

We talk to experts in the areas we’ve been looking to find out what they think the big challenges are. We then turn these challenges into positive “how might we?” questions which we try to answer as we carry out the experiment.

We’re lucky enough to work for a large organsiation with a lot of expertise in-house - we’re grateful to our colleagues who have given us their time and their knowledge. We’ve found that the Co-op name has opened doors for us, and external experts have also let us interview them.

From idea to software in 5 days

We tested this idea by building and testing a prototype in 5 working days. Our goal was not to test the effectiveness of the prototype, but to use it as a starting point for a conversation. Using the GV product design sprint framework, we were able to quickly test our assumptions and make something that felt real enough to try out on some potential users. The conversations we have with users as they interact with our prototype tell us if an idea is worth persuing or not.

One-to-one interviews with potential users

We only have a few days to organise interviews, which means we have to use the quickest way to find relevant people to talk to. We use social media and our professional and social networks to carry out highly targeted recruitment to find people who closely represent our target audience.

We observe and record our interviews, taking care to remove any details that would identify the participants. Everyone involved signs a participant agreement and we give vouchers to the people who take part as a thank-you for their time.

We use active listening techniques to get deeper insights into the reasons behind comments that people make. We have learnt a lot from experts in Co-op Digital such as Andrew Travers and James Boardwell that have helped us get the most from these interviews, thank you.

Further reading

This is a work in progress: Give feedback