Innovation at InsurTech startups, and how to work with them


Innovation at InsurTech startups, and how to work with them

This is adapted from a speech I made at TINtech in London in June 2017. I talked about how we, as a startup, think about innovation and creating new things. I contrasted that approach with what I’ve seen at established insurance businesses, and also set out how insurance incumbents need to behave if they want to work successfully with startups.

How we think about innovation

The primary point I want to make, is that we see innovation as a science, and not an art.

In my previous career in consulting I would see numerous large businesses (in insurance and elsewhere) pursue an ethereal concept of innovation, as some sort of soft, cultural principle that would save them from all of their problems, with a plethora of working sessions and away days arranged to that end.

Actually, what I would argue, is that the most successful large tech businesses have developed a formula for innovation. It is tried and tested, and importantly, built into every business process in their organisations.

At its crux, this is about testing things in a lean manner, using 3 core stages, each with their own implications:

  • Build
  • Test
  • Iterate, or kill
First, build.

The best tech businesses start building early. They don’t spend six months writing a business case, or doing ten phases of customer research. They talk to a few customers to understand their problems, take their best guess at what product might fix those problems, and then build it.

Clearly, that has some implications; primarily that you can’t spend too much on building it. Overall, it means making lots of small bets, not a small number of big ones. At its heart, this is an acceptance that failure exists, and rather than ignoring it, you can build a process which controls for it.

For us as a business, this has been a core focus. Our approach has always been getting to market as soon as possible. Even if what we launch with isn’t our perfect vision of the future, it will enable huge amounts of learning, and help us to understand what works and what doesn’t.


Second, test.

 Henry Ford famously said, “If I had asked the people what they wanted, they would have said ‘faster horses’”

You may not have spent much time asking the people what they wanted before, but now is the time to see whether they like it. Sure, you can ask someone whether they like something, and what their favourite bit is, but the strongest signal that customers like something, or that it fulfils a significant need that they have, is if they are willing to reach into their wallet to pay for it; or at the very least pull it out of their pocket 5 times a day to use it.

A commonly deployed example of this principle is what many very early stage founders will do to test their business idea. They spin up a homepage-only website, which sets out what you want the new enterprise to be about, and drive Facebook and Google ads to it. This allows them to assess whether there is any demand for the product they are offering, or want to offer, as well as giving them some idea of what their cost of acquisition might be. You could literally run that experiment with an afternoon of work from one person, and £100 of ad spend, and generate a huge amount of learning and validation.


Third, iterate or kill

 It’s very unlikely that the first attempt will work perfectly, especially given you have been relatively light on research, planning and investment, but you can quickly iterate towards something else which might work better, and you will learn much more from experimenting, than by a decade of desk research.

If you’re a big company, and you’re running a dozen of these projects, you also need to be brave enough to kill the ones that aren’t working.

In our business, we’ve already been doing this. We have hundreds of beta testers signed up, who are testing our product, and giving us feedback on how to make it work better for them. More importantly, when we add new features, we watch how people use them; if people are working hard to avoid using them, it’s a good sign that they need to be improved or culled.

We certainly think that most successful startups are following these stages, and you also see them enshrined in the values of the big tech companies. Facebook famously asks employees to ‘move fast and break things’. Amazon demand that managers have ‘bias for action’. Google talks about the need to ‘ship and iterate’.


How to work with innovative startups in insurance

Now, what does that mean for a bigger company wanted to partner with a startup?
If you’re aware that a startup is using the build, test, iterate approach, you can change the way you work with them, and I have two main headings for that:

  • Start small
  • Move quickly

The first point is to look for something small you can start working with them on. That probably means giving them access to the most vanilla, and best established products or services. Before spending money on an all singing, all dancing product for a new startup, work together to make an innovative spin on something that you already have.

The second point is to move quickly. Raising cash for a startup is hard, doubly so in a regulated industry. That means that time is always the enemy of a startup. Your bias in working with them should always be towards what you can get them now, or in the coming weeks, not what you could build for them over a period of months. If you take too long, they’ll simply have to look elsewhere.

If you found this interesting, or are interested in InsurTech more generally, do hit click to tweet below to share more widely, and / or follow me on twitter @JimmySwims