Throwing down the disruption gauntlet in insurtech

By David Beach | 28 September 2017

Alberto Chierici, CPO of SPIXII, a fintech specialising in chatbot technology, spoke to bobsguide about the insurtech landscape, its pioneering UX solution, and the company's plans for major insurance disruption in 2018. 


How did you and SPIXII find its way into insurance?

I am one of the three co-founders of SPIXII. I take care of product management and data science. With a highly technical background with degrees in Physics and Statistics, I started working in insurance as an actuary, and it was at this point that I became fascinated and frustrated by the industry. It’s a very interesting financial product as it’s one of the most social too - the unlucky few are covered by the majority - so ideologicially it’s a fantastic product. But, of course, the public perception doesn’t see it in that light.

From an insurance consumer perspective, it’s frustrating to fill in incredibly long forms to get a quote or not understanding the terms and conditions, and that’s from someone who works in the industry. In fact, I read an article from the BBC that said you apparently need a postgraduate degree to understand the jargon.

I wanted to do something new and innovative because, in the six years I worked, I found it difficult to innovate within the established industry. Following my move from actuarial work to data science consultation I became passionate about using technology and data to make people’s lives easier, and that was the key driver that allowed me to make the leap and co-found SPIXII.

Do you have any pain points with insurance as a customer?

A few! For instance, it simply didn’t make economic sense for me to have contents insurance after processing the maths through financial models; I found that it was better to accept certain losses with certain frequencies rather than to pay the premium.

The massive variation in prices is another pain point. For instance, I helped our CTO insure his Vespa. The aggregator gave us a massive price range and it seemed illogical to pay top-end premiums worth more than the vehicle. We went directly to the website of the cheapest quote and it was £50 cheaper. We investigated the underwriters and found it was even cheaper still.  

New for old exchanges is another one as well. I struggled with this because the insurance industry is based on the principle of immunity which means that the insurer guarantees you’ll be in the same financial position before and after the accident. Many policies don’t work on a new for old principle. The system is broken.

It seems that the customer experience is very important to SPIXII then.

Customer experience is one of the most broken things in insurance. You always hear about bad customer experiences but rarely good ones; it would be great to give a voice to good experiences and put back some confidence in the system.

As the insurance contract isn’t a tangible product but an experience, it’s really important to deliver a remarkable customer experience. After studying design, I came to understand that data and design is a very powerful combination that more companies could take advantage of. If data informs design and experience-building, with the right KPIs and measurements, insurance companies can achieve great things.

As an example, our chatbot replaces the generic Q&A forms, which is nothing fancy, but we collected behavioural data through the user-bot interaction and that informed our machine learning algorithms to optimise the type of experience, such as how long to wait before asking a question or whether to use a formal greeting or a more casual one.

By optimising the journey to receiving a quote we achieved a 100% uplift, without optimising. From quote to buy, we achieved an uplift of 20%.

Why specifically did you decide upon using chatbot technology?

It came from customer interviews, where consistent trends emerged. We noticed that there’s no single solution that was readily available to help you on your insurance journey.

The second thing we learnt was that 99% of our interviewees ask their friends for advice to cut through the jargon and price variation to tell them what to do next. We built SPIXII to be that personal expert. Initially, we developed with the aim of being B2C but we had so much traction from the B2B sector we went down that avenue, particularly as we didn’t have the right experience to go straight into B2C.

What does the SPIXII roadmap look like?

Our roadmap generally follows the market options of technologies in the area of conversational agents. The trend we’re seeing at the moment is text, which will be relatively short. The next will be voice and touchless, as we’ve seen Apple introduce facial recognition and gestures.

In 2017 we’ve become the well-recognised solution for insurers, particularly for our focus on customer experience in the insurance context. We’ve also been praised by CEOs of the Fortune 500 as a ‘no brainer’ for insurance companies. Naturally, we want to build and consolidate on this position with a view to launching internationally.

In terms of our technological roadmap, we have a set of tools but they’re not unified under the same product. We have a white label implementation for chatbots on insurers’ webpages and these are customised. We do have standardised elements and we need to achieve more standardisation. There are three key things to do well when considering our white label implementation.

The first is very light integration to incorporate more easily with legacy systems; for instance, the chatbot only requires three lines of non-invasive code for implementation.

The second is dialogue management i.e. building a workflow of the user’s A-B journey. We call this the Conversation Experience Designer and essentially this is a canvas where anyone (particularly non-coders) can visually build a conversation journey and publish it as a bot; that’s under development for release next year.

The third is that you need to learn from the conversations. We use the data to group and segment customers, and we use this information to better drive traffic. It’s also a good approach for understanding customer segmentation as proxies for risk and allocate accordingly. Again, this is a part of our product on our roadmap for release next year.

What about the security aspect, particularly as we’ve seen ransomware and major data breaches in 2017?

Data protection is top priority when building technology which is why we use bank-level security in every system we employ. We’re also working with top lawyers to understand the required levels of privacy and to ensure we’re GDPR compliant, as well as consulting clients on how they can prepare.  

Above all, we work to our business values first and foremost. Instead of building the tech and applying rusty ethics after, we start from the ethics and build the tech around that. Essentially this means putting ourselves in the shoes of the consumer.

Do you have any secret developments at SPIXII HQ?

A few… but the only thing I can mention is that we’ve noticed there are some claims experiences that can be fully automated. Flight compensation is a good example of that. There are little niches in the market that are ripe for innovation and automation.

Which side of the challenger vs. collaborator does that put SPIXII?

Ultimately, SPIXII exists to get more people better protected. The only way to do this is for insurance companies to collaborate with insurtech startups, adopt an innovative mentality and trial innovative ideas, both internally and outside of the company. Insurance companies are now much more active in nurturing innovation through accelerators. There's still a lot to learn in the industry, but also there's a lot that simply can't wait to be learnt - that's SPIXII's space.

For now, we're here to help insurance companies bring the best possible experience to their customers. It means challenging the status quo yet having significant impact thanks to collaboration.