It takes working in a variety of positions across the insurance ecosystem to gain a true understanding of the complexity of the industry. This has been the experience of Scott Field (pictured) who heads up international product strategy for Duck Creek Technologies and whose career to date has seen him work for insurers in finance, marketing and product design capacities as well as for technology vendors.
“It has been really interesting to see the industry from a lot of different people’s perspectives,” he said. “I think that’s really helped me when I’ve been working in machine learning aspects of my career – the ability to relate to an underwriter, or a claims person, or a CSR because I spent so much time interacting with them before machine learning threatened to take over the world. Now that it is threatening to take over the world, I don’t have the hubris of someone who has only done that data science piece. I understand what it’s like to have a dissatisfied customer on the other end of the phone.”
Given that the COVID crisis has whetted the already sharp corners of conversations around technology, innovation and the digital-first world of today, many insurance companies are having to come to terms with the pain points impeding their steps in that direction. For Field, who has seen the internal systems that govern these companies from every angle, the challenge ahead is for them to sort necessary complexity from unnecessary complexity.
“That’s the real challenge,” he said. “If you want to innovate, you need to be on a modern technology platform, because you need to be able to recruit people, to maintain that system and to not spend all your time regression testing a whole bunch of integrations with bespoke legacy systems downstream. If you don’t have a modern, simplified technology platform, you’ll spend all your time maintaining it. But when getting to those new platforms, one of the biggest mistakes people make is porting over all of the complexity of their existing business.”
Simply taking existing business logic and replicating it into a new platform – however modern and innovative that platform may be – will not make your business as efficient as you might hope. Then on the other side of the equation, there is necessary complexity, Field said, which is just as important to identify correctly.
He recalled a time early in his career when he was acting as a product manager for a firm going through a transformative programme that saw a ‘payroll deduct’ feature cut from the transition to the new system as it was only used by one affinity group in the entire company. The impact, however, was calamitous and saw that affinity’s group’s retention rate drop from 99% to 40%.
“And worse than that, a lot of customers lapsed, and they didn’t even know it because they didn’t know to look for a notification to pay their bill,” he said. “It would never have occurred to them that their insurance company would cancel them. These were people who’d been with their insurer for 15/20 years.”
It’s often difficult to identify the things that are important to take forward and to pinpoint those complexities which are essential and make your business unique. Field added that it’s a process made even more difficult by needing to accurately identify the stakeholders across an organisation who will be able to make those decisions correctly. Who can be trusted to make those calls, who has got the right opinions?
“It’s about finding the right people and through them get a [clear view] of what you really need, what you don’t need,” he said. “The minimum viable product approach is one I’m sceptical on when you’re doing a core transformation and moving to a new technology platform to innovate. Because once you mess something up - and in insurance, it can be really quick to mess up and can take years to fix – the customers are less forgiving when they’re no longer insured or not insured for what they thought they were than in other more transactional-based industries.”
The key, therefore, to successfully embarking on a transformational project is to formulate the right partnership and to work with the right providers who are well-versed in separating essential from non-essential complexity. Looking at what underpins those partnerships, Field noted that unfortunately there are a lot of technology companies that are promising a lot of things that they cannot deliver.
That’s particularly true in the AI space, he said. There’s a lot of talk in the market about firms using their own data to create a customer experience that guides consumers through the process of purchasing insurance. But to do that correctly, you have to consciously set up your systems in a way to capture that data and design a system that doesn’t just output that data but also asks users the right questions. To do so would likely take a couple of years to establish the right data sets but too much of the shopping behaviour he has seen around the industry is still treating AI like a magic trick, which it most assuredly is not.
“I’m sceptical, but it’s an earned scepticism, that one insurance companies’ data will have much bearing on another insurance companies’ data and people,” he said. “The executives already want the pre-trained models because they’re looking to cut time out of the machine learning and the AI process. I think that that’s the biggest mistake that we run into.
“And you have these companies that aren’t exceptionally well funded, that are trying to grow their customer count on the insurer tech side as fast as possible. Because companies aren’t really evaluated on their revenue today, they’re evaluated on their growth trajectory. That’s what fuels the multiple and so they’re just trying to grow as fast as possible in an industry that doesn’t move that fast. So, I think the economic incentives aren’t always aligned when you have insurtechs versus the insurers themselves.”
When it comes to selecting the right technology partners, he said, it’s about finding the sweet spot of companies that have enough scale that you don’t have to worry about them disappearing, but they don’t have so much scale that they’ve lost sight of what it means to innovate, that have become holding companies for other, older technologies. There’s not too many of those in the industry right now but there are a lot of people who will build systems for you that are exactly what you have requested. But that’s not necessarily what companies need in all cases, he noted, and the trick is knowing the difference.
