AI-enabled processes could improve all key financial KPIs, according to Cytora’s Juan de Castro. More so than ever, productivity and growth are well and truly under the microscope in insurance.
With employees all over the world now working remotely, important questions have emerged among leadership teams around how to maintain pre-pandemic levels of productivity – and how to win in an increasingly competitive market.
What’s more, with today’s hardening market, insurers need to make the most of the opportunities available to set themselves up for future success.
With all of this in mind, imagine being able to automate huge areas of manual or unnecessary work. By implementing AI and transforming processes today, you could gift underwriters the capacity to focus on the tasks that can drive real value and set insurers up for future success. There really is no better time.
Cytora recently spoke on an Insurtech Insights webinar, alongside speakers from AIG and Descartes Underwriting, about how to do just that.
According to Sebastien Piguet, co-founder and CUO at Descartes Underwriting, “the hardening cycle is clearly impacting our industry, and the gap between offer and demand is widening.”
In today’s hard market we’re seeing rates go up and capacity reducing. As a result, insurers are being flooded with a high number of out-of-appetite submissions. This is creating a number of challenges, specifically in the underwriting process.
Many insurers – both our clients and the wider industry – are asking how to capture the opportunities this market brings. And some of these opportunities will be in new lines of business that traditionally were out-of-appetite, but are now within appetite thanks to rate changes.
In that context, imagine if you could have real-time control over what your underwriters are working on. You’d need a consistent triage of submissions, which are supported by automated decisions about how submissions are filtered and prioritised. This allows your underwriters to spend time on the most attractive business, rather than following a first in first out process.
To make this a reality, insurers need to automatically represent the risk contained in a broker email submission. This will drive the automation of decision making, which in turn enables insurance leaders to tweak and change the appetite in real time.
This isn’t just about the here and now. According to AIG’s Sima Ruparelia (Chief Actuary, UK, EMEA, Global Specialty) we’ll inevitably start to hit a soft market. The key to succeeding is to keep our eyes on the prize – to understand what we want to write and what we don’t want to write when that soft market comes. Cleaning up data now and making sure processes are in place will mean that you have the best possible information to select the right risks when the market softens.
When considering the data sources to prioritise, there’s no one size fits all answer. However, Sima notes there are two main drivers for integrating this today.
“There is external data to enhance risk characteristics, which helps improve pricing models. But also, what you actually want in the workflow is a dashboard where that data comes in as you’re doing live pricing.”
When using granular risk characteristics, like flood data or adjacent property risk, those sources need to be reliable and accurate. But there’s a place for other data sources that are less revered for their accuracy, that can still help make a quick automated decision as to whether an underwriter should spend their time looking at a risk or not. In fact, there are hundreds of data sources for that purpose alone.
During the webinar, Sima noted that we’ve historically made processes inefficient for our underwriters. And as such, expense ratios are another big area driving transformation today.
Underwriters receive lots of submissions every day, but it’s still often unclear how they should be prioritised. Not only that, rekeying data means underwriters still have to input information three or four times.
According to Sima, “if you can make a clear underwriting workflow that reduces underwriter time spent on systems and processes, and increases time on key risks and interaction with brokers, that’s critical to success. It will make the whole process for the client better too.”
When considering use cases of AI in the underwriting workflow today, these have progressed in line with the technical advancements of recent years. We’ve already talked about augmentation; not just adding data, but delivering real insight to build risk profiles.
In addition, consider the automatic extraction of data from submissions. These still come in a number of formats, containing unstructured data. It’s been possible for AI to extract data fields from attachments in recent years, but now it’s even more mature.
Today, the AI can carry out a semantic analysis of the submission and ascertain the broader context. For example, if the broker was explaining a specific aspect of the risk, should it trigger an exception?
Secondly, and this is one of the most interesting use cases, how can you use AI to execute automated but sophisticated matching of the risk with your underwriting strategy? How can you make a decision on whether it’s in or out of appetite? Or predict the value of the risk?
AI can deliver these predictions, which aren’t just about conversion but also the expected loss ratio. In commercial insurance, it’s only possible to match a risk against others in your portfolio and understand expected loss when you have hundreds of data points to call upon.
The use cases above are no longer the science fiction of five or six years ago.
However, each insurer has a different set-up in terms of their operating model, tech stack and strategy. So for the technology to work, it needs to be modular and configurable.
What’s more, when it comes to being “bionic”, Sebastien believes it’s about adding skills into what an ideal team should be, rather than replacing anyone. He believes this part of the value chain will not – and should not – be automated.
Sima echoed this sentiment, noting that some companies have tried implementing AI and have felt uncertain about what it means for some roles in future. But it doesn’t mean replacing jobs. Rather, it allows people to spend time in areas where they add more value, like building relationships with brokers.
For relative newcomers to the market like Descartes Underwriting, these technologies make it possible to scale up and develop new products quickly. It also helps to “fail fast”, bringing the ability to test the market fit of a product and decide whether to make it a long term strategic focus.
But for incumbents and new players alike, all of this has the potential to hugely impact loss ratio. Taking admin away from underwriters and allowing them to maximise time spent on the right opportunities, means you end up with a much healthier book, more aligned to your underwriting strategy.
In fact, it’s fair to say that this approach can impact all key financial KPIs – expense ratio, loss ratio, and ultimately growth.
For more information on how you can use AI to support your underwriting teams, you can listen to Cytora’s blogcast here.
To watch the webinar in full on-demand, see here.
Cytora transforms underwriting for commercial insurance. Their platform helps insurers to underwrite more accurately, reduce frictional costs, and achieve productivity-led growth.
Chief Operating and Commercial Officer at Cytora
Share this article:
With world’s largest forum for insurance professionals, Insurtech Insights provides an invaluable platform for networking, new insights and exposure for thought leaders. We are always looking for new partners, content creators, and contributors to create value and deliver exceptional support to Insurtech Insights.
To share your knowledge, simply fill out the form.
