Analytics and claims - Key takeaways:
• The majority of insureds who change carrier cite either premium increases of poor claims handling as the driver
• Greater claims automation could assist with customer retention and reduce the cost of doing business
• Automation is well advanced in the personal lines space, particularly high volume, low cost and low complexity claims
• Intelligent process automation – a combination of workflow automation and deep learning via AI or machine learning – uses rules and pattern recognition to triage claims
• Commercial lines carriers are making use of bots to speed up labour-intensive tasks, meaning straightforward claims can be paid within a day, rather than weeks
• Analytics can help to narrow down more complex claims and focus human resources on those claims that require more attention to detail
Bill Pieroni, chief executive of ACORD, is troubled. He is troubled by the fact that while the cost of claims is equivalent to roughly three quarters of all premium dollars, (re)insurance companies do not dedicate 75% of their digital spend on data and analytics to improve their back office capabilities.
"It's not even close," he says. "The majority of spend is on marketing and advertising, and the UK is the poster child for that."
Yet so much rests upon the claims experience - most crucially, the industry's reputation. Claims is the shop window by which the (re)insurance industry is judged - and as ongoing Covid-19 business interruption lawsuits (and the column inches dedicated to them) indicate, it is often judged harshly.
"Almost all insureds who leave their carriers cite one of two reasons: either their premium rose over 10%, or a claim was mishandled," says Pieroni. "Claims [handling] is the moment of truth for a lot of carriers."
It is widely accepted that the quicker a claim can be settled, the less expensive it is ultimately going to be. In one German pilot, a motor insurer benefitted from a reduction in claims processing costs of up to 50% using AI-based claims validation tools, according to the Swiss Re Institute.
Greater claims automation retains customers and ultimately reduces the cost of doing business, thinks Pieroni.
"If you don't settle a claim as quickly as possible you will only increase the severity," he says. "You always need to seek out early settlement opportunities and human beings don't do this very well. Technology tends to identify these settlements quickly, and it increases customer satisfaction and it decreases the severity of the claim."
"You're then left with the claims that have to be managed and adjusted and technology can help you determine who is the best colleague internally or third party to handle those claims."
Freeing up talent
This claims triage process operates best at (re)insurance organisations with a high level of digital maturity, according to Pieroni - companies ACORD describes as being “digital competitors”.
"Your scarcest resource is high skill, high will talent," he says. "I would make an argument that superior digital competitors have the best people."
"Yes, a lot of the process is digital - but those who remain are the ultimate in terms of skill and will, and understanding how to deploy those resources in a value-driven, selective way is critical. Digitisation and data analytics enable you as a carrier to allocate those scarce resources where they can have the biggest impact."
In the personal lines space, claims automation is already advanced. Beyond the hype (think Lemonade's highly-marketed three-second claim settlement for a lost mobile phone), super-fast claims payments have been around for a while. Many carriers have adopted the “one and done” approach to short-circuiting the process: one notification of loss, one inspection and loss estimate and the file is “done”.
Where claims are high volume, low cost and low complexity the whole process can be automated end-to-end, with little, if any, need for human intervention.
Notification apps, big data, robotic process automation (RPA) and predictive analytics are among the tools and insights used to extract and transmit data, validate coverage, adjust and assess the settlement amount and separate genuine losses from spurious notifications.
Officially, the process is called intelligent process automation - or IPA. This combination of workflow automation and deep learning via artificial intelligence (AI) or machine learning is the key to shifting and sorting through a plethora of claims notifications, using rules and spotting patterns that sort the claims into different buckets and directs them accordingly.
Claims triage in action
Such an approach is more of a challenge within specialty insurance where complexity becomes a hindrance. Here, there are still paper-based practices and the need for manual intervention.
Nevertheless, important strides are being made. Axa, for instance, is using AI bots to help staff with repetitive admin work within its commercial property, motor and liability claims team.
The aim is to free up staff to focus on more analytical work, with bots - Harry, Bert, Lenny and Como - saving over 18,000 working hours each year.
Meanwhile, Allianz Global Corporate & Specialty (AGCS) is using robotics to speed up labour-intensive tasks, such as reading emails or documents, collecting and processing claims data, as well as automating certain tasks, such as claims authorisation and payment.
According to Raymond Hogendoorn, property and engineering claims specialist at AGCS, it means more straightforward claims can be paid within a day, rather than the current industry average "of weeks".
Claims is a central pillar of Lloyd's modernisation efforts under the banner of the Future at Lloyd's initiative. During lockdown, the small claims auto settlement (SCAS) pilot went live for the automation of small value settlement transactions, with the first claim processed through the service on 12 May.
“These are challenging times so I'm really pleased to be able to say we've already delivered some improvements to our processes to make sure that money gets into the hands of our customers as quickly as possible," said Paul Brady, business owner for the Future at Lloyd's claims solution, in a video blog.
In most of these examples, the value offered by technology is in the ability to 'triage' claims so that genuine and straightforward losses can be settled quickly and efficiently, freeing up time to focus on claims that require the human touch.
"Essentially, automating the claims process is the Pareto principle at work - the law of the vital few: of many outcomes, around 80% of consequences come from 20% of causes," observes Bart Patrick, managing director for Europe at Duck Creek Technologies.
"In the claims world, 80% of claims, broadly speaking, should go through smoothly as they are valid. Then the vital few 20% need extra focus, either because the insured needs attention because they are in a difficult situation, or there is subrogation or a total loss."
"Analytics can help identify the vital few, focusing costly, experienced staff onto the right claim at the right time to help delight customers and drive down the cost of indemnity," he continues. "A good analytically-driven claims system can reduce the cost of handling claims dramatically. Settle 80% of your claims fast and efficiently, and focus your expertise on the 20% of challenging cases."
While some critical claims automation milestones have been achieved during lockdown, there have also, inevitably, been delays to the claims handling process.
This is putting (re)insurers on the back foot at a time when COVID-19 is impacting both asset and liability sides of the balance sheet.
An initial lull in claims notifications (in some classes) at the start of the pandemic is proving short-lived. The impact of Covid-19 on commercial liability lines of business is still unfolding, with the additional pressure of social inflation on classes such as cyber, medical malpractice, commercial motor and D&O.
Now, more than ever, there is an urgent need to make up for lost time, using technology where possible to expedite claims settlement.
