By Patrick O’Neill, Founder and President, Redhand Advisors
The days of error-prone, time-intensive claims processing are gone. Risk teams today can use artificial intelligence (AI) and machine learning (ML) models with their risk management information system (RMIS) to predict claim developments, make informed decisions, and reduce risk costs.
You might think you don’t need this technology if you use a third-party administrator (TPA) to manage claims. But, the truth is, every organization should own their risk analytics, even if another party manages their day-to-day claims as was discussed in a recent webinar hosted by Redhand Advisors.
Mark Tainton, Ventiv Technology’s global head of decision analytics, and Elvin J. Rosado, ARM, Asplundh Tree Expert’s risk analytics-risk management manager, joined me in delving into how AI and ML technology are improving claims administration for all parties involved.
The benefits AI and ML in claims handling
AI and ML features in a RMIS can help risk teams predict how claims will develop — including claims’ severity, ultimate incurred, resolution time, and litigation potential. With these predictions, risk teams can:
- Push low-severity claims through straight-through processing (STP). AI and ML models can identify low-cost claims, so risk teams can automatically handle them with STP and allocate resources to more complex claims.
- Act on claims with litigation potential. AI and ML models can predict which claims have a high likelihood of litigation, so risk teams can prioritize settling them to reduce costs.
- Help adjusters close claims, regardless of their experience. AI and ML models help adjusters of all levels make informed claim decisions. And with case severity predictions, risk teams can confidently assign claims that fit adjusters’ experience levels at first notice of loss.
These benefits help risk teams identify areas of claims that aren’t obvious and prioritize the most urgent claims. This can lead to hours of time and millions of dollars in actual savings by reducing total claims costs.
Keep in mind, this return on investment doesn’t happen overnight. Onboarding for a RMIS with AI/ML features can take anywhere from two to six months. From there, it often takes six to 12 months for organizations to see a reduction in claim costs and increase in efficiency as adoption accelerates and users better leverage the data.
Even with time, the ROI of this technology still isn’t a guarantee. Claims teams must work with their RMIS vendor and TPA to make sure your underlying data and tracking is sound. If organizations don’t have the high-quality data they need, AI and ML technology likely won’t produce the returns they want.
How Asplundh Tree Cutters lowered its TCOR with AI and ML
Asplundh Tree Cutters provides line clearance services all around the world and has over 38,000 employees.
Because 60% of the company’s staff works six feet or higher above the ground every day, predicting when regular tasks will turn into a claim is a benefit for the organization and their 12-person risk management team. The group handles 3,000 claims, worth $80M+, and 40,000 audits and assessments annually.
Managing such a volume of claims became difficult for Asplundh in early 2020. The company was struggling with year-over-year total cost of risk (TCOR) increases and labor shortages due to COVID-19. The risk team had a TPA, but there was a backlog of old claims that needed settling.
That’s when the company’s risk management team turned to Ventiv for help. Luckily, the organization’s RMIS platform — and its AI and ML-based software in particular — was what Asplundh needed.
Over three months, Asplundh and Ventiv worked together to tailor the RMIS to meet the company’s needs. The risk team took advantage of the RMIS’ sentiment predictive analysis features of Ventiv Predict to sift through claim notes, as well as the platform’s safety and TPA efficiency analysis features.
With the system’s advanced predictive analytics, Asplundh has seen a number of tangible benefits already, including:
- 10% reduction in TCOR (and an estimated 18% reduction next year)
- 17% improvement in reserve accuracy
- 8% improvement in closing ratios
Now, instead of waiting on their TPA to present data, Asplundh’s risk team oversees the claims process with their RMIS’ AI and ML features. They can bring concrete risk information to their administrator and redirect the TPA if needed.
While it’s more typical to let the TPA own the claims process, the best outcomes come when the business takes ownership and recognizes that you have the most power in closing claims and getting the best outcomes.
The future of AI and ML in claims processing
There are many opportunities to apply advanced analytics towards risk mitigation and claims management. With AI and ML models in their RMIS, risk teams already do, or will soon have access to:
- Future-proof risk management models to prevent risk and severity, tracking claim outcomes and identifying patterns to inform early claims decision-making.
- Geospatial simulations that provide insights on when, where, and why risk events might happen and how they can prevent danger based on previous claims.
- Claim perception analytics thatidentify high-priority claims through negative survey responses. This can include claims risk scoring and claims duration ratios.
- Litigation correlation features that identify high-priority claims with a likelihood of litigation.
To improve claims processes and mitigate risk, harness the AI and ML capabilities of your RMIS alongside your TPA. For further guidance on how your organization can improve claims outcomes with a RMIS that provides advanced analytics, schedule an inquiry call with Redhand Advisors.