This is just some announcement which people will want to pay attention to! Learn More
This is just some announcement which people will want to pay attention to! Learn More

How estimating the cost of worker’s compensation claims changed the way a leading insurance broker priced their risk

Predicting cost of worker’s compensation claims is becoming more challenging

Rising cost of care and lifestyle diseases leading to greater comorbidities is making predicting the probability and cost of a worker’s compensation claim even more critical to an insurer. This challenge gets further exacerbated due to almost 3-5% increase in complex claims. These cases tend to have delayed recoveries and much higher medical costs.

Further, catastrophic claims (defined as burn injuries, acquired head injuries, spinal cord injuries and multiple trauma injuries) account for less than 1% of all Worker’s Compensation claims but as much as 20% of Worker’s Compensation claims losses. One of the shortfalls of predicting claims losses is the ability to predict the catastrophic claims more accurately.

A global leader in insurance broking and risk management sought a solution

A global leader in insurance broking with presence across the world with over 5000 employees as part of a larger conglomerate was struggling with the unpredictable costs of catastrophic claims. Before reaching out to CrowdANALYTIX, the client had attempted to build predictive models in-house but couldn’t reach precision levels that could be high enough for them to confidently replace their existing underwriting or risk pricing methods.

CrowdANALYTIX tackled the problem

When hired to attempt a solution, CrowdANALYTIX drew on the skills of more than 300 data scientists to create accurate predictive models of potential compensation claims using historical claims data. Claims were classified in multiple ways to attempt more accurate predictions, including factors such as claim severity, possibility of lost time, claimant age, and detailed body part and cause codes. After evaluating over 500 different models from the 300 data scientists, CrowdANALYTIX picked the top 3 based on precision scores and the top predictive factors that made most intuitive business sense.

We then deployed the model on the cloud and enabled a dashboard for predicting the probability and cost of individual claims which could be further used to explore the cost trends of workers’ compensation claims.


The solution was integrated with existing pricing and underwriting tools and was used as an additional data point while pricing premiums for certain high-risk businesses.

Data used: Historical claims data

Share This Post

Share on facebook
Share on linkedin
Share on twitter
Share on email

More To Explore

Thanks to Gertrūda Valasevičiūtė via Unsplash

Why we need ensemble models in AI

When are machine learning models not enough? AI companies like CrowdANALYTIX give their customers the best AI models for their needs, and that means offering

Thanks to Possessed Photography via Unsplash

General AI: Is it achievable?

What is General AI? Although most people outside of the industry do not consider AI in these terms, there are actually two different types of