Originally established as Farmers Mutual Insurance Company in Madison, Wisconsin, in 1927, American Family Insurance Group provided motor insurance to farmers. In 1963, it rebranded itself as American Family Mutual Insurance Company after broadening its product line throughout the years. According to the company’s financials, revenue increased from $14.4 billion in 2022 to $17.1 billion in 2023.
In an effort to enhance client relations and operational effectiveness, the American Family Insurance Group purchased Networked Insights in 2017 to strengthen its data analytics and artificial intelligence capabilities. Ryann Foelker, Strategy Director at American Family Insurance Group, recently discussed how to create compelling AI use cases in an Emerj podcast. According to Foelker, her organization is concentrating on addressing enduring issues for which there are no practical AI remedies.
In order to demonstrate how Family Insurance Group’s AI projects are actively assisting its strategic business objectives, this paper examines two compelling use cases:
Simplifying subrogation with the use of computer vision and machine learning: Subrogation claims can be automated with machine learning to find recovery opportunities and computer vision to assess damage, increasing speed and accuracy.
Using machine learning to optimize business rules and policy updates: reducing reliance on IT and increasing product delivery speed by centralizing decision rules and automating business logic extraction through machine learning.
Simplifying Subrogation Using Machine Learning And Computer Vision
According to the Washington State Office of Insurance Commissioner, subrogation is the legal procedure by which an insurer pursues compensation from the party causing the loss after paying a claim to its policyholder. The insurer is able to recoup expenses thanks to this approach.
However, insurance firms have difficulties with the conventional subrogation procedure. These difficulties, according to Tata Consultancy Services (TCS), include ineffective operational models and a significant reliance on manual procedures, which eventually result in lost chances for recovery and biases in decision-making. Insurers frequently require assistance with conflicting priorities that divert adjusters’ attention from subrogation duties, leading to mistakes and hold-ups.
The TCS Source also emphasizes how AI and machine learning assist in resolving these issues by automating repetitive processes, spotting subrogation opportunities at every stage of the claims process, and strategically assigning cases to the appropriate adjusters at the appropriate moment. It increases productivity, lowers recovery leakage, and eventually makes the entire claims procedure for insurers better.
In order to increase speed, efficiency, and accuracy, the insurance group used Tractable’s AI subrogation solution throughout all of American Family Insurance Claims Services (AFICS), according to a news statement issued by Tractable. American Family Insurance Group established an Innovation Lab in 2021 to investigate cutting-edge technology that might improve its claims procedures. The news release cited above claims that the partnership with Tractable, which focuses on incorporating AI into their business processes, was a result of the innovation lab project.
Tractable AI is a technology startup that uses deep learning and computer vision to create AI solutions for evaluating damage to automobiles and property. The news release also listed the partnership’s goals, which are as follows:
Enhanced Efficiency: Shortening processing times by automating the subrogation claim screening procedure.
Improved Accuracy: Making use of AI to consistently assess claims based on photographic proof.
Better Customer Experience: Giving claims adjusters more time to spend interacting with customers instead of handling paperwork.
According to the same news release, American Family Insurance Group used Tractable’s AI Subro technology to support its incoming subrogation activities at the time.
According to the company’s website, Tractable’s AI Subro solution automates the settlement of third-party claims and is used for the subrogation procedure. It is a component of the Tractable Auto Reviewer product reviewer.
The company’s promotional material claims that this AI technology operates as follows:
Data Input: When a claim is filed, along with any pertinent supporting evidence, including pictures of the damaged car or property, the procedure starts.
Image Analysis: To determine the amount of the damage, Tractable’s AI algorithms use computer vision techniques to examine the uploaded photographs. Based on past data, this analysis pinpoints certain areas of concern and calculates repair costs.
Automated Evaluation: The application automatically assesses the claim in accordance with predetermined criteria and policies established by the insurance provider. It computes possible payouts and assesses whether the claim satisfies the requirements for subrogation.
Recommendation Generation: The system creates a thorough report with suggested actions, anticipated expenses, and any supporting paperwork required for the subrogation procedure based on the study.
Claims Handling: In order to make prompt, well-informed choices, the claims team examines the insights and recommendations produced by AI. When managing subrogation claims, this simplified method improves overall efficiency and cuts down on processing time.
According to Tractable, their AI-powered technology evaluates the claim and files the report in under 15 seconds.
There are valuable inferences to be drawn from examining the American Family Insurance Group’s financials, even if it has not yet made any outcomes of this collaboration publicly available. Strong customer satisfaction and retention rates were reported on the company’s website, and at year’s end, there were 14 million policies in effect, a 3.8% increase over 2022. By expediting the processing of claims, artificial intelligence (AI) tools such as Tractable can improve customer service by increasing customer satisfaction and loyalty.
Additionally, premium growth and investment income drove American Family Insurance Group’s revenue increase from $14.4 billion in 2022 to $17.1 billion in 2023. Faster claims processing and maybe higher customer retention are two ways that AI’s efficiency in the claims process could support this revenue growth.
Using Machine Learning To Optimize Business Rules And Policy Updates
The cumulative impact of climate change is one of the many obstacles the insurance business faces when it comes to policy updates.
According to a Deloitte study co-authored by Karl Hersch, a guest on the “AI in Business” podcast, the global economic losses resulting from natural disasters in 2023 came to US$357 billion. Only 35% of these damages, however, were covered by insurance, leaving a 65% protection gap, or US$234 billion. In places like the Middle East, Africa, and Asia, this disparity is particularly noticeable.
The American Family Insurance Group implemented decision management software to enhance product offerings and ultimately provide better value, according to a press release from Sapiens Americas.
The company website claims that Sapiens offers a platform for decision automation that helps companies manage intricate business rules and convert policies into code. Its insurance decision platform automates and simplifies decision-making, enabling businesses to improve operational efficiency, manage business rules centrally, and expedite product delivery without heavily depending on IT resources.
According to the company website, Sapiens Decision uses user-friendly tools to centrally manage the logic in the policy administration system. It asserts that companies do not have to depend on IT for policy updates or tear down and rebuild existing policy management systems.
According to the Sapiens Decision Management Platform’s product literature, the software automates and maintains the business logic of insurers through a systematic procedure. Automated logic extraction, decision management, and decision execution come first.
Automated Logic Extraction: It transforms current business logic from legacy systems into decision models that are independent of technology by using machine learning. It is said that this process will remove the hazards and complications that come with using old code.
Decision Manager: It is accessible to business analysts without technical knowledge by enabling them to model, validate, and test these decision rules through a visual interface that requires no coding. It guarantees that the regulations governing activities are consistent and clear.
Decision Execution: It provides complete traceability of decisions made as well as various integration possibilities with current systems. With this strategy, businesses can minimize operational risks, streamline their decision-making procedures, and easily adjust to emerging technological advancements.
Using the digital interface accessible via their website, Sapiens offers a product tour:
Although neither Sapiens nor the American Family Insurance Group have released any particular findings from the aforementioned collaboration, Sapiens did release the following findings for Hiscox, one of its clients:
Customer satisfaction ratings increased from 83% in October 2019 to 93% in September 2020.
Increase in Net Promoter Score (NPS) from October 2019 to September 2020 (from 68.3 to 82.0).
Without involving IT, 90% of modifications may be created and implemented.
It is clear from American Family Insurance Group’s 2023 financial statements that the business experienced several difficulties, including a record $3.5 billion in catastrophic claims and a $1.7 billion net underwriting loss. Despite mentioning efforts to control costs, AFI’s combined ratio of 110.8% indicates that they are spending more on expenses and claims than they are making from premiums.
Monitoring these expenditures shows that, despite their efforts to reduce costs and enhance claims handling, the financial outcomes imply they have not yet been successful. Consequently, the available data does not clearly demonstrate that these changes have been made thus far, even though employing decision management software such as Sapiens could aid in process development.