Just under half of property/casualty insurers recently surveyed by WTW say they plan to incorporate AI into their analytics (49 percent), but most have yet to bring “traditional” analytics into areas other than pricing, WTW reports.

“While there is much benefit to be had in adoption of new technological solutions such as generative AI, most companies are still not close to realizing the full benefit of more ‘traditional’ AI approaches such as machine learning, and there are immediate and real gains to be had by applying these techniques throughout the organization,” says the WTW’s 2024 P/C Insurance Advanced Analytics Survey report titled “Advanced analytics: Bridging the gap between ambition and real-world success.”

The report summarizes the responses of 47 P/C insurers in the United States and Canada to questions about aspects of different insurance functions—underwriting, claims and reserving, expense management, marketing—where they are using advanced analytics today and where they plan to use analytics in two years time.

Comparing the results to an earlier survey in 2021, the report says, “Our 2024 reveals significant growing pains.”

For example, 36 percent of carriers surveyed in 2021 said they were using advanced analytics for claims triage in 2021, with another 42 percent saying they would do so within two years. But three years later, the percentage saying they are now using advanced analytics for claims triage is not even close to 78 percent (36 percent + 42 percent). Instead, 33 percent now say they are using advanced analytics for this purpose. Another 40 percent expect to do so by 2026.

“The aspiration is there, with many insurers saying they intend to bring analytics into areas as diverse as claims, marketing, distribution and other areas within the next two years. But looking at actual use of analytics over the last few years, the results demonstrate that progress has been significantly slower than hoped for and in some cases non-existent,” the report notes, referring to similar results for other aspects of claims handling (evaluation of future severity, litigation and fraud potential, for example), reserving (where only 17 percent use advanced analytics) and aspects of marketing (where usage is less that 30 percent for all activities WTW asked about).

Ambitions are significant in these areas. Forty-two percent of the insurers surveyed plan to use advanced analytics for case reserving within two years, for example, and one-third want to use it for customer segmentation aspects of the marketing functions.

In contrast, the report shows that insurers have achieved their ambitions in the area of risk premium modeling—essentially rating or pricing. But they had already achieved those ambitions two years ago. Eighty-three percent said they were using analytics for pricing in 2021 and 84 percent said the same this time around.

Other aspects of underwriting and risk selection came in lower, with the lowest percentages recorded for underwriting workflow management or submission triage. In 2021, one-third of carriers said they were using advanced analytics for underwriting workflow management, with the another 39 percent expecting to do so within two years. For the 2024 survey, the figure for current use came in at just 22 percent, not the 72 percent expected based on the 2021 survey (33 percent + 39 percent). But the aspiration is the same, with 49 percent of carriers surveyed now saying they expect to use advanced analytics for underwriting workflow management by 2026 (which would bring the total of those currently using and planning to use it to 71 percent).

Asked what has been holding them back previously, insurers selected IT bottlenecks (52 percent) and data warehouse constraints as the top two barriers. And those loom larger now than in 2021, when just 39 percent identified IT issues as a problem, and only 27 percent said they had the infrastructure/data issue.

The survey also asked insurers about top-line and bottom-line impacts of using advanced analytics (top-line significant), the impacts of InsurTechs on the market, the extent of skills or resource-oriented challenges they encounter in using analytics, the attention their organizations give to model bias, and about the cultural aspects of their organizations that support the use of analytics.

In terms of culture, while just about half (49 percent) said “leadership commitment” is regularly present at their companies, only 27 percent said the same about the presence of a “clear vision and strategy.”

Discussing resource-oriented challenges, the report highlights the interplay between the three most often identified by survey respondents: integration with business expertise (how effectively businesses leverage analytical output), lack of insurance knowledge, and shortage of advanced analytical talent.

“Data scientists bring a wealth of skills related to data manipulation and advanced analytical methods, but leveraging these capabilities will depend on a commitment to foster effective collaboration with actuaries and business stakeholders and to bridge the gap between insurance knowledge and a solid grasp of data analytics,” the report says.

A timely finding from the report relates to how carriers are thinking about the impacts of inflation and adjusting models to account for inflation. Here, 30 percent of carriers said they have adjusted current models for inflation and another 30 percent said they had no plans at all to adjust advanced analytics models for inflation.