Uncovering the Hidden Profit Potential: Startup Lessons from Healthy Averages

In 2025, U.S. property and casualty (P&C) insurers experienced their most successful underwriting year in decades, with the industry’s net underwriting gain nearly tripling from the previous year. This positive trend continued into 2026, as Verisk and the American Property Casualty Insurance Association reported a $31.7 billion USD net underwriting gain for the first half of the year, marking one of the strongest half-year results in recent history.

Despite these achievements, a report highlighted significant performance disparities based on business line and geographical location. Commercial auto, umbrella liability, and other casualty lines continued to face challenges, indicating that even in a record-breaking year, some insurers might be missing out on potential profits due to their assessment of their portfolios.

This issue mirrors a common challenge faced by startups: decision-making based on averages, which can obscure individual underperformers.

The Importance of Segmented Thinking in Insurance 

Insurance claims outcomes are binary – either a claim occurs or it doesn’t. As a result, insurers often assess the viability of their models by grouping policies into segments with shared risk or expected claims characteristics, such as personal auto, homeowners, or commercial lines.

However, this segmentation approach can lead insurers to unknowingly retain individual policies within supposedly profitable segments that are actually losing money. While a segment may appear profitable overall, there may be hidden negative-profit policies within it.

According to McKinsey, enhancing policy-level precision could result in a 30-50% improvement in underwriting results. This potential has attracted a wave of AI companies aiming to delve deeper into insurance analysis beyond the segment level.

For example, Boston-based Earnix offers tools to segment customers more accurately for health, life, and P&C insurers. Newer entrants like Soteris, a machine learning startup that recently secured over $8 million USD in seed funding, focus on identifying unprofitable policies individually.

The Significance of Identifying Blind Spots 

Soteris’ founder and CEO, Sunit Shah, highlighted the challenge in auto insurance specifically. He emphasized that insurers traditionally categorize policies into segments like ‘married couples’ or ‘single-driver policies,’ treating all policies within those groups uniformly.

Shah noted that insurers already possess sufficient data during the application process to enable individualized analysis, underscoring the importance of advanced tools in this regard.

Soteris identified that some carriers were holding onto profit-reducing policies, amounting to as much as 30% of their portfolio. Their predictive tools have been instrumental in assessing policy-level expected loss ratios, leading to significant insights for carriers.

The company’s latest tool goes further by pinpointing individual negative-profit policies and estimating their impact on profitability and EBITDA.

By dropping unprofitable policies while maintaining the rest of the segment, insurers can potentially enhance their bottom-line profit significantly, as demonstrated by early proofs of concept.

These findings underscore the importance of looking beyond surface-level averages to uncover hidden opportunities for improvement.

Avoiding Common Pitfalls in Startup Operations 

While startups may not manage millions of insurance policies, many operate on segment-level thinking, focusing on factors like pricing tiers, customer cohorts, or sales channels that may appear profitable overall.

However, beneath these aggregated metrics, there may be a subset of accounts that incur higher costs than revenue generated. Understanding profitability at the customer level is crucial for sustainable growth.

As AI facilitates cost-effective analysis, founders are encouraged to move away from managing by averages and prioritize profitability over indiscriminate growth, aligning with investor expectations.

Balancing Precision and Strategy 

While granular analysis can yield valuable insights, discontinuing individual policies poses challenges related to fairness, transparency, and communication with customers and regulators.

Startups face a similar dilemma, as cutting unprofitable customers can boost margins but may impact reputation and future growth if not executed thoughtfully. Strategic accounts that currently operate at a loss may hold long-term value beyond immediate profitability.

The key takeaway from the insurance industry is not to indiscriminately cut underperforming entities but to identify and understand the reasons behind their performance.

Featured image: Getty Images via Unsplash+