In-House AI Development: A Key Competitive Edge for Commercial Insurers

Artificial intelligence is reshaping the software development ecosystem in commercial insurance, positioning it as a critical differentiator alongside pricing strategies. This transformation isn't just a trend; it's a shift that could define the industry for years to come. Companies that adapt quickly could find themselves at a substantial advantage, while those that lag may struggle to keep pace as the landscape evolves.

John Swigart, co-founder and CEO of Pie Insurance, argues that the future of AI investments in the insurance sector hinges on a decisive choice: building proprietary software in-house versus relying on third-party tools. This choice isn't merely about initial expenses; it's about long-term strategy and identity in a highly competitive market. Insurers that outsource software development might find themselves unable to distinguish their offerings, a situation that could lead to diminishing returns over time.

Cost-Effective Tool Development

Swigart suggests that with AI, firms can democratize the development of proprietary software solutions, which was once reserved for organizations boasting extensive engineering capabilities. This shift means smaller players can compete more effectively with established giants. The democratization of tech development could enable insurers to innovate faster and tailor offerings to specific customer needs. He emphasizes that to remain competitive, insurers must develop their technology rather than solely depending on external vendors, which can create vulnerabilities and limit adaptability.

The Buy vs. Build Dilemma

In small commercial insurance, discussions surrounding AI investment are increasingly wrestling with the tension between purchasing existing software and creating tailored solutions. On one hand, buying off-the-shelf tools can speed up implementation. Yet, Swigart argues for the value of developing bespoke underwriting models, particularly where specialization could enhance a carrier's portfolio or address niche markets that standard solutions can't cover. However, building such models demands access to significant data, which often necessitates licensing from external sources—a hurdle that may prove insurmountable for smaller firms lacking deep pockets.

What becomes clear in this debate is that the right choice often depends on individual circumstances, including existing resources and market positioning. For many newcomers, the allure of buying over building can be tempting; however, they must weigh the long-term implications of such a decision.

Data Challenges and Opportunities

Swigart points out that commercial lines lag behind sectors like personal auto insurance concerning the availability of third-party data. The vast differences across businesses—from operational risks to classifications—complicate effective data modeling efforts. The intricacies involved in accurately capturing and utilizing this data can confound even seasoned carriers, as regulatory requirements and diverse operational landscapes present significant challenges. Consequently, many small carriers lean heavily on external data and technology solutions instead of developing proprietary resources that will enable them to stand out.

Moreover, Swigart highlights a striking contradiction; many commercial carriers continue to rely heavily on manual processes, even when marketing themselves as tech-savvy companies. This disconnect raises questions about their actual level of digital transformation. For instance, companies like Next Insurance have achieved substantial scale with over 600,000 policyholders and roughly $548 million in revenue. However, this model has been built over a decade of accumulating data, and sustainability in growth will greatly depend on their ability to adapt and innovate using AI and new data sources.

The Value of AI in Commercial Lines

Despite the challenges that come with implementing AI, the potential advantages for commercial insurance are considerable. Increased productivity at both the individual and organizational levels could redefine operational efficiencies. Swigart advocates for a practical approach to AI integration—emphasizing experimentation and immediate application over lengthy pre-planning. By being agile and responsive, institutions can begin to see the value of AI in real-time, rather than waiting for exhaustive models to be built before testing them.

According to research by Deloitte, early implementations of agentic AI have delivered efficiency gains in underwriting of up to 36% and reductions in claims processing times nearing 40%. That’s not just fluff; it’s a measurable shift in operational capacity.

Key Data Insights for Underwriting

A significant area poised for growth is the acquisition of accurate firmographic data—comprehensive, real-time insights about insured businesses. This kind of data is not just important; it’s essential for correctly pricing and underwriting various commercial coverages. For instance, accurate classifications for workers' compensation are paramount; misidentifying the nature of a plumbing business can lead to inflated premiums. Small errors can have heavy consequences.

It's also essential to understand that the National Council on Compensation Insurance manages around 700 classification codes, and even a slight misclassification can drastically affect premiums. With this complexity in mind, Swigart advocates for AI-driven data streams that can proactively identify potential operational risks rather than waiting for issues to emerge during policy renewals or claims.

Strategic AI Approaches for Insurers

The emergence of AI has pushed small commercial insurers toward four strategic areas, where in-house software development proves pivotal: creating tailored software solutions, implementing customized underwriting models supplemented with licensed data, promoting broad adoption of operational AI through hands-on experimentation, and building accurate firmographic data pipelines to minimize classification errors in real-time.

Swigart warns that relying solely on third-party tools makes insurers indistinguishable from their peers, which in turn limits their competitive edge. For decision-makers in executive roles pondering future vendor expenditures, these strategic focus areas serve as a roadmap for potential investments. This framework not only helps identify which carriers are equipped to innovate independently but also reveals those who will merely follow market trends, lacking the differentiation necessary to attract discerning clients.

Future Implications and Significance

The implications of this AI evolution in commercial insurance are significant. As the industry matures, insurers must adapt or face the consequences of remaining stagnant. Those who invest wisely in data acquisition, proprietary technology, and AI models will likely emerge as leaders. The risk is real: as more companies adopt advanced technology, the competition will intensify, and being an average player will no longer suffice.

If you're working in this space, these insights should prompt you to reassess your current strategies. The challenges are formidable, but so are the opportunities. The question is, are you ready to take a leap into this transformational era of commercial insurance?