Why have expense ratios only gone up in the P&C insurance industry despite the influx of technologies like RPA, Computer Vision and AI?
The insurance industry processes billions of claims every year and has access to vast amounts of data. This wealth of information presents great potential for the industry to reduce costs and improve efficiency by adopting technologies like robotic process automation (RPA), computer vision (CV), and artificial intelligence (AI). However, despite the availability of these technologies, expense ratios in the US insurance industry have remained high, at around 25-30%.
Several factors contribute to this:
The cost of implementing these technologies can be prohibitive. RPA, CV, and AI are complex technologies that require significant investment in hardware, software, and training.
Integrating these technologies with existing systems can be difficult. Insurance companies often have complex legacy systems that can make it challenging to integrate new technologies. According to a Gartner market report on #hyperautomation, only 5% of the market opportunity for these technologies is attributed to them, while the other 95% is due to the need to integrate with other software and systems.
There is a shortage of skilled workers who can implement and operate these technologies. The demand for skilled workers in the insurance industry is outpacing the supply, making it difficult to find qualified people to implement and operate RPA and AI technologies. One solution is to retrain existing resources with these new skills instead of outsourcing.
These factors have slowed the adoption of RPA, CV, and AI technologies in the insurance industry, which has resulted in expense ratios that have not declined as much as they could have.
However, tech consulting, digital transformation, and system integrators are well positioned to help the industry address this challenge by working together with product/tools companies to create ecosystems that insurers will want to partner with.
What other reasons do you think could explain why this metric has remained relatively unchanged?
