Insurance is in the early stages of a dramatic transformation driven by data analytics, and unstructured data will play a key role in the industry’s analytics. (See our prior post.)
Unstructured data is the information that does not have a pre-defined data model or is not organized in a pre-defined manner (here). The sheer volume of data has increased dramatically, and today unstructured data makes up approximately 80% of enterprise data and is growing at the rate of 55% to 65% per year (here, here). The data-centric world is driving a change in how data is used, and companies that can effectively utilize unstructured data will have a competitive advantage.
We’re now starting to see the rise of massive, complex systems built around data – where the primary business value of the system comes from the analysis of data, rather than the software directly. From Andreesen Horowitz: Emerging Architectures for Modern Data Infrastructure
For insurance organizations, particularly distribution entities, unstructured data includes large amounts of emails, pdfs, photos, voice, and text information (here). In addition, insurers and agents/brokers have access to third party data including websites and social media (including images and videos from smartphones). Extracting value from unstructured data is a considerable challenge but advances in AI are creating opportunities to uncover actionable insights.
Acrisure’s acquisition of Tulco for approximately $400 million earlier in 2020 received minor press coverage (here, here, here) but could be an early warning for insurance distributors. Tulco is an expert in advanced artificial intelligence (AI) and had been working in partnership with Acrisure prior to the acquisition. Additional insights on the acquisition can be found in this Marshberry interview with Greg Williams, CEO of Acrisure (here).
Accenture provides three examples of AI insurance applications (here):
- Intelligent document analysis. Uses AI techniques including natural language processing, entity extraction, semantic understanding and machine learning to analyze content, extract meaning and reliably aid process automation and decision-making.
- Data storage optimization analytics. Lets IT departments, CIOs and content owners identify content eligible for deletion or migration to lower-cost storage, improving data quality and reducing costs.
- Sentiment analysis. Also known as opinion mining, uses AI to automate the process of identifying opinions about a specific subject from a piece of content. In its simplest form, it categorizes a sentiment as positive or negative.
Other applications may include fraud detection using social media, and personalized customer communication (here).
As AI applications become more advanced, more sophisticated methods of analyzing insurance data and developing unique insights will provide an edge to leading insurance organizations.
Innovate Insurance – Innovation & Entrepreneurship in Insurance
Also, see the related Specialty Insurance Blog – News & Commentary on Specialty Insurance – with an Emphasis on Professional Liability Insurance