The Key Role AI Plays in Optimizing Underwriting Processes

Digital Transformation - October 27 2021

“The pandemic has created a lot of change, a lot of pressure to digitize, to get that self-service and that digital capability. It's also impacted workforce and staffing issues and putting pressure on underwriting,” stated Deb Smallwood, Senior Partner, Strategy Meets Action, during our recent webinar titled Identify the Best Submission to Quote in Minutes with AI.

The rules of engagement are really changing. Customers have different expectations and demands today. No longer do brokers and agents want to wait a week or two to receive a quote or to find out their application does not meet a carrier’s risk appetite. There is an expectation of a fast seamless insurance buying process. However, 50% of commercial lines business is still being submitted via email and PDFs, which is creating a lot of work and a lot of delays on the carrier side, and we are not capitalizing on all the data that is available. It's still very timely and costly in terms of delays, getting quotes out, or even acknowledgment of receipt or appetite.

The submission process is still plagued with paper submissions, e-fax submissions and more than 50% is static PDFs that still require someone to read through each document to manually extract key data points. Resulting in data gaps and long processing times, as we don’t always get all of the information needed to assess the risk. The entire underwriting process is flooded with layers of spreadsheets, worksheets, processes, rules, and data fields. Unstructured data is trapped across a myriad of documents requiring underwriters to spend copious amounts of time on non-value-added tasks like data rekeying across multiple systems and handoffs.

McKinsey estimates front-line underwriters are spending between 30% and 40% of their time on manual processes. In fact, based on a recent poll conducted with insurance executives we found that 44.4% want their underwriters to focus on high-value strategic judgment work and 38.9% want their underwriters to decrease the amount of time spent rekeying data. Others indicated that they want to eliminate data gaps and shorten the submission cycle to accelerate speed to quote.

Poll 1

Further to this, 39.3% indicated that their underwriting teams are spending 25 to 35% of their time on mundane, admin tasks, while 29.3% revealed their teams are spending upwards of 40 to 50% of their time on mundane admin tasks.

Poll 2

Identify the Best Submissions to Quote in Minutes with AI

Artificial Intelligence, Natural Language Processing, and Machine Learning solutions enable commercial insurers to re-imagine their underwriting processes by streamlining the submission intake process, applying intelligence to automatically and quickly identify the best submissions to quote by extracting the data from submissions making it possible to rate them faster and easier, to prevent underwriting leakage, and free up underwriting teams to focus on more strategic work.

By automating the submission intake process and eliminating the need for time-consuming data entry, cut and paste, and the time-consuming manual effort involved with extracting data from emails and multiple attachments, and populating downstream systems, underwriters can focus their time and effort on high value strategic work like deepening broker relationships.

Today, the submission intake process for mid-market and enterprise commercial lines insurance is messy, laborious, and steeped in repetitive manual effort. Reading lengthy submissions in various formats to arrive at the decision of whether or not the submission is in or out of appetite and the significant effort to master the data that's needed to rate is very onerous and that effort gates how much data is captured from the submission. Additionally, processing submissions is typically done in a first-in, first-out basis, based on the underwriter’s intuition or how they feel about the originating broker, or based on just the order in which it appeared. While underwriters are executing all of the manual effort to identify the best submissions that meet risk appetite to quote, the submissions are aging, and distribution partners are getting frustrated. It's not effective or systematic.

“AI really does have the potential to really transform the overall underwriting value chain from submissions to triage, risk appetite, exposure area and we know that not only does it help there, but it can also help extract and organize submission data,” stated Deb Smallwood, Senior Partner, Strategy Meets Action, during our recent webinar. “The great part about natural language processing is it helps us categorize and organize a lot of this unstructured and semi-structured data, unstructured data in front of the underwriter in forms and e-mails and PDFs that we're getting. And it will also help enable triage.”

Natural Language Processing provides insurers with the ability to identify, extract, standardize, contextually summarize, and ingest unstructured data from submission emails and attachments. Essentially, this capability can be adapted to a broad set of unstructured and semi-structured documents. Clearance and registration (C&R) elements like account name, broker name, quote by date, to more specific entities like the coverage needs, the claims history elements, can be extracted and structured and auto populated into downstream systems eliminating the need for manual data entry and rekeying.

“AI reduces the rote transcription effort during the intake process. It’s like a chainsaw in the submission intake world,” stated Colin Toal, Chief Technology Officer, Chisel AI. “Insurers can respond faster, cut costs, expand underwriting capacity, and refine their understanding of the data faster because you're capturing more of it, resulting in a better business experience, and winning more of the right business. The goal of implementing natural language processing is to improve the quote to bind ratio, by quoting the right business faster and more automatically because they can be systematic about how they access key data.”

Digitizing components of the underwriting process with applied AI can help to significantly improve the underwriter and broker experience by reducing the time it takes to quote. While at the same time freeing up underwriters and operations staff to focus on much higher value activities like understanding customer needs more deeply and advising customers on products and coverages that meet their needs more effectively, rather than spending countless hours on rekeying data across multiple systems.

Toal further explains that there is one benefit of AI that often gets overlooked, “as companies adopt machine learning in deeper and more specific use cases, having a complete view of the market, so seeing all of the submission data, extracting all of the submission data, not just the submissions that you want to quote and not just the portion of that submission you need to generate a quote, but the completeness of the data gives you an asset that you can use in the future to evaluate things like appetite decisions.”

Choosing the Right AI-Powered Submission Intake Solution

Artificial intelligence (AI) brings a powerful new set of tools to commercial lines underwriting, but choosing the right solution is key. If you’re thinking about adopting AI It’s important to understand the key attributes, key considerations and questions to ask AI providers when evaluating a viable AI solution. Here are seven key factors to consider:

  1. Purpose-built for commercial insurance
  2. Domain knowledge
  3. Pre-built workflows like Submission Intake and Submission Triage
  4. Trained insurance-specific AI models
  5. Supports insurance industry standards like ACORD and CSIO
  6. Learning based system with a continuous feedback loop
  7. API Integration

Change is Happening

According to a recent research study conducted by Strategy Meets Action, 80% of mid-market commercial lines executives expect big changes in five years and 94% of executives believe and expect significant change in 10 years. And 79% really believe that AI can help in underwriting risk analysis, and 41% really believe that it can help in the submission process, with 56% for triage and prioritization.

Watch the on-demand webinar to learn more about the digital forces that are putting pressure on commercial lines insurers to rethink their underwriting processes and hear firsthand from Colin Toal, Chief Technology Officer, Chisel AI how AI works to streamline the submission intake and triage process enabling insurers to accelerate speed to quote and write more business.

Digital Insurance On-demand webinar blog post graphic

Browse different topics

Recent Posts