Commercial P&C Insurers Guide to AI
How AI-Powered Data Extraction Technology is Being Used to
Solve 3 of the Biggest Challenges in Underwriting
Introduction
Data is the lifeblood of Commercial P&C insurance, and for underwriters, it plays a vital role in determining what actions they need to take next.
Unless your underwriting workflows are fully digitized, there’s a good chance your underwriters spend 40% of their time on repetitive activities such as data extraction and rekeying data.
This is where the power of AI technologies can help. SortSpoke, for example, combines powerful AI models + humans in the loop to extract data from your unstructured documents (with 100% accuracy) in a scalable and consistent way.
We’ve helped 25+ carriers, MGAs, and service providers implement SortSpoke into their underwriting workflows, and now we’d like to share these experiences with you.
- Boston Consulting Group
“The insurance companies that lead in the digital era won’t replace humans. Instead, insurance leaders will intertwine technology, people, and processes.”
Who is this guide for
If you’re a Commercial P&C insurer and parts of your underwriting cycle involves data extraction and rekeying data, this guide is for you. The goal of this guide is to help Commercial P&C carriers evaluate their current workflows, open dialogue between teams, or identify opportunities for improvement.
For example, a few months ago, we met with a P&C insurer and their two underwriting teams. After asking them some of the self-assessment questions from this guide, they quickly discovered each team processed new submissions differently!
To help you avoid a similar situation, we suggest answering the self-assessment questions at the conclusion of the following sections:
Challenge 1 - Triaging Submissions to Find Profitable New Business Opportunities
Commercial lines underwriting is not easy, especially finding profitable new business opportunities in your submissions. Based on conversations with our Commercial P&C customers, they can get anywhere from 200 - 500 new submissions daily.
Unfortunately, triaging new submissions for underwriters often requires manually opening, reviewing, and extracting information from PDFs, such as:
- Supplemental Applications
- Statements of Value
- Loss Runs
- Financial Statements
- ACORD Forms
- And more
These manual and repetitive processes often lead to missed opportunities to quote and bind profitable business because the first carrier that can provide a quote to brokers typically wins the business.
SortSpoke’s Triage AITM solves this challenge by automating the submission workflow by analyzing and classifying incoming emails and attachments. Triage AI will then score each submission based on your underwriting guidelines to prioritize submissions via:
- Identifying on-appetite submissions for priority underwriting
- Declining business that doesn’t meet underwriting guidelines
- Identifying incomplete submissions for broker follow-up
By augmenting your underwriting staff with Triage AI, underwriting teams can now prioritize submissions in a scalable and consistent way to win more new business and avoid wasting time on submissions that don’t meet your underwriting guidelines.
- Brandon Robinson, Director of Customer Success, SortSpoke
Case Study - How One P&C Insurer Turned Small Savings into Big Gains
Before using SortSpoke, one of our Commercial P&C customers spent 5 - 7 minutes pulling and rekeying data on every new business submission. After implementing SortSpoke, they reduced the submission intake workflow to less than 1 minute for a 5x productivity boost!
Over the course of a day, these small savings add up. For example, if each of your underwriters is saving 3.5 minutes per submission and they process 40 submissions a day, that’s 140 minutes saved each day by your underwriters!
This time savings will allow your underwriters to respond to submissions faster and provide more personalized service to your clients and brokers.
Challenge 1 - Self-Assessment Questions
Every good journey starts with a map. To see if implementing AI technology can solve your submission challenges, we typically ask the following questions to see if they can generate immediate ROI for your organization. Click here for the complete online version of the self-assessment.
Note: If you answer “Yes” to two or more questions, your organization would probably be a good fit for a solution like SortSpoke.
- Are you leaving 20% + submissions unreviewed?
- Is your submission volume expected to increase over the next two years?
- Do your underwriting teams follow the same processes for triaging submissions?
- Does your submission workflow require you to open and review each submission to find on-appetite submissions?
- Are you losing business to your competitors due to slower response times?
Challenge 2 - Increasing Underwriting Capacity
In commercial lines underwriting, these activities often include gathering information and extracting data from emails and PDF documents (e.g., applications, loss runs, etc.) to process new submissions and renewals.
SortSpoke can increase your underwriting capacity by combining the power of AI technology + your underwriters to turn highly varied and complex documents into actionable data (with 100% data accuracy).
In as little as a few hours, SortSpoke will automatically identify 80% - 90% of the data fields you need to extract. This time savings will free your underwriters to focus on what they do best, which is assess & price risk.
According to Accenture, underwriters spend approximately 40% of their time on non-core and repetitive activities.
Case Study - Specialty Insurer Generates a 5x Productivity Boost in Underwriting
One of the challenges of generating quotes for new business and renewals in P&C commercial lines is that underwriters spend a considerable amount of time manually reviewing and extracting information from financial statements.
For example, underwriters at a leading specialty insurer would spend anywhere from 3 to 8 minutes (sometimes longer) reviewing and rekeying information from financial statements to generate a quote.
In an effort to free up their underwriters from administrative tasks, this leading specialty insurer engaged in a proof of concept with SortSpoke. After implementing SortSpoke into its underwriting workflow, they reduced the time it took to review and pull data from financial statements from 3 - 8 minutes to 45 - 120 seconds per submission.
In addition, this company also reported it took SortSpoke’s AI only 25 - 50 documents to get up to speed. Furthermore, since moving into a production environment with SortSpoke, this specialty insurer reported SortSpoke’s models have only gotten better over time as they have added additional users and documents.
SortSpoke’s ability to accurately extract data from various types of documentation and adapt to new documentation on the fly (No templates or IT resources required) enabled this insurer to boost its underwriting productivity by 5X!
Challenge 2 - Self-Assessment Questions
Every good journey starts with a map. To see if implementing AI technology can solve your underwriting capacity challenges, we typically ask the following questions to see if they can generate immediate ROI for your organization. Click here for the complete online version of the self-assessment.
Note: If you answer “Yes” to two or more questions, your organization would probably be a good fit for a solution like SortSpoke.
- Do you have a standard approach for evaluating new submissions to determine who works on them?
- Would you like to capture more data in underwriting to enhance the accuracy of your in-house and third-party models?
- To increase your underwriting capacity are you planning on hiring more people?
- Do you have more submissions than capacity to review?
- Are you losing business or experiencing customer satisfaction issues due to slower response times in underwriting?
Training and Retaining Staff
Replacing institutional knowledge due to staff retirements is one of the biggest issues P&C insurers face. According to The Jacobson Group & Ward Group, 65% of P&C insurance companies added staff from July 2022 - July 2023 alone3.
This is why some forward-thinking insurers have implemented AI technology to eliminate non-core activities (e.g., extracting data from unstructured documents) in underwriting to boost staff retention and accelerate employee onboarding.
SortSpoke captures your institutional knowledge by using the feedback of your experienced underwriters to train our AI models on how to read and interpret complex documents. This will allow you to onboard new employees faster because they won’t have to learn how to read complex documents from scratch.
In addition, it will also help with employee retention by keeping your employees happy and more productive. In a study published by The Institutes, 60% of Millennials found analyzing risks and recommending ways for companies to overcome them to be appealing4
Case Study - TAI Empowers Staff to Spend More Time on Higher Value Activities
TAI is an industry leader in reviewing complex insurance treaties, allowing its customers to effortlessly digitize, store, search, and export key treaty data.
Insurance treaties are some of the most complex and difficult documents to read in the insurance industry. Each treaty alone is hundreds of pages long, and when you factor in 30 - 40 addendums, you’re looking at thousands of pages in total.
Before SortSpoke, this work required TAI experts to spend up to 8 hours reviewing each treaty and manually extracting, structuring, and comparing this information to historical data within a customer's admin system.
After implementing SortSpoke into their workflow, TAI reduced their treaty review process from 8 hours to 35 minutes, allowing TAI experts to focus on higher-value activities.
While this example above is from a different line of business, it’s one we often share with P&C insurers to demonstrate how SortSpoke can handle even the toughest unstructured documents.
Challenge 3 - Self-Assessment Questions
Every good journey starts with a map. To see if implementing AI technology can solve your challenges around training and retaining staff, we typically ask the following questions to see if they can generate immediate ROI for your organization. Click here for the complete online version of the self-assessment.
Note: If you answer “Yes” to two or more questions, your organization would probably be a good fit for a solution like SortSpoke.
- Do you have processes in place for capturing institutional knowledge in underwriting?
- Are you losing talent in underwriting faster than you can replace it?
- Does your onboarding process require training new hires on how to read and extract data from documents?
- Are your underwriters spending 3+ hours looking for data in documents?
- Are your documents a mix of unstructured text and tables?
The Cost of Conducting Business As Usual
Traditionally, Commercial P&C insurers followed the mantra of “If it ain’t broke, don’t fix it” regarding their legacy systems and processes. But, times have changed.
Loss of institutional knowledge, inflation, competition from InsurTech startups, higher workloads + fewer employees are just some of the challenges Commercial P&C insurers face.
However, insurers who augment their staff with AI technology can boost short-term performance and set themselves up for long-term success by:
- Consistently winning profitable new business opportunities by responding to broker submissions faster than their competition
- Adapting to marketplace changes on the fly by using AI technology to scale their operations without adding headcount
- Retaining underwriting staff by allowing them to focus on higher-value activities such as working with customers to solve risk mitigation issues
- PwC’s Annual Global CEO Survey for 2023
“37% of Global Insurance CEOs think their organization will no longer be economically viable in ten years’ time if they continue on their current course.”
Conclusion - 3 Steps Your Organization Can Take to Remain Competitive
By leveraging the strengths of humans and AI, insurers can increase staff productivity by up to ten times without increasing risk!
Based on our experiences of successfully implementing SortSpoke with 25+ insurers and service providers, we recommend taking the following steps:
- Benchmark the time your underwriters spend reviewing and extracting data from documents
- Identify and prioritize your use cases that score the highest on our self-assessment criteria
- Consult with SortSpoke to help us assess which use case will have the highest impact and plan a 3-week proof of concept (no obligation required)