Orbiseed
Insurance Vertical
Problem
The core users of Orbiseed’s insurance product are Risk Engineers. They are highly paid knowledge workers, typically with a Master's degree in mechanical or electrical engineering who are primarily responsible for helping their insurance carrier prevent losses and streamline underwriting processes. They are similar to home inspectors, however, they mainly inspect large commercial, industrial, or manufacturing facilities. They spend over 80% of their workweek reviewing commercial property risk reports of these large facilities. This entails reviewing 200 to 2,000-page reports looking for risk factors to populate an Excel property schedule that will eventually be reviewed by an underwriter to determine if a policy should be quoted for a prospective client’s properties. This is a time-consuming, tedious, and highly error-prone process that costs the US P&C industry US $12.5 billion annually.
My walkthrough of Orbiseed’s core IDP insurance solution.
Product & Solution
The product demo to the left showcases how Orbiseed automates the commercial property risk report review.
We use Natural Language Processing and Machine Learning to ingest, extract, standardize and summarize the property data in seconds
Risk engineers simply upload their PDF-based property risk reports, we detect over 50 risk factors, assign risk ratings, and provide a visual summary that can be exported to Excel or core insurance systems.
Evolution of the Product
Orbiseed’s insurance product was directly inspired by the work we were conducting in the construction vertical. Through our work in construction, I learned that on average a construction company spends .25% of its gross profits on innovation and new technology and we wanted to determine if a larger market with higher spending could benefit from our solution.
I conducted over 200 customer discovery interviews with insurance and risk professionals across the world trying to validate a core hypothesis: risk engineers and consultants with our technology could assess a prospective client’s overall risk profile and highlight specific risks faster and more accurately than traditional, manual methods. It became evident very quickly that we had discovered a painful problem that was begging to be solved. The body language of the people I was interviewing shifted when they were asked to describe their frustrations. Reviewing commercial property risk reports on a daily basis was described as, “nails on a chalkboard on repeat all day”, “I wish I didn’t have to go home every day with red, watery eyes”, and finally, “There has to be a better use of my master’s degree!” Additionally, through the customer discovery process, I met two retired insurance executives who loved what we were building so much that they joined our board of advisors.
Utilizing our new advisors, we built a hardcoded demo of our product and showcased it on over 25 sales calls with large insurance carriers and brokers across the world. Eight months later we closed our first insurance carrier out of Asia-Pacific and began helping them digitally transform their risk review processes. Today, Orbiseed is on its way to becoming the primary unstructured data aggregator and extractor for risk engineering teams and underwriters suffering from data overload. The product is successful and is used by risk professionals around the world.
P.S. Annually, the average insurance company spends anywhere from 3.5% to 4.5% of its gross written premiums (GWP) on innovation and new technology.
Achievements
Orbiseed speeds up risk review time by 6X from 2.5 hours manually to 10 minutes.
Orbiseed’s product is used by risk engineers in the US, UK, Mexico, Singapore, and Israel processing over 1,000 commercial risk reports per month.
We have processed over 100 GBs of data points from our active users.
We have processed over 100 different formats of risk reports and consistently extracted key risk factors with 95%+ accuracy.
Key Learnings
AI First B2B SaaS is hard! However, here are some ways I’ve learned to win in this environment.
Good demos don't have to be perfect for the product. They have to be perfect for the audience. Start with the discomfort - the issues, challenges and opportunities that your ideal clients face. The product can come later.
Love the problem, not the solution. There’s a reason over 200 customer discovery interviews were conducted for Orbiseed’s core product: I wanted my team to fall in love with the problem. Starting with a solution is like building a key without knowing what door it will open. You can try testing your key on lots of doors, or you can start with a door you want to open. When you fall in love with the problem versus your solution, you start building keys to doors that actually take you places.
Build a moat!
Integrate into existing tech.
Integrate into existing processes.
Product support is just as, if not more important, than the service you deliver.
Understand your customer’s buying journey and remove friction from that journey.
Nobody wants a sales call before knowing what your product does.
Supporting Documents
Insurance User Journey Map (Note: On desktop or laptop, to enter fullscreen please hover your mouse over the top right corner and click the expand button.)