Table of Contents

1. Built for This Moment

A big part of the Investor Day was spent walking through the same ideas they frequently cover in a bit more detail. They think AI and agent-based traffic is profoundly positive for their structural competitive positioning. It’s not like they have a data quantity advantage over the big boys. Instead, they’ve carefully built a system and foundation that’s better at processing and utilizing that data — across all departments. Their platform is AI-native to the core, as leadership spoke about the need for this technology more than a decade ago before it was on anyone’s radar. They’ve been building for this moment since their founding, and while their tech stack leadership is abstract, the outcomes they deliver with it are concrete:

This is how they settle claims in seconds with 70% of cases now automated. This is how they onboard new customers in minutes. This is how they routinely undercut everyone in the field with a strong gross loss ratio (GLR). This is how they explosively scale their business without growing headcount (which is expected to continue). This is how they maintain 3x lifetime value to customer acquisition cost (LTV/CAC) amid brisk growth. This is how they offer steep discounts for autonomous vehicle (AV) owners before others can. This is how their 5% loss adjustment expense (LAE) ratio is already lower than giant competitors’ despite their highly important economies-of-scale advantage. The lead keeps growing, with LAE at 9% two years ago and expected to keep falling from here.

Those are the effects. Its DNA is the cause. Lemonade’s DNA is a living, breathing web of perpetually growing data connections. It’s how they’re able to reliably improve product quality with every prediction and interaction showing them the precise way. No guesswork. For review, this is all done by the central Lemonade OS, which is used to power end-to-end optimizations across every facet of its business. With this operating system, building and optimizing can be done with tight performance testing and without complex coding or software engineering. They conducted a demo where Chief Claims Officer Sean Burgess built and shipped a new subagent on stage. He built an agent to customize plans for loss-of-use claims, which scored at 98% accuracy vs. 95% accuracy for humans and so was shipped. They added a live example where the agent was able to verify the losses, calculate appropriate payment and send it to the customer.

In contrast, the companies with sizable market share are built on 100-year-old tech stacks and disparate M&A stitched together like Frankenstein (as Lemonade leadership describes). As leadership says, “fragmented systems create fragmented context” that isn’t capable of using data to holistically automate and uplift workflows with AI. Alternative insurance platforms are clunky, slow, frustrating and not positioned for rapid innovation or agent-based traffic. Paperwork and people are still common pieces of basic insurance workflows elsewhere. That’s why “happy insurance customers” is such an oxymoron for everyone besides Lemonade and its uniquely lofty service scores.

Everyone has the same models. Not everyone is able to use these models or unleash data in ways that Lemonade can. While that is qualitative and abstract, the aforementioned cost and pricing leads when they shouldn’t yet have them alongside strong underwriting despite rapid growth are all concrete.

2. Lead Expands as Agents Proliferate

This foundation means Lemonade moves faster and has a far easier time retrofitting its operations for agent-based access. As that source of customer demand booms, Lemonade thinks its advantages will simply get bigger. That’s for 2 connected reasons. First, most insurers are still struggling to safely scale integrating this next-gen technology with their plan offering and onboarding systems. Lemonade’s platform has completed that task and also trailblazed connections between popular AI agents and every other useful part of its database and tech stack. Whether it’s asking follow-up questions, filing claims, issuing appeals or anything else to customize the experience, agents can tap into it with Lemonade far more seamlessly than with any large insurer. This means plans better tailored to a person’s interests.

“Everything in our system is MCP enabled meaning agents can operate all the building blocks of our system autonomously.” — Co-Founder/Co-CEO Daniel Schreiber

Second, this is where Lemonade’s aforementioned pricing advantages shine brightest. Agents don’t care about creative marketing campaigns supported by $5B+ annual budgets. They poked some fun at the various mascots of each giant competitor, but took the commentary a bit deeper and to a more critical place. The dominant insurance vendors in the USA have worked hard to essentially eviscerate price comparison websites by blocking access whenever possible. In the UK, where these sites are highly popular, smaller vendors like Lemonade can rank first on results pages more than anyone else (like they do) without spending much on ads. That’s not true here. Instead of competing on things customers actually care about (price & service), the big boys decided to make it a game of who can spend the most on awareness. This is why the insurance industry spends such a humongous amount on ads, effectively trying to drown out all other (often better) options so customers automatically think of them when deciding what plan to purchase. 

Finding the best option instead of picking the one we already know used to be a daunting task. It required navigating the web and searching for a needle in a haystack that might not even exist. Now… agents do all of that for us. They find the needle and can definitively tell us if there isn’t one. It’s even better than the backdrop in Great Britain, as the automated, machine-based price comparison can happen across every vendor rather than just the ones listed on a marketplace.

And? These agents don’t care about which company has the biggest marketing budget. They care about which vendor is better suited to work with them to craft the cheapest possible plan that customers will be happiest with. That’s not always Lemonade. But it’s Lemonade at a frequency level that’s far, far higher than its current market share. As agents take over, Lemonade has consistently over-indexed on output inclusion because of this. Furthermore, the platform makeup we’ve spoken so much about is well-suited to handle the explosion in activity and paperwork stemming from agent proliferation. It’s much easier for a customer to send detailed evidence of appeals when they can just ask Muse or Claude to do it for them. Companies have to be ready for this boom in activity, and Lemonade is.

3. More Examples of How Lemonade OS Provides Automated Help & Value

Leadership walked us through a series of demos showing how their end-to-end platform drives value across various parts of the business. On pricing, they worked through an example where an agent found a potential discount for customers, had it initially rejected, sought support from competing regulatory filings, found it, submitted it and had that change embedded throughout its pricing. They’ve constantly found small opportunities like this to improve pricing accuracy and uncover more customer value while still seeing their trailing 12-month gross loss ratio (TTM GLR) fall from 83% to 60% in a little over 2 years. Lemonade can gear these autonomous systems towards whatever key performance indicator (KPI) it wants, with gross profit dollar maximization the common choice.

They also went through product and growth demos where offerings were subtly tweaked on the fly to improve the customer experience and budgets were shifted in real time to bolster LTV/CAC. End-to-end automation touches everything Lemonade does, with 99% of code now written by AI (vs. 2% in 2024) and human verification built in when needed to ensure systems are acting appropriately. It’s this more efficient work that drives more efficient claims handling and overall customer-facing operations.

  • Overall Lemonade service automation rates are now 82% vs. 67% Y/Y and 51% 2 years ago.

  • Efficiency gains saved $400M of personnel expense since 2022.

4. A Lot More on Car Insurance

Upgrade below to read more about car insurance, new financial targets and my updated take on the investment.

logo

Subscribe to our premium content to read the rest.

Become a paying subscriber to get access to this post and a boatload of other subscriber-only content. Read the stock market newsletter read by Fortune 500 CEOs.

Upgrade

Already a subscriber? Sign in

Reply

Avatar

or to participate