This is a lightly edited version of my introduction to Marble's September 2021 investor update.

‍

There are two insurtech phrases you can deploy as a founder that will reliably signal innovation. You can say you are working on: "Plaid for insurance" or "embedded insurance."

Today, I'm going to pick on the latter.

‍

Capital and talent have poured into embedded insurance ventures. Brilliant people are committing to this massive market opportunity and big retailers and financial institutions are correctly tantalized by the revenue opportunity (I would know, I built one).

But is it so simple? Is embedded insurance the sure thing that will “reshape” insurance distribution?

I would argue: It's complicated.

A nearly instant, context-rich, embedded insurance experience makes a lot of sense on a whiteboard and when it comes to selling certain types of insurance. Unfortunately, it makes a lot less sense for distributing the insurance products that make the most money.

Embedded insurance distribution is, for a few reasons, really great for warranties and warranty look-alike insurance policies. For example, things like small cyber policies, device protection plans, travel insurance, and concert ticket protection are really well suited. 

• The sticker price is low enough for these products that they don't create a ton of decision friction within the larger sales cycle (“embedded insurance” must tautologically be, you know, placed within something else after all)
• The underwriting process is typically simple, or even done at the group level
• The risk is less exposed to catastrophe-type risks (these are typically policies that “travel” with the insured)

This last point, in particular, is important because it means that single carriers have more capacity to accept risk, which in turn enables vendors to sell more policies with just one embedded relationship — basically creating a virtuous cycle.

Where it gets more complicated though is, tragically, exactly where insurance companies make more money: The embedded distribution of the bigger ticket policies.

Embedded insurance is constrained by the same limits that have constrained traditional distribution tactics: Choice, trust, price, technology, and risk exposure. 

‍

Auto insurance, homeowners insurance, 10 year+ term life, other particularly niche insurance products — the “four figure class” ($1,000+ in annual premium) to try to coin a phrase — face deep, structural blockers when it comes to distribution via embedded channels.

• By and large, the largest incumbent carriers offer the best price on these four figure policies (in aggregate at least). However, and those of you who have spent time under the hood of an embedded insurance provider will immediately recognize the problem: these are also the hardest insurers to integrate with and the least likely to enter competitive or unpredictable marketplaces
• Compounding this problem is the fact that these policies are naturally most adjacent to expensive products that already have highly complex and stressful sales processes (e.g. buying a car or financing a house)
• This is a natural place to bring up the 1.2 million trained professionals in the room: Agents. Our research, and plenty of other’s, makes it clear that customers still do really like talking to a professional when committing to a big policy like homeowners. Finding an elegant operational solution to introduce informed agents into a larger and often times regulated sales funnel is...hard
• Identifying and building an embedded offer at the correct point of attachment is fickle, requires constant adjustment, and can expose a brand to broader risks outside of their core competency
• Finally, while choice is the priority for insurance shoppers, it is very, very hard to achieve in application, in large part due to regulation and the nature of risk sharing. A retailer or fintech who wants to deliver true choice and competitive pricing (i.e. 3+ bindable quotes) across multiple product lines and states will have to commit to a patchwork of embedded insurance integrations
  

To put the problem more simply, embedded insurance is constrained by the same limits that have constrained traditional distribution tactics: Choice, trust, price, technology, and risk exposure. 

Until there is a structural change in insurance, this limit will persist, and while it does, truly embedded insurance technology will stay in the two or three figure club. The four figure class, even those who can embed, will end up relying on traditional channels to scale.