Reading Embedded Insurance Forecasts Without Betting the Business Case on Them

Anyone selling or buying embedded insurance has heard the headline numbers. They appear in decks, funding memos and board packs. Most of them answer a question that no single carrier, MGA or platform has to answer: how big is the whole market in 2030?

The question that matters for a program is narrower. Does this partner, with this product, in this market, produce profit for the carrier, the MGA and the partner?

What the published forecasts say

BCG (June 2025) states that embedded insurance could account for more than $70 billion in gross written premiums by 2030, up from around $13 billion at the time of writing. It also reports that conversion rates for traditional insurers that have embraced the model are already higher than for separate insurance for the same products.

Deloitte (July 2023) cites forecasts for embedded property and casualty sales by 2030 that range from US$70 billion in the United States to US$700 billion globally. It adds its own scenario: if as much as 20% of the US personal auto market went the embedded route by 2030, at least US$50 billion in premium could move away from traditional distribution channels.

Read side by side, these figures do not conflict so much as measure different things. They differ in geography, line coverage and definition of "embedded." Their authors say as much by framing them as forecasts and scenarios. A bottom-up business case cannot borrow any of them.

Even historical numbers depend on definitions

Consider the MGA market, which carries a large share of embedded programs. Morningstar DBRS puts MGA-sourced direct premium written at $90.4 billion in 2024, about 9% of the US P&C market. Gallagher Re's 2025 report puts 2024 MGA premium at nearly $100 billion, including MGAs not reported under NAIC Note 19. Both can be right. They count different populations. If past figures vary with definitions, a forecast to 2030 deserves more caution still.

Build the case from the partner outward

A program-level case has inputs that only you and your partner can supply. We would not publish a generic number for any of them. Gather them from the partner and from your own data:

Eligible transactions. How many partner transactions per period are eligible for the product, in the states or countries where you can sell?
Offer rate and attach rate. How often is the offer shown, and how often accepted? Obtain a pilot figure instead of assuming a figure from another program.
Premium and loss expectation. Priced by the carrier's actuaries from the product's actual exposure.
Who gets paid what. Partner compensation, MGA fees, carrier retention, reinsurance terms. The licensing structure may limit how partners can be paid, so check this with counsel.
Cost to serve. Claims handling, customer service, compliance monitoring, producer oversight for point of sale, and the carrier's oversight under the MGA arrangement.
Ramp and sunk cost. Integration, licensing, filings and the time until the program pays back.

The first model should fit on one page and run in three scenarios. If the program only works at the optimistic attach rate, do not launch it at scale. Pilot it.

What the forecasts leave out

Deloitte's analysis flags that many nonfinancial partners are unfamiliar with the insurance transactions they will facilitate. That has direct cost implications. Partner training, licensing structure, compliance review of screen copy and ongoing monitoring are line items in the business case, and they scale with the number of partners, not with premium. A program with many small partners can carry more fixed oversight cost per dollar of premium than one with a few large ones.

On the technology side, BCG stresses that embedded insurers must work with partners on coverage, pricing and customer experience tailored to each business. Tailoring is not free. Every partner-specific product is another product to filter, govern and service.

Our point of view

Use market forecasts for strategy, not for pricing a partner deal.
Insist on a pilot with a defined stop rule before scaling any partner program.
Show the partner the full economics, including their share of risk and compliance duties. Partners who understand the structure stay longer.
Report the same four metrics to the carrier, MGA and partner every month: eligible transactions, offer rate, attach rate and loss experience. Disagreements about the program usually begin as disagreements about the numbers.

Clark Embedded Advisors builds partner economics and pilot designs for embedded programs. Book a call and bring your current GWP volume and technical stack details: embedinsurance.net/contact.

General information, not investment or legal advice. Forecasts are third-party estimates and may be revised.