Embedded insurance launched as a side project stays a side project. The organisational moves that turn an insurance feature into a revenue line someone actually owns.
Here is the pattern we see most often in platforms whose embedded insurance underperforms: the integration shipped fine, the offer converts respectably, and nobody in the company is responsible for the number. Insurance was scoped as a feature — a squad built it in a quarter, demoed it, and moved to the next roadmap item. Features get maintained. Revenue lines get grown. The difference between the two is not technology; it is organisational, and it is the subject of this piece.
The thesis is blunt: embedded insurance becomes a real P&L line at the moment one named person carries a revenue target for it, and not before. Everything else — team shape, rituals, finance integration — follows from that decision or substitutes badly for it.
Why the side-project trap is the default
Insurance revenue arrives quietly. It is a commission on someone else's product, it grows without marketing spend, and in the early months it is small next to the platform's core line. Every incentive in a product organisation points away from it: the squad that built the embed is measured on shipping, the growth team on the core funnel, finance on closing the books. The insurance number appears in a dashboard nobody's bonus depends on, and the programme coasts on its launch configuration — same placement, same copy, same panel — for a year.
Coasting is expensive precisely because embedded insurance responds unusually well to iteration. Attach rate moves with placement, framing, pre-fill quality and product mix; renewal revenue moves with journey design. A programme nobody tunes is leaving most of its ceiling unexplored, and the gap never shows up as a failure — only as a smaller number than it could have been.
One owner, one honest number
The fix does not start with hiring. It starts with assigning the number to someone senior enough to negotiate across product, engineering and finance — typically a growth or commercial lead at first — and making the number honest: platform commission net of refunds and clawbacks, not gross written premium, which is the insurer's scale metric and flatters everyone while committing no one.
Give that owner a real target and the authority the target implies: to claim engineering time for funnel experiments, to renegotiate panel composition with the infrastructure partner, to say no to a placement change that would juice a quarter and damage trust. An owner with a target but no roadmap influence is an accountability decoration.
What the team looks like at each stage
Full-time headcount comes later than most platforms think, and in a different order.
- At launch, insurance is a fraction of three existing people: the owner carrying the number, a product manager who treats the insurance journey as part of their surface, and an engineer familiar with the integration. The infrastructure layer is doing the heavy lifting — licensing, carrier connections, documents — which is the point of using one.
- When the number is material enough to be missed, the first dedicated hire is usually an insurance product owner: someone who runs the experiment backlog, owns the relationship with the infrastructure partner and carriers, and reads the quality metrics — cancellations, complaints, claims signals — not just conversion. Analytics support matters more than domain pedigree at this stage; the insurance expertise you lack can be borrowed from your partners, the accountability cannot.
- Programmes that reach multiple product lines and markets grow toward a small commercial pod — product, analyst, operations — and only then does hiring deep insurance experience from the industry usually pay for itself.
The finance seam is where programmes quietly fail
The least glamorous move matters most: get the insurance line into the company's actual financial rhythm. That means finance can reconcile the partner's commission statements against per-policy reporting; refunds and clawbacks are booked against the line rather than surfacing as mysterious debits; and the number appears in the monthly management pack next to every other revenue line, at the same altitude, judged by the same standards.
A revenue line exists when finance reconciles it monthly and someone explains its variance in the management meeting. Until then it is a feature with income.
The variance conversation is the real ritual. When the insurance owner stands in the same review as the core business and explains why attach dipped or renewal beat plan, insurance has organisationally arrived. When the line is skipped because it is small, it will stay small.
The honest caveats
Two limits to this argument. First, sequencing: none of this rescues a programme whose offer is wrong for its context — organisation multiplies a working product, it does not create one. Prove basic attach before you assign targets, or you will manufacture pressure without a lever. Second, proportionality: a platform doing modest volumes does not need a pod, a pack slide and a quarterly offsite; it needs one owner, one honest number, and a monthly hour of attention. The trap is not under-investment in headcount. It is that nobody owns the number at all.