Building RiskPe — Why We Rejected the Commission Model in a Multi-Billion-Dollar Insurtech Market

A founder’s account of building a fee-based, zero-commission insurance advisory from Jaipur rather than a metro hub — what we gave up by refusing commission, what it bought us, and why the harder revenue model was the point.

The simplest way to build an insurance business in India is to take commission. The licences are established, the money arrives from insurers rather than from customers who have never paid for advice, and the entire market is configured to work that way. When we started RiskPe in 2025 we decided not to, and that decision has cost us real revenue in every month since.

This is the reasoning, including the parts that have not been comfortable. It is a founder’s account rather than a neutral analysis — the market thesis version, with the sector figures, is a separate piece linked at the end.

Three views of the same broken thing

RiskPe came out of three people who had each seen the industry from a different position, which is the only reason we agreed on the diagnosis quickly.

My father, Surendra Sharma, spent over a decade inside the industry, heading commercial lines distribution for one of India’s largest insurers, before leaving to start an independent agency in Jaipur in 2011. He built it on a rule that was unusual enough to be a differentiator: explain everything, sell nothing the client does not need. Watching that work, and watching what it cost him in commission he chose not to earn, was my education in the problem.

I saw the other side. After studying mechanical engineering I spent two years managing an insurer’s relationship with India’s largest online insurance marketplace, and took my licentiate certification from the Insurance Institute of India along the way. That job taught me exactly how the aggregator model works — what it is genuinely brilliant at, and what it structurally cannot tell a customer.

Saket Prag, my college friend and our CTO, came at it as a software engineer with a question neither of us had asked properly: why does an industry built entirely on modelling risk provide people with almost no tooling to understand their own?

The thing we could not design around

We spent our first serious weeks trying to build a better version of what already existed — cleaner comparison, better content, more honest recommendations, still commission-funded. It kept collapsing on the same problem in every model we drew.

If the money arrives only when a policy is sold, then a whole class of correct answers becomes unaffordable. "Your existing cover is fine, buy nothing today" is an answer no commission-funded business can be paid for. Neither is "consolidate these four policies into two", or "we will spend six weeks contesting your rejected claim from a sale two years ago that we already booked and spent".

We could design an honest commission-funded business. We could not design one that got paid for telling someone to buy nothing — and that was the answer our best conversations kept arriving at.

You can staff around this with integrity, and many good agents do. What you cannot do is make it survive scale, pressure and a bad quarter. Incentives do not have to corrupt anyone to shape outcomes — they only have to make one set of conversations profitable and another set free.

What refusing commission actually cost

It would be dishonest to present this as a clean win, so here is the bill.

• We ask customers to pay for something the entire market has trained them to receive at no visible cost. That is the hardest sentence in every sales conversation we have.
• The distribution cost inside a premium is invisible, so our fee reads as an additional expense rather than a substitution for one. We are competing against a price of zero that is not actually zero.
• We gave up the largest and most reliable revenue pool in the industry, in a market where it was available to us.
• Growth is slower. Trust does not respond to advertising spend the way a comparison funnel does, and we cannot buy our way to scale.
• There is no licensing moat in this category, because India has no fee-only fiduciary tier for insurance the way SEBI created one for investment advice in 2013.

Anyone evaluating this business should weigh those five points seriously. We do.

What it bought

What we got in exchange is a short list, but each item is something the commission model cannot replicate rather than merely do worse.

• We can tell a client to buy nothing, and we do — regularly. It is the single most trust-generating thing we say.
• Our claim recovery practice is a service with its own economics, not a favour extended to past buyers. We review a rejection against the policy wording, and if we conclude the claim is genuinely not payable, we return the review fee in full.
• Nobody at RiskPe earns more from one insurer than another, so there is no product our recommendations drift toward under pressure.
• Revenue recurs from the relationship rather than from a transaction, which means a year in which a client buys nothing is not a failed year.
• We can build technology for the client’s benefit rather than for conversion, because conversion is not what pays us.

That last one is why our AI policy analysis pipeline exists. It reads a policy document and surfaces the exclusions, sub-limits, waiting periods and coverage gaps that decide claims years later. In a commission model that tooling is a conversion aid and gets built to persuade. For us it is the product, and it gets built to inform — which turns out to produce very different software.

Why Jaipur

We are asked about this more than anything else, usually with an implied suggestion that we should move. We do not intend to.

The practical case is straightforward: lower burn, and a team that stays. Bangalore and Gurugram insurtech salaries and churn are a structural tax on a business whose growth curve is trust-led rather than capital-led. A slower, cheaper, more stable base is the correct shape for this model, not a compromise forced on it.

The substantive case matters more. India’s protection gap is roughly 70 to 80 per cent, and almost none of it sits in metro India. The underinsured customer is in Jaipur, Kota, Udaipur and several hundred cities like them — running a business with an unreviewed fire policy, or holding a family floater covering elderly parents that nobody has explained. Building for that customer from inside that market is not a disadvantage to be apologised for. Most of our category is built by people who have to research the customer we live next to.

It also keeps us honest in a way a metro office would not. When a claim goes wrong for someone in this city, they can come to our office in C-Scheme and say so to our faces. That is a useful constraint on a business that talks a lot about being on the client’s side.

What we are actually betting on

The bet is that Indian insurance follows Indian investing with a lag. SEBI separated advice from distribution in 2013 with the Registered Investment Adviser framework, and a generation of fee-only planners built real practices in a category that had not existed commercially before — while remaining chronically undersupplied, at roughly 967 registered advisers for more than 20 crore investors as of August 2025.

Insurance has no equivalent framework and no incumbent in the category. Every structural signal points the same way: penetration static at 3.7 per cent of GDP, a protection gap of 70 to 80 per cent, mis-selling named a significant concern in IRDAI’s 2024-25 annual report, and $4.44 billion of insurtech funding that has made buying easier without making advice independent.

We may be early. Willingness to pay for insurance advice is genuinely unproven at scale in India, and that is the risk we carry rather than one we have argued away. But the alternative was building another business that could not afford to tell people the truth, and there are already enough of those.

The market thesis behind this, with sector figures and the case against: why fee-based insurance advisory is India’s next insurtech opportunity. The category map: mapping India’s insurance distribution models.

More on the team and the company: about RiskPe, Krit Sharma, Surendra Sharma. What we charge: pricing. What we do when a claim is rejected: claim recovery.