Effort-Linked Commission: The Distribution Reform Proposal Nobody Has Modelled

Commentary published on 5 September 2026 described two remuneration ideas in play at IRDAI. The first spreads commission across the life of the policy instead of front-loading it at the sale. The second ties an agent's pay to the effort put into sourcing and servicing the customer.

The first has been picked over at length, including in our own coverage of the consultation. It is arithmetic. You can model a trail shift on a spreadsheet: move a share of the lifetime commission from year one into later years, make each later slice conditional on the policy staying in force, and you can see exactly what happens to a book.

The second has barely been discussed, and it is the harder of the two. Effort is not a number that arrives from a policy administration system. It is a judgement about what a person did. Before an effort-linked rupee can be paid, somebody has to define what counts as effort, somebody has to produce evidence of it, somebody has to check the evidence, and somebody has to decide what happens when the check fails. Every one of those steps is a cost, and under the structures being discussed most of that cost lands on the advisor.

A commission formula pays on inputs it can read. To pay on effort, a regulator or an insurer has to pick a proxy for effort that a system can record. There are only a few candidates, and each one breaks in a predictable way.

1. Documented needs analysis at the point of sale. Records a form, not a conversation. It rewards the advisor who fills the form fastest, which is not the advisor who advised best.
2. Contact logs across the policy year. Counts touches. It rewards frequency of contact, so the advisor who calls a client four times about nothing scores above the advisor who solved the problem in one call.
3. Servicing tickets raised and closed. Counts problems handled. It quietly rewards books that generate problems, and it under-counts the advisor whose placement was clean enough that nothing broke.
4. Persistency of the book. The cleanest available proxy, because it is an outcome rather than an activity, but it is already the mechanism the trail proposal uses. If effort is measured by persistency, the second proposal collapses into the first.
5. Claim-stage assistance. Genuinely valuable and genuinely hard to fake, but it only fires on the small minority of policies that produce a claim, so it cannot carry a whole remuneration formula.

The list is short and the failure modes are consistent. Anything that counts activity can be manufactured. Anything that measures outcome duplicates the trail proposal. That is the modelling problem sitting under the phrase "effort-linked", and no published version of the idea has resolved it.

Whatever proxy is chosen, an insurer paying against it has to be able to defend the payment. That means a file per policy: what was recorded, when, by whom, and in a form an auditor can inspect after the fact.

The infrastructure for that is already being built. On 31 July 2026 IRDAI tightened intermediary rules, requiring greater sales traceability and additional disclosures from intermediaries. Traceability was framed as a conduct measure, and it is one. It is also the substrate an effort-linked payout would run on, because a rule that pays for effort needs a trail of the effort to pay against.

Read that as a change in what an advisor is expected to keep. Under a volume structure, the record that matters is the policy. Under an effort structure, the record that matters is everything around the policy, and the advisor who cannot produce it has not merely lost an argument, they have lost the payment.

The choice of proxy decides who gains and who loses, and the two ends of the market sit at opposite ends of the effort distribution.

A commercial advisor may hold a small number of accounts and spend hours per renewal on each one: reworking a schedule, chasing a valuation, arguing a clause, sitting through a survey, walking a claim through the adjuster. Effort per policy is high and volume is low.

A volume motor advisor may write hundreds of policies a year that take minutes each and are never touched again until renewal. Effort per policy is low and volume is high. That is not a criticism of the work. Retail motor is priced and structured to be transacted quickly, and the customer is served precisely because it takes minutes.

An effort measure that counts activity per policy transfers income toward the commercial book. An effort measure that counts total activity across the book transfers it toward volume. A measure built on persistency favours whichever line renews best, which in practice means neither of the two consistently. There is no neutral setting. Any formula that ties pay to effort is also a decision about which kind of advisor the market should have more of, and that decision deserves to be made explicitly rather than fall out of a proxy chosen for being easy to measure.

Trail commission is contingent on a fact anyone can check: the policy was in force on the date the slice fell due. Effort-linked commission is contingent on an assessment, and assessments are arguable.

That difference creates a dispute channel that does not exist today. An advisor who believes the servicing was done and the insurer's record says otherwise needs somewhere to take it. Somebody has to run that process, hold the evidence, decide the case, and pay or not pay at the end of it. On a book of any size, an advisor is not going to contest a small deduction on one policy, which means small deductions will not be contested at all.

The administrative weight scales badly. Life insurers sold about 2.70 crore new individual policies in 2024-25. A per-policy effort file across a base of that size is a significant cost, and it is a cost carried on both sides: the insurer builds and audits the system, and the advisor feeds it. Whatever mis-selling reduction the mechanism buys has to be weighed against that, and no published version of the proposal has attempted the comparison.

The complaint data does not point where the reform is pointing.

Complaints under unfair business practices rose to 26,667 from 23,335 a year earlier, an increase of about 14 percent. Total life complaints were flat over the same period, 1,20,726 falling to 1,20,429. Mis-selling therefore rose as a share of all life complaints, from roughly 19 percent to about 22 percent, in a complaint pool that did not grow. The problem is getting more concentrated, not more widespread.

Much of the loudest mis-selling has come from bank counters: policies presented as fixed deposits, or bundled into a loan at the moment the borrower is least able to refuse. Individual agents still account for roughly 51 percent of individual new-business premium and bancassurance about 33 percent, with direct, online and broker channels making up the rest. The last mile runs on roughly 26 lakh individual agents.

Commission reform lands hardest on that agent, because an individual agent's pay is visible, itemised and straightforward to cut. A bank's economics are not only commission, and the pressure that produces a policy sold as an FD sits in branch targets rather than in a commission schedule. Cutting or conditioning agent commission does not reach it. We have written separately on what bancassurance scrutiny opens up for brokers; the point here is narrower, which is that the instrument being reached for does not act on the channel the complaint data implicates.

Scale the numbers and the mismatch sharpens. Across roughly 26 lakh individual agents and about 2.70 crore new individual policies, the average is close to ten new policies per agent per year, and the true agent-channel average is lower still, since agents source only part of that total. The evidence regime under discussion would apply per policy across all of it, to address a complaint pool of 26,667.

On 10 August 2026, Smt. Girija Subramanian assumed duties as Member (Distribution) at IRDAI, appointed for a five-year term. She was previously Chairperson and Managing Director of The New India Assurance Company Ltd, which is to say her operating experience is general insurance rather than life.

That is worth noting without over-reading. Distribution reform is being discussed mostly in life terms: front-loaded commission on long-tenure savings products, persistency, policies sold as deposits. A general-insurance background brings a different default picture of what an agent does, one in which annual renewal is the normal rhythm, servicing is continuous, and the premium is re-earned every twelve months rather than banked once at inception.

If the reform arrives written in that idiom, effort-linked pay may look less like a life-industry mis-selling remedy and more like a servicing standard applied across lines, including general insurance books where front-loading was never the issue. Advisors who write both should watch for that, because a rule framed for one problem can end up scoped across both.

Nothing here justifies restructuring a practice around a proposal. It does justify a handful of moves that pay off whether or not effort-linked pay ever arrives.

• Start keeping your own servicing record. Date, client, what was asked, what you did, what it resolved. If an effort formula lands, you will be asked to evidence exactly this and the insurer's system will hold only its half. If it never lands, you have a client history that makes renewals easier.
• Know your own per-policy economics. How much time a motor renewal actually takes you against a commercial one, and what each earns. Any effort proxy will reprice that ratio, and you cannot see what it does to you if you do not know where you stand now.
• Do not assume your book is average. A book concentrated in servicing-heavy commercial lines and a book of volume retail face opposite outcomes under the same rule. Work out which one you have.
• Build the renewal income that does not depend on the formula. Persistency pays under an upfront structure, a trail structure and any effort structure that touches outcomes. The renewal book as an advisor annuity is the one asset none of these proposals penalise.
• Respond to the consultation when it publishes. The chairman said the commission cap's upper limit would be set after stakeholder feedback. The definition of effort is the part most worth commenting on, because it is the part least worked out, and the people who will have to evidence it are the ones best placed to say what evidence actually exists.