2.7 Lakh Festive Gig Hires Against One Insurance Declaration Set Months Ago

Festive-season hiring in 2026 is expected to add up to 2.7 lakh gig and temporary jobs, according to staffing-industry estimates reported by The Economic Times and The Daily Jagran on 17 August 2026. Moneycontrol reported on 14 August that festive gig hiring would rise by as much as 20 percent, with quick commerce the dominant source of demand, and followed on 19 August with quick-commerce firms bracing for a worker crunch as that demand outran supply. Bhaskar English reported on 23 August that the boom had spread past the metros, and the staffing estimates put roughly 45 percent of the demand in tier-2 and tier-3 cities.

For a platform's insurance programme, that is a headcount problem before it is anything else. Group personal accident (GPA) and employees' compensation covers for delivery fleets are almost always written on a declared number of lives at inception, with premium charged on that number and an adjustment clause at expiry. The declaration for a policy incepting in April or July was built on a base-season roster. A cohort that swells 20 percent for six weeks in September and October sits outside the assumption the premium was calculated on.

The exposure is not merely numerical. Redseer's estimate, reported by People Matters on 11 August 2026, puts India's gig workforce on a path to 21 million by 2030, and its 18 August finding that gig work delivers roughly 2.5x higher net earnings than the alternatives explains why the inflow keeps arriving. Much of that inflow is first-time riders, working unfamiliar routes, in cities where the platform has thin operational history. That is the cohort an underwriter needs to price. The declaration in force describes a base-season roster.

A declaration-basis group policy has three moving parts, and the peak-season problem sits between them.

1. The estimated number of lives at inception, on which deposit premium is charged.
2. The basis of declaration, which is either an opening and closing average, a monthly or quarterly average of lives on cover, or a peak count.
3. The adjustment clause, which trues up premium at expiry against the declared basis, usually subject to a minimum retained premium of 50 to 75 percent of the deposit.

Where the basis is an opening and closing average, a six-week spike that starts after inception and ends before expiry can pass through the arithmetic almost invisibly. The insurer collects a small adjustment, and the platform pays for the peak at a fraction of its true exposure. That sounds like a buyer's win until a claim arrives from a rider who joined in week two of the surge.

The problem is not that the additional premium was underpaid. It is that a material change in the risk was never communicated. Indian group policy wording carries a change-in-risk condition requiring the insured to notify any material alteration during the currency of the policy. A 20 percent increase in insured lives, concentrated in a cohort with no operating history, is material on any reading. An insurer told about it after a fatality is entitled to ask why it was not told before.

Peak-week claims rarely fail on a coverage exclusion. They fail on identity. The two recurring failures are these.

The named insured is wrong. Platform groups run multiple entities: a parent, an operating company that holds the app, and often one or more logistics subsidiaries that hold the rider contracts. The GPA is issued in the name of one of them. The rider who died was engaged by another. Nothing in the wording covers a person engaged by an entity that is not an insured, and the fix at claim stage requires the insurer's consent, which is expensive to obtain after a loss.

The category is undeclared. Insured-person clauses in fleet GPA are frequently drafted as "all permanent delivery partners of the insured", or "all riders on roll as at the date of inception". Either formulation excludes a festive hire recruited in September. A clause that reads "all delivery partners engaged by the insured, whether directly or through a manpower supplier, whether permanent, seasonal or temporary, during the policy period" covers the cohort without needing an endorsement per intake.

The declaration mechanics and the insured clause must agree with each other. A monthly-average declaration basis attached to a fixed roster of named lives at inception is internally inconsistent, and the inconsistency is resolved against the insured when the claim is examined.

Most festive-season riders arrive through a staffing vendor rather than the platform's own onboarding. The insurance question is whether the platform, as principal, carries statutory liability for them.

The Employees' Compensation Act, 1923 fixes liability on the employer for personal injury caused by accident arising out of and in the course of employment. Section 12 extends that liability to a principal who, in the course of its trade or business, contracts with a contractor for the execution of work: the principal is liable to pay compensation to the contractor's employees as if it had employed them directly, with a right of indemnity against the contractor. Section 12 is why a platform cannot treat the vendor relationship as a clean transfer of statutory risk.

Whether a specific festive rider is an "employee" for Act purposes turns on the facts of engagement rather than on the label in the contract. Courts examine control over the manner of work, allocation of tasks, the presence of penalties and deactivation rights, and economic dependence. Platform engagement models vary enough that the answer is not uniform, and a platform structuring cover should assume the exposure exists rather than assume it does not.

The practical consequence for the programme is that workers' compensation cover should be written to respond to claims where the platform is joined as principal employer, not only where it is the direct employer. Two drafting points matter:

• The definition of employee should include persons engaged through contractors and manpower suppliers for whom the insured may be liable under Section 12.
• The policy should respond to legal costs of defending a claim in which the platform is impleaded alongside the vendor, which is the usual procedural shape of these matters.

Then check the vendor's own cover. A vendor certificate issued in June and expiring in September leaves the surge weeks bare, and a lapsed vendor policy converts a contractual indemnity into an unsecured claim against a staffing company with limited balance sheet. Ask for the certificate, the expiry date, and confirmation that the declared headcount includes the festive intake.

The Code on Social Security, 2020 creates the national framework for gig and platform worker benefits, funded by aggregator contributions set at 1 to 2 percent of turnover subject to a cap of 5 percent of the amounts paid by the aggregator to gig and platform workers. State measures, including the Rajasthan Act of 2023 and the Karnataka Act notified in 2025, operate on a similar welfare-fund model with accident insurance among the benefits.

Three things follow, and platforms routinely get them wrong.

The contribution funds a scheme, not your policy. A welfare-fund benefit is a statutory entitlement paid from a pooled fund on terms the state sets. It does not indemnify the platform against a compensation award under the Employees' Compensation Act, and it does not discharge liability under Section 12. These are separate obligations sitting in separate places on the risk register.

The benefit quantum is not underwritten to your exposure. Fund benefit levels are set by policy, not by reference to what a rider's dependants would recover in a compensation proceeding or a civil suit. The gap between the two is the platform's retained exposure and belongs on the sum insured decision.

The registration base and the insured base diverge during a surge. Welfare-fund registration is worker-initiated and lags actual engagement. A rider who started on 12 September and died on 24 September may be unregistered, which affects the statutory benefit and says nothing at all about whether your GPA responds.

The right structure treats the contribution as a compliance line item and the GPA and employees' compensation programme as the loss-financing layer. Neither substitutes for the other, and a benefits stack presented to riders should describe them separately so that expectations at claim stage match the documents.

The single most common cause of delay on a peak-week fatality claim is the platform's inability to prove, from a system of record, that the deceased was engaged on cover on the date of the accident. A surveyor or claims manager asks a narrow question: was this person an insured person at the time of loss, and what was the wage or benefit basis applying to them?

Build the answer before the season, not after a death.

• Freeze a daily active-rider extract from the platform's operational system, retained for the policy period, showing rider ID, city, engaging entity, vendor name where applicable, and first and last active date.
• Reconcile that extract monthly to the number declared to the insurer, and keep the reconciliation. This is what converts a declaration from an assertion into evidence.
• Record the engagement document for each vendor-supplied rider: the vendor's letter of engagement, the rider's KYC, and the date of first assignment.
• Capture wage or earnings basis on the same record, because employees' compensation quantum is calculated on monthly wages and a rider's festive-week earnings differ materially from a base-season month.

Run one dry-run claim file in the last week of August. Pick a real rider record at random and walk it through the documents an insurer would ask for. Whatever you cannot produce in that exercise is what will hold up a real claim in October.

Work this list before the first surge week, in this order.

1. Confirm the declaration basis in the current policy schedule. If it is an opening and closing average, write to the insurer with the expected peak headcount and ask for a monthly-average endorsement for the festive months.
2. Notify the expected uplift in writing. A dated email to the broker recording the anticipated increase discharges the change-in-risk condition and is the cheapest protection in this list.
3. Fix the insured clause. Name every engaging entity. Replace roster language with a category description that captures seasonal, temporary and vendor-supplied riders engaged at any point during the policy period.
4. Check the sum insured against the wage base. Employees' compensation quantum moves with monthly wages, and festive earnings run above base-season levels for the same rider.
5. Collect vendor certificates with expiry dates. Reject any that expire before the end of November, and require the vendor to declare its festive headcount to its own insurer.
6. Amend the vendor contract to require cover on a declaration basis for the full engagement period, a certificate before the first deployment, notice of cancellation or non-renewal to the platform, and an indemnity that survives the assignment.
7. Confirm territorial and city scope. With roughly 45 percent of demand in tier-2 and tier-3 cities, verify that the policy has no city schedule or location limitation carried over from an earlier programme.
8. Set the endorsement route for mid-term additions and name the person authorised to send them, so a Sunday intake of 400 riders does not wait for a Monday approval chain.

Two documents carry most of the value on this class, and both are written by the broker rather than the insurer.

The first is a declaration reconciliation schedule: month by month, the number declared, the number actually on cover from the platform's extract, and the variance. Presented at renewal, it does two things. It shows the underwriter the true shape of the risk, including the seasonal peak, which supports a rate built on real exposure rather than a loaded guess. It also removes the argument, after a claim, that the peak was concealed.

The second is a liability map covering each engaging entity, each staffing vendor, the cover held by each, the certificate expiry dates, and the contractual indemnity position. Where the same rider cohort could be claimed against two entities, the map should say which policy responds first and whether the wordings are consistent on that point. Inconsistent other-insurance clauses across a platform's own policy and a vendor's policy produce contribution disputes that stall payment to a dependent family while the entities argue.

Both documents are also the honest test of whether the programme was structured for the business the platform actually runs. A fleet insurance programme built around a stable roster and priced on a fixed declaration is a programme designed for a company that does not exist during the festive quarter. The gap between the two is the exposure, and it is measurable well before it becomes a claim.