The first-party data playbook written for Western direct-to-consumer brands assumes something most Indian brands do not have: a moment when the brand and the consumer actually meet.
India’s dominant commercial structure is indirect. Fast-moving consumer goods (FMCG) move from manufacturer to distributor to wholesaler to kirana before reaching a consumer who pays cash. Pharmaceuticals reach the patient through a doctor’s prescription and a chemist’s counter. Financial services sell through agents and direct selling agents. In each case an intermediary owns the relationship, and the brand’s knowledge of its own consumer ends at the point where its invoice does. The distribution shape of the FMCG market has shifted only modestly in a decade; the figures below describe its structure in the mid-2010s and remain directionally true today — treat them as illustrative orders of magnitude rather than audited shares.
| Channel | Share of FMCG volume (mid-2010s, illustrative) | Consumer data available to the brand |
|---|---|---|
| Kirana / general trade | ~75–80% | None — cash transactions, no data sharing |
| Modern trade | ~15–20% | Category-level retailer data; not consumer-level |
| Owned channels | <1% | Full transaction data, negligible volume |
| E-commerce | Low single digits | Consumer-level only on the brand’s own site |
Across a decade of deployments, the pattern repeats. In discovery work at a large Indian conglomerate’s foods business in the mid-2010s, the finding was stark: no first-party consumer data at any usable scale. The only consumer insight source was a syndicated retail panel — aggregated, delayed, expensive. Social listening existed, but the raw data was owned by the agency that ran it, not by the brand. Behavioural data was simply not captured, because there was no touchpoint at which to capture it. Onto this foundation, global technology firms — mine included, at the time — proposed personalisation engines, next-best-action decisioning, and unified consumer profiles. The methodology was correct. The data prerequisite was absent. We were selling the destination without scoping the journey, and the honest verdict on that era is that the analytics stack was premised on a transaction record that did not exist and could not exist without years of prior investment in capture.
That verdict yields the operating principle this article exists to state: in an indirect economy, the first investment in a first-party data strategy is not a platform, it is a channel through which data can be captured at all. Everything else is sequencing error.
The capture channels are now well understood, and each has an Indian track record. The pack itself is a data-capture instrument: a quick-response (QR) code on packaging that pulls the consumer into a registration or reward flow converts a product the brand already distributes at national scale into its largest owned media property. Loyalty programmes are the heaviest-duty vehicle — in indirect categories they are frequently the only mechanism that attaches an identity and a consent to a purchase. Owned direct-to-consumer (D2C) commerce is the third channel, and here the framing matters more than the build: a D2C channel in an indirect market is a data investment, not a revenue channel. Its returns are not in gross merchandise value — which will remain a rounding error against general trade for years — but in consumer understanding that improves the efficiency of the intermediary-channel marketing where the real volume sits. The Indian D2C wave from 2019 onwards was driven by exactly this logic, whatever the investor decks said about revenue diversification. Finally, retail-media partnerships with the large e-commerce and quick-commerce platforms give brands campaign-level, consumer-grade attribution without owning the relationship — second-party data as a structural proxy. It is worth noticing who built that answer: the marketplaces, not the brands or their technology vendors.
Two lenses should govern which of these a given brand funds. The brand lens asks whether the capture mechanism gives the consumer a genuine reason to identify themselves — a reward, a service, a community — because capture that feels like surveillance produces junk data and burned trust. The evidence lens asks one number: consented, contactable profiles per rupee spent, and the rate at which those profiles are subsequently reachable. A D2C site judged on revenue will be shut down in eighteen months; the same site judged on data yield may be the cheapest consumer research the company has ever run. If the metric on the review slide is sales, the strategy has already been misfiled.
What did not work is equally instructive. Buying the unification layer first — a Customer Data Platform (CDP) procured before any capture channel produced volume — delivers an empty warehouse. Treating syndicated panel data as a substitute for first-party data confuses market measurement with consumer relationship; the panel tells you what moved, never who moved it or why. And waiting for the intermediary layer to digitise itself is a strategy only in hindsight: the Goods and Services Tax (GST) regime from 2017 and the business-to-business trade platforms that followed did digitise a meaningful slice of the distribution chain, but they produced retailer-level data, not consumer-level data. The kirana still transacts in cash. No policy change is coming that hands brands the consumer record; the capture channel must be built.
On Monday, a senior leader in an indirect category should do the sequencing test on their own roadmap. List every line item in the marketing technology and data budget, and sort it into three buckets: capture, unification, activation. If unification and activation spend exceeds capture spend while your consented profile base is below a few percent of your consumer base, the roadmap is inverted — reallocate until capture leads. Then pick one capture channel and give it a real target denominated in profiles, not sales: an on-pack programme across your highest-velocity stock-keeping unit, or a loyalty proposition designed around data yield. Judge it quarterly on consented profiles per rupee and reachability, and hold the platform purchases until the channel produces enough volume to justify unifying anything. The Western playbook starts with the profile store because the Western brand already meets its customer. In India, you have to engineer the meeting first.