What changed · Retail
Five ways Indian consumer brands sell, and what each one changes
Nearly every article about customer data, retention and attribution silently assumes one route to market. Here is what changes when yours is different, including when it is several at once.
20 August 2026 · 7 minute read
Where a brand sells determines what it can know about its customers, what retention even means, and which marketing activities are possible at all. A direct-to-consumer brand on its own site can see everything and act immediately. A marketplace brand gets volume without identity. A general trade brand has to persuade a retailer and a shopper separately, with opposed interests. Most Indian consumer brands are two or three of these at once, with different data and different constraints in each, which is why generic advice so often fails to land. The useful first step is being honest about which mix you are, because it decides which of the standard approaches apply to you and which are simply unavailable.
The five, and what each one gives you
| Route to market | What you know about the customer | What retention looks like |
|---|---|---|
| Your own site, Shopify or custom | Everything: identity, history, behaviour | Ordinary. Group them and make contact |
| Marketplaces: Amazon, Flipkart, Nykaa, Myntra, Meesho | Orders without identity, contact masked | Mostly unavailable. Reviews and branded search instead |
| Delivery and quick commerce: Swiggy, Zomato, Blinkit, Zepto, Instamart | Orders, no identity, a two-minute decision | Availability and position matter more than messaging |
| General and modern trade | Nothing automatic. Only what you collect | Possible if you build a phone list. Nobody does by default |
| Your own outlets or franchise | Whatever the billing counter captures | Genuinely possible, and usually untapped |
What each one makes hard
- Your own site. Nothing structural. Your problems are traffic and cost, and every standard approach applies. This is the situation almost all marketing writing is aimed at, including most of ours before this piece.
- Marketplaces. No customer list, no direct retention, margin set by someone else, and a platform that may compete with you. Effort goes into reviews, branded search and off-platform pull.
- Quick commerce and delivery. All of the above, plus a decision made in under two minutes from a deliberately narrow assortment. Going out of stock behaves like being delisted.
- General and modern trade. Two audiences with opposed interests, no automatic customer data, and a shelf you are effectively renting. Consumer pull is what gives you negotiating power.
- Your own outlets. The best data position outside e-commerce, and the most commonly wasted, because billing data sits with finance and marketing has never asked for it.
Most brands are a mix, and that is the real problem
A brand doing 40% modern trade, 35% marketplaces, 15% quick commerce and 10% own site has four data situations, four competitive dynamics and one marketing team. The mistake is running one plan across all four, usually the plan that suits whichever channel the head of marketing came from.
What transfers across all five
- Brand. What you stand for and how you sound works identically everywhere, and it is the only asset not rented from a platform or a retailer.
- Being asked for by name. Branded search on a marketplace, a customer requesting you at a counter, and someone typing your name into a browser are the same underlying thing. Every route rewards it and none can take it away.
- Being named when someone asks what to buy. Whether the question goes to a friend, a shopkeeper or an AI assistant, being one of the two or three names that comes back sits upstream of every channel you sell in.
What does not transfer, and is worth admitting
Attribution. Retention programmes. Audiences built from your own purchasers. Anything that requires knowing who bought. If two thirds of your revenue comes through channels that mask the customer, then two thirds of your business cannot be measured or contacted the way the textbooks describe, and a plan that assumes otherwise underdelivers in a way nobody can explain afterwards.
The workable version is to run the measurable channels properly, run the unmeasurable ones on brand and availability, and resist the pressure to report all of them with one metric.
A starting point for a mixed brand
- Write down the revenue split across the five routes. Most teams have never had this on one page and are surprised by it.
- For each route, note what customer data you actually hold. Not what you could theoretically obtain.
- Pick the one route where you hold identity, even if it is the smallest, and build the retention work there properly. A visible result there funds patience elsewhere.
- For the rest, concentrate on being asked for by name, since it is the only lever that works everywhere.
- Stop reporting all five against one attribution model. They are not comparable and the comparison misleads you.
A brand with four routes to market and one marketing plan is running the plan that suits whichever channel its head of marketing came from. Usually that is the one they can measure.
If you are an agency
The revenue-split-by-route exercise is a strong opening conversation for any Indian consumer client, and remarkably few have it on one page. It also protects you: agreeing upfront which channels are measurable and which are not prevents the review meeting where you are asked to justify unmeasurable spend with measurable logic.
How we do this ourselves
Nothing in the platform assumes you sell online. It works from whatever customer data you hold, including a spreadsheet of phone numbers, and the parts that matter most to offline and marketplace brands are the ones about being asked for by name: what people are saying, and what AI tells someone deciding what to buy.
What the platform covers