Zero to ₹11 lakh a month, fully prepaid: the Svasthyaa teardown

The starting line: a snack brand with zero online revenue

Svasthyaa makes chana jor – a roasted-gram snack in a category dominated by loose, unbranded street purchase. When we started, online revenue was literally zero: no D2C motion, no ad account history, no baseline to optimise. That sounds like a disadvantage; it was actually the opportunity. A blank slate meant every structural decision – payment policy, order economics, rollout geography – could be made correctly from day one instead of retrofitted later, which is where most snack brands burn their first year.

Worth pausing on the category too. Chana jor is roasted, flattened black gram – protein-dense, honest street food. That gave the brand a real story: a traditional snack cleaned up for the health-conscious online buyer. It also set up a hard problem, because the reference price in the buyer's head belongs to the street vendor, not to a premium D2C pack landing by courier.

Three decisions defined the whole engagement. Each of them contradicts the default Indian D2C playbook.

Decision one: no COD, from day one

The default advice for a new Indian D2C brand is that COD is unavoidable – most shoppers expect it, so you eat the RTO rate as a cost of doing business. For a low-ticket snack brand, that maths is brutal: COD orders that get refused or never picked up cost you the product, both shipping legs and the ad money that bought the order, and they inflate every platform-reported metric while doing it. The full damage model is in our breakdown of COD, returns and real ROAS.

Svasthyaa launched fully prepaid instead. Yes, that filters out some buyers. But every order that comes through is real: paid for, wanted, and honest in the analytics. The 3.36× ROAS this account runs at is measured on money actually collected – not on COD orders that might come home in three weeks. For a thin-margin snack, that integrity is the difference between scaling profitably and scaling a leak.

Decision two: fix AOV before scaling spend

A single snack pack can't pay for its own customer acquisition – shipping and CAC eat the order value before margin enters the conversation. This is the structural trap of the category, and we've written the general maths in snack brand ROAS and low-AOV math. The fix is not better ads; it's a better order.

So before spend scaled, the store was rebuilt around bundles: multi-pack and assortment offers as the default way to buy, positioned as the sensible choice (stock-up value, variety to share) rather than a forced minimum. Bundle-led ordering moved the average order into territory where prepaid shipping and ad costs fit inside the economics – and it did a second job people miss: a household with a stocked multi-pack finishes it and reorders. AOV architecture and repeat behaviour are the same lever pulled twice.

Decision three: city-first, not nation-first

The tempting move with Meta ads is to target all of India on day one – the audience is bigger, the CPMs look friendlier, the dashboard fills up faster. The problem: your spend spreads thin across geographies you can't learn from, delivery times vary wildly, and you build no density anywhere.

Svasthyaa launched city-first: concentrate the budget on one city, learn which creatives, offers and audiences convert there, tighten delivery, then expand city by city with a playbook that already works. Concentration buys you three things – faster statistical signal on every test, better delivery experience (which protects repeat rates), and local word-of-mouth density that national scatter never creates. It feels slower for the first few weeks. It is faster by month three.

City-first also disciplined the creative. Ads could speak to a specific place and a concrete delivery promise instead of generic national messaging, and every learning cycle ran against one consistent audience. When the playbook moved to the next city, it moved as a checklist – offers, hooks, audiences, delivery SLAs – not as a hope.

What the machine produced

With those three structures in place, the media playbook itself was disciplined rather than exotic: creative testing into the winning city, doubling down on proven hooks, expanding geography only when the previous city held its numbers. Four months in, the account crossed ₹11 lakh/month at a 3.36× ROAS – fully prepaid. Every rupee of that revenue was collected at checkout; none of it was COD paper waiting to bounce.

And the D2C engine turned out to be the opening act: proof that the product sells at a healthy AOV, city by city, is exactly the evidence retail buyers respond to – the brand's rollout has since expanded toward shelves, with the online numbers as the pitch deck. The full case, with the visuals, is on the Svasthyaa case study page.

The other thing the numbers don't show: nothing about this account is exotic. No hack, no one-week spend surge, no discount spiral. The ROAS held because the order economics were built before the scaling started – which is exactly why the model transfers to other brands. It's structure, not magic.

What another snack brand can copy

The transferable playbook, in order:

• Decide your COD policy on maths, not fear. Model RTO cost per order honestly; for low-ticket food, fully prepaid (or heavily prepaid-incentivised) usually wins.
• Build the order before the ads. Bundles-as-default until your AOV covers shipping + CAC with margin left. No ad creative fixes a broken order size.
• Concentrate geography. One city until the playbook proves out, then replicate. Density beats reach at this stage.
• Measure collected money. If your ROAS isn't computed on delivered, paid revenue, you don't know your ROAS.
• Treat D2C proof as a retail asset. The dataset you build online is your distribution pitch offline.

The wider category playbook – positioning, creative angles, channel mix – is in how to market a snack brand in India.