To grow online sales for a D2C food brand in India, build in this order: first trust (reviews, clear ingredients, cash-on-delivery controls so returns do not eat your margin), then repeat purchase (sampling and WhatsApp reorders), then founder-led content, and only then paid ads and quick commerce listings. Paid media scales a brand whose first order already makes money and whose customers come back; it cannot rescue one that does neither. Quick commerce now handles over two-thirds of India's online grocery orders, so a packaged food brand that wants scale has to be on Blinkit, Zepto and Instamart eventually.
- Build in order: trust → repeat → founder content → paid ads → quick commerce.
- Quick commerce handled over two-thirds of India's e-grocery orders in 2024 (Bain).
- The first five reviews lift purchase likelihood by about 270% (Spiegel Research Center).
- COD orders come back 25-30% of the time against 2-3% for prepaid (Razorpay), and returned food usually can't be resold.
- Quick-commerce ad spend roughly tripled to about ₹4,000 crore in 2025: shelf visibility is bought.
The question usually arrives in a familiar shape, from a founder selling something like healthier Indian sweets on Shopify: decent product, a few hundred orders, Meta ads that stopped working, and no idea what to do next.
The answer below works through each step with Indian numbers, and gives you a calculator for the one figure that decides everything: how much you can afford to pay for a customer.
Where do Indians actually buy packaged food online?
Increasingly on quick commerce. Bain and Company's How India Shops Online 2025 report found that quick commerce (delivery in under 30 minutes) handled more than two-thirds of e-grocery orders in 2024 and about a tenth of all e-retail spend, and projected it to grow over 40% a year to 2030.
Bain and Company's report found India had over 270 million online shoppers in 2024, the second-largest base in the world, and almost 60% of new shoppers since 2020 came from Tier-3 cities and smaller towns.
What each channel is for
That shapes the channel plan. Your Shopify site is where you earn margin, own the customer and test new products. Quick commerce is where most repeat grocery-style buying happens in the metros, and some estimates quoted by Agency Reporter put its share of FMCG e-commerce sales as high as 70%.
| Platform | Orders in FY26 |
|---|---|
| Blinkit | 916 million |
| Zepto | About 640 million |
| Instamart | 412 million |
See the data as a table
| Quick commerce ad spend (₹ crore) | |
|---|---|
| 2024 | 1325 |
| 2025 | 4000 |
| 2026 (projected) | 6000 |
Source: Pitch Madison Advertising Report 2026, via Agency Reporter (Aug 2026) · Zepto alone reported ₹1,636 crore of ad revenue in FY26, up 151%.
The chart is the warning. Quick commerce advertising tripled in 2025 to about ₹4,000 crore (Agency Reporter), so listing on these apps is no longer enough; visibility on the shelf is bought, by brands with bigger budgets than yours.
That is why it comes last in the order, after you know your unit economics. My ranking of Indian retail media networks by ad revenue covers the platforms in detail.
Why do reviews matter more for food than for most categories?
Because food is a trust purchase with a body attached: buyers worry about taste, freshness, allergens and whether "healthy" means anything. Reviews are the cheapest way to answer those worries at scale, and the effect is large and front-loaded.
In late April 2016, iD Fresh Food, the Bengaluru maker of idli and dosa batter, put small fridges at the gates of apartment complexes and walked away. There was no shopkeeper and no CCTV. Residents took a pack of batter, chapatis or parathas and dropped the money into a locker fixed to the fridge. The first 17 of these "Trust Shops" opened around Bengaluru, at places such as Bharat Golf View in Domlur.
The reason was a trust problem. iD's products had no preservatives and a short shelf life, and the company sensed that buyers doubted the claim. Its marketing team expected the shops might lose money. Instead, its marketing manager told The News Minute that collections ran at 90-95% of stock value, and on some days passed 100%, as people paid extra or settled earlier purchases. Founder PC Musthafa said residents saw the fridges as a reminder of values they thought were lost.
The News Minute: Unmanned trust shops selling home products becoming a hit in Bengaluru (23 May 2016)Online, you cannot leave a fridge at the gate. Reviews, clear ingredient lists and honest ratings are how a food brand shows the same trust at a distance.
| Finding | Number | Source |
|---|---|---|
| Purchase likelihood with 5 reviews vs none | +270% | Spiegel Research Center, 2017 |
| Lift for lower-priced products vs higher-priced | +190% vs +380% | Spiegel Research Center, 2017 |
| Star rating where purchase likelihood peaks | 4.0 to 4.7 | Spiegel Research Center, 2017 |
| Effect of a "verified buyer" badge | +15% odds of purchase | Spiegel Research Center, 2017 |
| Bad review vs good review | A 1-star review moved sales more than a 5-star one | Chevalier and Mayzlin, 2003/2006 |
Using review and purchase data from retail partners, the centre found that a product with five reviews was 270% more likely to be bought than one with none, with most of the gain arriving with those first five. Reviews raised conversion more for higher-priced items (380%) than lower-priced ones (190%), purchase likelihood peaked at 4.0 to 4.7 stars rather than a perfect 5.0, and verified-buyer badges raised the odds of purchase by 15%. The limit: US retail data, and the report does not publish its full sample, so treat the numbers as directional for India.
Read the report (free)The practical lesson is that the first five honest reviews on each product page are worth more than the next fifty, and that a wall of perfect 5.0 ratings reads as fake.
The authors compared the same books' reviews and relative sales ranks on Amazon.com and BarnesandNoble.com, which let them separate the effect of reviews from the appeal of the book itself. Better reviews on one site were associated with higher relative sales on that site, and a 1-star review hurt sales more than a 5-star review helped. The limit: books, not food, and sales ranks rather than unit sales. The asymmetry is the part that travels: one unanswered complaint about stale stock costs more than a good review earns.
Read the paper (free)
How do you stop cash on delivery and returns from killing your margin?
Treat COD as a risk to be priced, not a default. Razorpay, citing ET Prime Research, puts COD at 60 to 65% of Indian e-commerce orders, with 25 to 30% of COD orders returned to origin (RTO) against 2 to 3% of prepaid orders.
- Nudge to prepaid. Offer a small prepaid discount or free shipping on prepaid orders, and send a WhatsApp payment link to COD buyers right after they order.
- Gate COD by risk. Set a cart value floor for COD, block pin codes with high refusal history, and turn COD off for first orders above a threshold.
- Confirm before dispatch. A WhatsApp confirmation for every COD order catches fake and impulse orders before you spend on shipping.
- Measure it weekly. Track RTO by pin code, product and ad campaign. A campaign with a low CPA and, say, a 35% RTO rate (an illustrative premise, not a market figure) is your most expensive one.
The full playbook is in my guide to reducing COD and RTO for Indian D2C.
How much can a D2C food brand afford to pay for a customer?
Only as much as the customer's contribution margin over the period you can wait for payback. Food has low order values and needs repeat orders to work, so the first order alone rarely covers the cost of acquiring the buyer.
Plug in your own numbers.
Break-even customer acquisition cost for a D2C food brand
If your paid CAC is above the first number, every new customer loses money until they reorder, so you are financing growth with repeat rate. If it is above the last number, stop scaling ads and fix retention first.
Paid media scales a brand whose first order already makes money and whose customers come back; it cannot rescue one that does neither.
My CAC payback and LTV guide covers cohorts and payback windows properly.
Do samples and WhatsApp reorders really work for food?
They are the two cheapest ways to earn the repeat purchase that the calculator above depends on. A sample removes the taste risk for a new buyer; a timely reorder message catches the buyer when the pack runs out. For a product eaten weekly, the second order is where a food brand becomes a business.
- Sampling: trial packs priced near cost, inserts in every order ("give this to a friend"), and tie-ups with gyms, offices or cafés that match your positioning. Track every sample code to a first full order.
- WhatsApp reorders: time a reminder to your product's real consumption cycle, include a one-tap reorder link, and keep the message useful (a recipe, a festival box) rather than a discount every time. Meta has charged per delivered message since July 2025, with India's marketing message rate raised from January 2026, so measure revenue per message sent.
- Festive gifting: Diwali, Raksha Bandhan and wedding season are when Indians buy sweets and dry fruit for others. Corporate and family gift boxes bring larger orders and new households who later buy for themselves.
Setup, templates and opt-in rules are in my guide to WhatsApp Business API marketing in India.

Is founder-led content worth the time for a food brand?
Yes, because food content has a natural story (where the recipe came from, how it is made, what is not in it) and short video is where Indians discover food. It also costs time rather than money, which suits a brand that cannot yet afford a positive-CAC ad engine.
The best-performing formats for food are:
- Process: making a batch;
- Comparison: your ingredient list against a market leader's;
- People: the cook, the farmer, the founder's family recipe.
Two rules keep it honest and useful. Make claims you can back: "no refined sugar" is checkable on the label, "healthy" is not. And post on a schedule the algorithm can learn from; my explainer on how Instagram ranks Reels covers what it rewards.
When should a food brand start paid ads and quick commerce?
When your first-order contribution is positive or your repeat rate covers the gap, and when your reviews and site convert well enough that paid traffic is not wasted. Then expand in this order, with a test budget for each.
| Channel | Cost to start | Time to a clear signal | What it does best |
|---|---|---|---|
| Reviews, COD controls, site fixes | Low: time and an app or two | 2-4 weeks | Raises conversion on all later traffic |
| Sampling and WhatsApp reorders | Low to moderate: product cost and per-message fees | One consumption cycle | Repeat rate, the base of LTV |
| Founder content on Reels and YouTube Shorts | Time | 2-3 months | Discovery and trust at no media cost |
| Meta and Google ads to your own site | Moderate: a test budget per creative angle | 2-4 weeks per test | Scaling what already converts |
| Marketplaces and quick commerce listings | High: margins to the platform plus ad spend for visibility | 1-3 months | Volume and repeat in metros |
Food brands are also part of a wider story about Indian taste; the Rasa world on this site covers the food science and regional history behind what sells. If you want a second pair of eyes on your own plan, my work page shows what I have built for other brands.
What's new
- Quick commerce advertising heads for ₹6,000 crore in 2026 as Zepto's ad revenue rises 151% Agency Reporter, Aug 2026
- Razorpay updates its COD guide: 25-30% of COD orders return to origin against 2-3% for prepaid Razorpay, Jun 2026
- Meta raises India's WhatsApp marketing message rate from January 2026 and moves Indian billing to rupees Meta for Developers, Jan 2026
Questions people ask
Should a new food brand start on Amazon or its own website?
Start where you can learn fastest and keep margin, which is usually your own site plus a small marketplace presence for search demand. Your site gives you customer data for reorders; marketplaces give volume but take a commission and keep the customer relationship.
How do I get my first reviews without faking them?
Send product to early buyers, friends of customers and relevant communities, then ask for an honest review three days after delivery with a direct link. Show 4-star reviews and reply to complaints publicly; buyers trust a 4.5 with visible replies more than a perfect 5.0.
Is quick commerce worth it for a small food brand?
Only once your unit economics hold at the platform's margin and you can supply dark stores reliably. Quick commerce ad spend tripled in 2025, so organic visibility on the apps is thin; budget for ads there from day one.
How often should I send WhatsApp messages to customers?
Time them to your product's consumption cycle and festivals rather than a fixed weekly blast. Since July 2025 Meta charges per delivered marketing message, so track revenue per message and cut segments that do not reorder.
Why did my Meta ads stop working?
Usually creative fatigue, a narrow audience exhausted at small budgets, or rising CAC against a low order value. Check whether the problem is the ad (click-through falling) or the site (conversion falling) before changing either.
Sources
- Bain and Company: How India Shops Online 2025 (2025)
- Agency Reporter: Retail media heads to ₹6,000 crore in quick commerce advertising (2026)
- Razorpay: Cash on delivery and its challenges (2026)
- Spiegel Research Center, Northwestern: How online reviews influence sales (2017)
- Chevalier and Mayzlin: The Effect of Word of Mouth on Sales, NBER Working Paper 10148 (2003)
- Meta for Developers: WhatsApp Business Platform pricing (2026)
- The News Minute: Unmanned trust shops selling home products becoming a hit in Bengaluru (source for the iD Fresh story) (2016)

