Key takeaways
- Set up the pixel with Conversions API, a synced catalogue and a verified domain before you spend, and confirm every order is tracked once, with value and currency.
- Keep cash on delivery, but protect margin with prepaid incentives, order confirmation and pincode checks: Ads Manager still counts COD orders that later come back.
- Work out break-even ROAS with COD fees and RTO included. It can sit well above the number the standard formula gives you.
- On a small budget, run one campaign with one ad set so every purchase feeds the same learning.
- Plan the festive season six to eight weeks ahead, test early, and scale only proven ads as auction costs rise.
Facebook ads work for D2C brands in India when three things are in place before you spend: tracking that reports every order once, an offer whose maths survives cash on delivery and returns, and creative that shows the real product to real Indian buyers. Start with one consolidated campaign across Facebook and Instagram, judge it on profit from delivered orders rather than the ROAS Ads Manager shows, and scale only what the numbers prove.
How do you set up Meta ads for a D2C brand in India?
Skip any of these steps and your first month of data can't be trusted.
Business portfolio, ad account and billing
- Create a business portfolio (what Meta used to call Business Manager) and add your Facebook Page, Instagram account and ad account to it.
- Add two admins with two-factor authentication, so one lost login can't lock you out.
- Set currency and time zone deliberately, usually INR and IST. Changing either later is disruptive.
- Use a reliable payment method. A failed charge pauses delivery.
- Add your GSTIN in billing settings if you're GST-registered, and ask your CA how to treat the tax on Meta's invoices.
Meta Pixel plus Conversions API
Install the Meta Pixel and switch on the Conversions API together. The pixel reports what happens in the shopper's browser. The Conversions API sends the same events from your store's server, catching orders the browser misses because of ad blockers, patchy connections or privacy settings. On Shopify, the Facebook & Instagram app handles both.
Before launch, confirm that each order fires exactly one Purchase event, that each Purchase carries a value and a currency, and that browser and server events are deduplicated. Our Shopify pixel and Conversions API checklist walks through every check.
Catalogue and domain
Connect a product catalogue in Commerce Manager (Shopify's app syncs it for you). You need one for Advantage+ catalog ads, formerly dynamic product ads, which can show shoppers the products they viewed. Use the same integration for pixel and catalogue so product IDs match, keep prices and stock in sync, and verify your domain in Business Settings. Then place one real order yourself and confirm it appears once, with the right value, in Events Manager.
Should you offer cash on delivery?
For most new brands, yes, but price it, confirm it and nudge shoppers towards prepaid. Many Indian buyers won't pay upfront to a brand they've never heard of, so switching COD off can cost more first orders than it saves. Test it rather than assume. The real enemy is RTO (return to origin): COD orders that are refused or undeliverable, which you pay to ship both ways after the ad cost is spent.
There's a reporting trap too: Ads Manager counts a purchase when the order is placed, so a COD order that comes back still sits inside your reported ROAS. COD-heavy stores often look healthier in Ads Manager than in the bank.
Ways to cut RTO without killing conversion
| Lever | How it works | Trade-off |
|---|---|---|
| Prepaid incentive | A small discount, free shipping or a free sample for paying online | Costs margin on orders that would have prepaid anyway |
| COD fee | A modest charge for cash on delivery, shown before checkout | Some shoppers drop off |
| Order confirmation | WhatsApp, SMS or IVR confirmation before dispatch, and hold unconfirmed orders | Needs a tool or a person |
| Partial COD | A small advance by UPI, the rest on delivery | Adds a step at checkout |
| Pincode and address checks | Block unserviceable pincodes, flag incomplete addresses and repeat refusers | Needs courier data and clear rules |
| Fast NDR follow-up | Act on non-delivery reports within a day | Ongoing operations effort |
Price for the maths, not just the market
Your price has to absorb product cost, shipping, payment fees, RTO and the cost of winning a purchase on Meta. If a purchase costs ₹350 in ads and the product sells for ₹399, no amount of optimisation rescues the first order. Combos, bundles and "buy two" offers lift average order value so each purchase can carry its ad cost.
Show the price in the ad. If you show a struck-through MRP, it must be a price you genuinely charge. Invented reference prices mislead customers and break Meta's advertising policies.
What ad creative works for Indian audiences?
Creative that feels made for the viewer: their language, the real product in a real hand, proof they can check and a price they can weigh up in a second.
Speak the buyer's language
Ads in Hindi, Tamil, Telugu, Marathi, Bengali, Kannada or Hinglish can feel far more personal to people who think in those languages. Pair a regional-language version with location targeting for the states where it's spoken, and test it against your English or Hinglish ad on the same angle and offer. Translate the idea, not the words. If producing language variants is the bottleneck, Fly ROAS builds image and video ads from your real product photos in 95+ languages.
Show the real product in a real hand
Phone-shot footage of the product being opened, applied, poured or worn answers every first-time buyer's question: is this real, and is it what I'll get? UGC-style ads from genuine customers or creators do this naturally. Keep packshots accurate. If the label or colour doesn't match what arrives, you invite refusals at the door.
Lead with proof shoppers can check
Real ratings and review counts, "Cash on delivery available", delivery times, a clear returns policy, licences and certifications you actually hold (FSSAI for food, for example) and a founder on camera all lower the trust barrier that pushes people towards COD. Never invent reviews or results. Health and "miracle" claims in categories such as ayurveda and supplements carry legal as well as ad-policy risk.
Anchor the price, too. Put the number in the first frame or the headline, and give it context: a per-day cost ("under ₹15 a day"), a combo saving ("3 for ₹999") or a genuine MRP comparison.
Build your first batch around three or four angles (problem and solution, demo, proof, offer) with two or three hooks each. Cut them in 9:16 for Reels and Stories and 4:5 for Feed, with burnt-in captions for people watching without sound.
How should you structure campaigns on a small budget?
Consolidate. Start with one Sales campaign optimised for purchases, one ad set using broad targeting or Advantage+ audience, Advantage+ placements, and four to six ads. Every extra ad set splits the same money and purchase signals, so Meta learns more slowly.
Meta's long-standing guidance is that an ad set needs around 50 optimisation events in the week after its last significant edit to leave the learning phase. Check the current wording in Meta's Business Help Center, as it has been adjusted before. Small budgets won't get there quickly, which is exactly why splitting them is expensive.
Worked example: what a small budget can support
Example: your target CPA is ₹400 (about $4).
- Fifty purchases a week at ₹400 costs ₹20,000 a week, roughly ₹2,850 (about $29) a day for one ad set.
- With ₹1,500 a day you'd get around 26 purchases a week at target. That won't exit learning, but one ad set can keep improving on it.
- Split the same ₹1,500 across four ad sets and each gets six or seven purchases a week: too few for Meta, or you, to learn anything.
Plug your own order value, CPM, click-through and conversion rates into the Meta ads budget calculator to see what your revenue goal needs. For typical costs, see how much Meta ads cost in India. For when to move budgets to the campaign level, see CBO vs ABO.
Two more rules:
- Optimise for Purchase once you can get even a handful of purchases a week. Optimising for add to cart can bring cheaper, lower-intent traffic.
- Skip a separate retargeting campaign at first. A broad campaign already reaches recent visitors, and small warm audiences burn through frequency quickly.
How do you measure true profit when orders are COD?
Work out a break-even ROAS that includes COD fees and RTO, then judge campaigns on delivered revenue. The standard formula, break-even ROAS = price ÷ (price − costs), assumes every order is delivered and paid for, which flatters every number a COD-heavy store sees. Our guide to calculating break-even ROAS covers the basics. Here is the COD version.
Worked example: break-even ROAS with COD and RTO
Example: a ₹999 (about $10) skincare combo. Product cost is ₹250, shipping ₹70 an order, the gateway takes 2% on prepaid orders (₹20), the courier charges ₹40 per delivered COD order, and each RTO costs ₹170 (₹70 out, ₹70 back and ₹30 to repack). Of every 100 orders, 60 are prepaid and 40 are COD, and 10 of the COD orders come back. GST is left out for simplicity. Use net-of-GST prices in your own sums.
| Per 100 placed orders | Amount |
|---|---|
| Revenue Meta reports (100 × ₹999) | ₹99,900 |
| Revenue collected (90 delivered × ₹999) | ₹89,910 |
| Product cost (90 × ₹250) | −₹22,500 |
| Shipping on delivered orders (90 × ₹70) | −₹6,300 |
| Gateway fees (60 prepaid × ₹20) | −₹1,200 |
| COD fees (30 delivered COD × ₹40) | −₹1,200 |
| RTO costs (10 × ₹170) | −₹1,700 |
| Contribution before ads | ₹57,010 |
You can spend up to ₹57,010 on ads for those 100 orders and break even: about ₹570 (roughly $6) per placed order. Against the ₹99,900 that Meta reports, that is a break-even ROAS of about 1.75.
Ignore COD and RTO and the sums say something different: a ₹659 margin per order and a break-even ROAS of about 1.52. A campaign showing 1.6 in Ads Manager looks profitable on the naive maths and loses money in reality.
To model this in the break-even ROAS calculator, enter your price and normal costs, then add the gap (here about ₹89 an order) as an extra fee.
Track delivered revenue every week
Each week, divide delivered revenue from your store or courier dashboard by total Meta spend. This delivered-revenue version of MER can't be flattered by attribution windows or undelivered COD orders. If Ads Manager ROAS rises while this ratio falls, find out why before you scale.
How do you plan Facebook ads for Diwali and the festive season?
Start six to eight weeks out, test while costs are closer to normal, and in the final weeks scale only what's already proven. For many Indian categories this is the biggest sales window of the year and the most crowded auction: marketplaces run their big sale events (Amazon's Great Indian Festival and Flipkart's Big Billion Days usually fall in this period), and every brand wants the same attention. Plan budgets around higher CPMs than your September numbers.
A festive timeline
- Six to eight weeks before Diwali: lock the offer and gift packs, confirm stock and courier capacity, and brief festive versions of proven angles.
- Three to five weeks before: test festive creatives and offers at normal budgets, and build warm audiences.
- One to three weeks before: scale proven ads in steps and add delivery cut-offs ("Order by Friday for delivery before Diwali").
- Diwali week: keep spend on proven ads, and stop promising delivery once couriers can't make it.
- After Diwali: retarget festive visitors and move to wedding-season and year-end gifting angles.
Regional festivals such as Onam, Durga Puja, Ganesh Chaturthi, Pongal and Bihu can matter more than Diwali in some states. Pair them with regional-language creative.
Recalculate break-even before you discount
A discount raises your break-even ROAS. In the example above, cutting the combo from ₹999 to ₹799 shrinks the naive margin from ₹659 to about ₹463, and break-even ROAS climbs from about 1.52 to roughly 1.73 before COD and RTO are even counted. Run the numbers before you promise 20% off.
A 30-day launch plan for your first Meta campaigns
| Days | Focus | What to do | Decide at the end |
|---|---|---|---|
| Before day 1 | Foundations | Tracking tested with a real order, catalogue synced, COD rules live, break-even and target CPA set, 8–12 creatives ready | Is every order tracked once, with value? |
| 1–7 | Launch and hold | One Sales campaign, one broad ad set, 4–6 ads. No edits for three or four days. Pause ads that spend 2× target CPA without a purchase | Which angles earn clicks and add-to-carts? |
| 8–14 | Second batch | New hooks and formats on the best angle. Check RTO on the first COD orders. Tighten COD rules if needed | Which ads beat target CPA? |
| 15–21 | Consolidate and scale | Keep winners, pause the rest. Raise budget in steps (for example +20–25% every two to three days) | Does CPA hold as spend rises? |
| 22–30 | True-profit review | Reconcile delivered revenue, RTO and spend. Check delivered-revenue MER. Plan next month's creative | Scale, fix the offer, or rethink the product |
Pre-launch checklist
- Business portfolio with two admins, and its currency, time zone, payment method and GSTIN set
- Pixel and Conversions API live and deduplicated, with one Purchase per order carrying value and currency
- Domain verified and catalogue synced with matching product IDs
- COD rules, prepaid incentive and order confirmation in place
- Break-even ROAS and target CPA calculated with COD and RTO included
- 8–12 creatives across 3–4 angles, in the languages your buyers use
- A weekly slot booked to reconcile delivered revenue against ad spend
How Fly ROAS helps
Fly ROAS is an AI media buyer for Meta ads. It reads your brand from your website link in about 90 seconds, decodes the angles behind competitors' long-running Meta Ad Library ads without copying them, builds ads from your real product photos, and launches them in your own ad account, paused by default.
Autopilot runs 48-hour tests, by default stops any test that spends 2× your target CPA without a sale, and scales winners in steps inside a monthly cap and daily limit only you can raise. In the default "Ask me first" mode, nothing goes live without your approval. Give it a target CPA based on your COD- and RTO-adjusted break-even. The 14-day test flight costs $19.99 (₹1,939), and Meta bills ad spend directly to your ad account. Start your test flight.
Start narrow, measure honestly, then scale
Get tracking right, price for COD and RTO, show the real product in your buyer's language, and keep the structure simple until your data earns more complexity. Judge every week on delivered-order profit, not the dashboard's ROAS. Then scale what's working in steps, and have your festive plan ready before the season starts.
Frequently asked questions
How much should a D2C brand spend on Facebook ads in India?
There's no universal figure, so work back from your target cost per purchase. A practical starting point is enough daily budget for a few purchases a day at target in a single ad set, for example ₹1,500 to ₹2,500 a day if your target CPA is ₹400. A budget calculator fed with your own order value, CPM, click-through and conversion rates will size it more precisely.
Should a D2C brand offer cash on delivery?
For most new brands in India, yes, because many first-time buyers won't prepay a brand they don't know. Protect your margin with a prepaid incentive, a modest COD fee or partial COD, and order confirmation before dispatch. Then include COD fees and RTO costs in your break-even ROAS so your ad decisions reflect what you actually collect.
What is a good ROAS for Facebook ads in India?
A good ROAS is one comfortably above your own break-even ROAS, which depends on price, product cost, shipping, payment fees, COD share and RTO rate. Two brands with the same ROAS can make very different profits. Work out your break-even first, then set a target with a safety margin above it.
Do regional language ads work better in India?
They can, particularly in states where that language dominates, but results vary by category and audience. Test a regional-language version against your English or Hinglish ad using the same angle and offer, and let the numbers decide. Translate the idea and the idiom, not just the words.
When should I start running Diwali ads?
Start preparing six to eight weeks before Diwali and test festive creatives three to five weeks out, while costs are closer to normal. In the final weeks, scale only proven ads, because auction costs usually rise as the festival approaches. Add delivery cut-off messages so you don't take orders you can't deliver in time.
