How much should a D2C brand spend on Meta ads in India? A budget guide for 2026

Most founders ask this question backwards. They want a number — ₹20,000 a month, ₹2 lakh a month — before they have worked out what that number needs to buy. Meta will spend whatever you give it. Whether that spend teaches the algorithm anything, or just burns cash while it "learns", depends on maths you can do in five minutes, on your own phone, before you touch Ads Manager.

What Meta actually requires — and what it doesn't promise

Meta's own help pages on ad set budgets and on the "learning phase" describe two separate ideas that get run together constantly.

The first is a technical minimum: the smallest budget Meta's delivery system will accept without warning you that it's too low to deliver consistently. This exists so an ad set doesn't starve itself of impressions. It is not a promise of results.

The second is the learning phase: the period after you launch a new ad set, or edit an existing one significantly, during which Meta's system is still working out who to show your ad to. Meta's guidance is clear that a new or heavily edited ad set needs a meaningful run of the result you've asked it to optimise for, within a short window, before delivery settles and your cost per result stabilises. What Meta does not publish is one fixed figure — a single number of conversions, in a single number of days — that applies across every objective, bid strategy and account. If you've read that a magic number exists, treat it as a rule of thumb passed between agencies, not something Meta guarantees in writing. The two pages linked at the bottom of this post are where that guidance actually lives; read them before trusting anyone's paraphrase of them, including this one.

The practical upshot for a founder: a budget that clears Meta's technical minimum can still be far too small to ever leave the learning phase, because it never buys enough of the result you're optimising for.

Why we won't quote you a CPM or CPC range

You'll find plenty of blog posts quoting an exact average cost-per-click or cost-per-mille for India. We looked for a citable, reputable, dated report with India-specific Meta ad cost figures we could stand behind, and couldn't find one we were confident hadn't been quietly generated to look authoritative rather than measured. Costs also move by quarter, by category and by creative, so a number that's accurate in September could be stale by December. Rather than repeat a figure we can't stand behind, here's a method that works whatever your category's real costs turn out to be.

Work out your own test budget

Before you pick a monthly number, answer three questions about your own business:

• What's your average order value (AOV)?
• What's your contribution margin per order — after product cost, packaging and shipping, before ad spend?
• What conversion rate does your product page already get from cold traffic (or your best honest guess, if you don't have data yet)?

Those three numbers tell you the maximum you can pay to acquire an order and still make money on it. From there, the budget question becomes simple: can your daily spend buy enough orders, at that maximum cost, to actually mean something?

Contribution margin per order	Maximum sensible cost per purchase	What a thin daily budget buys you
₹300	up to ~₹300	At ₹1,000/day, roughly 3 purchases — enough to start seeing a pattern
₹600	up to ~₹600	At ₹1,000/day, roughly 1–2 purchases — thin, but workable over a longer test
₹1,200+	up to ~₹1,200	At ₹1,000/day, well under 1 purchase — the budget is too small to learn anything from

That table is illustrative arithmetic, not a benchmark — the point is the shape of the problem, not the exact numbers. If your own maths lands in that bottom row, the fix isn't a cleverer campaign structure. It's a bigger budget, a cheaper offer to test with first, or fixing the product page so the same traffic converts at a higher rate before you pay to send more of it there.

Why a very small test teaches little

This part isn't sourced from Meta. It is our own reading of the arithmetic above. Below a certain spend, a month is too short a window to produce enough signal to learn from: the same "not enough volume" problem Meta's own learning-phase guidance describes for a single ad set launch, applied to a whole month of running an account. Where that line sits depends on your price and your margin, which is why there is no single number here for everyone.

If a proper test is more than your business can spare right now, that's a genuine answer, not a sales objection. Run the ads yourself, in your own ad account, at whatever budget you can manage, and fix the page first. Often the product page or the offer, not the ad spend, is what's actually capping results.

What to do this week

• Pull your last 90 days of orders and work out AOV and contribution margin per order — not revenue
• Decide the maximum cost per purchase you can tolerate and still be profitable within roughly 60 days
• Check whether your planned daily budget can buy at least a handful of purchases per day at that cost — if not, it's too thin to learn from, not just too thin to see results from
• Read Meta's own budget and learning-phase help pages (linked below) before accepting anyone's specific number, including ours

If the maths above says your business isn't ready for a real test yet, don't guess at a smaller number instead. Fix what the traffic lands on first. If you sell on Amazon, a free listing check gives you a written note on five things anyone can see on your listing, within two working days.