The most common question on a first call is some version of: "What's a good Meta ads budget to start with?" The honest answer is that there is no industry number — there is only your number, and it falls out of three things you already know: what a customer is worth, what you can afford to pay for one, and how many data points the platform needs before it can learn.
Here is how we work it out.
Start from what a customer is worth, not from what you can spare
Write down two figures:
- Average order value or deal value. For D2C, your typical cart. For lead-gen, the revenue from one closed client.
- Gross margin on that. After cost of goods, shipping, returns and payment fees — not before.
A skincare brand with a ₹1,400 AOV and 60% margin has ₹840 of gross profit per order. If you're willing to spend a third of gross profit acquiring a new customer, your target cost per acquisition is roughly ₹280. That single number governs everything that follows.
For lead-gen the chain is longer but the logic is identical: if one closed client is worth ₹40,000 in profit and you close one in eight qualified leads, a qualified lead is worth ₹5,000 — and paying ₹600 for one is a good trade.
Working backwards to a monthly Meta ads budget
Meta's algorithm needs conversions to optimise. As a working rule, an ad set needs somewhere around 50 conversion events a week before it stops guessing. You don't always get there on day one, but a budget that can't get there ever is a budget that will never produce stable results.
So: target CPA × 50 events × 4 weeks ≈ the monthly spend at which the account can actually learn.
| Target CPA | Events needed | Realistic monthly floor |
|---|---|---|
| ₹280 (D2C purchase) | ~50/week | ₹55,000 – ₹60,000 |
| ₹600 (qualified lead) | ~50/week | ₹1,20,000 |
| ₹150 (raw lead / form fill) | ~50/week | ₹30,000 |
If those numbers are higher than you can commit to, that is useful information — not a reason to start anyway at a quarter of the budget. A starved account produces noise, and noise gets misread as failure.
A budget too small to learn on isn't a cheap test. It's an expensive way to buy a wrong conclusion.
Budget for creative, not just for media
The single biggest mistake we see: ₹2,00,000 allocated to media and ₹0 allocated to making things to run. In 2026 the account isn't the lever — the creative is. Most ads you launch will lose. That's the process working, not failing.
Set aside roughly 10–20% of media spend for producing new creative every month, whether that's UGC, static concepts or creator content you can whitelist. Plan to put 8–15 genuinely different concepts into the account monthly — different angles, not different crops of the same photo.
Give it a fair window before you judge it
Three weeks is not a test. A fair first read needs:
- Weeks 1–2: learning. Expect volatile CPA. Do not switch everything off on day four.
- Weeks 3–6: the first honest read. Winners emerge; losers get cut.
- Weeks 7–12: scaling on what proved itself, plus a steady stream of new tests.
Budget for at least 90 days. Brands that commit for one month and then stop have paid the full price of the learning phase without collecting any of the return on it.
Where the money actually leaks
Before raising spend, check that what you already pay for is arriving somewhere useful. In most audits we run, the account isn't the problem:
- Traffic landing on a slow homepage instead of a page built for the offer.
- Leads sitting unanswered for six hours — by which point they've spoken to three competitors.
- Conversion events firing wrong, so the algorithm optimises toward the wrong people.
Fixing any of those usually beats a 30% budget increase, and costs less.
A simple starting point
If you want one rule of thumb: start at 50 × your target CPA per week, commit for 90 days, and reserve 15% for creative. Then let the numbers — not your nerves — decide what happens in month two.
Before changing budgets, check whether your ad sets can clear the learning threshold at all with the learning phase calculator.
When the budget starts going up rather than being set, scaling without losing your return covers why efficiency falls and which part of the fall is avoidable.
Questions people ask
What is a good Meta ads budget to start with in India?
There is no industry figure, because the number falls out of your own economics. Work out what a customer is worth in gross profit, decide what share of it you will spend to win one, and multiply by how many you want a month. Then check that the result clears the conversion volume the platform needs to learn from — if it does not, the budget is the constraint, not the targeting.
How much of the budget should go to creative?
Enough that you are not running the same three ads in month three. At higher spend creative becomes the binding constraint rather than targeting, so a media budget with no production line behind it stalls on its own.
Can I start with a very small daily budget?
Only if it still produces enough conversions a week for bidding to stabilise. Below that the account cannot learn, every reading is noise, and you are paying for data you cannot act on rather than running a cautious test.
How long before I judge the budget?
Give it the learning phase plus a readable window on top — in practice six to twelve weeks. Changing the budget every fortnight restarts learning each time and guarantees you never see a clean result.
If you'd like us to run those numbers against your actual margins, tell us what you sell. We'll say plainly whether paid is the right move for you yet.