How Much to Spend on Meta Ads

The Direct Answer

Work backwards from revenue, not forwards from a random number. The formula: (monthly revenue target ÷ average order value) = customers needed; customers ÷ close rate = leads needed; leads × target CPL = your Meta budget. A business wanting 10 extra customers, closing 20% of leads, at ₹400/lead needs 50 leads — a ₹20,000 media budget plus learning buffer, so realistically ₹30,000–₹35,000.

The Learning-Phase Floor

Whatever the formula says, Meta's delivery system has a floor: it optimises properly at roughly 50 conversion events per ad set per week. If your market CPL is ₹300, that's ₹15,000/week. Budgets far below this still work, but exit learning slowly and swing week to week. This is why ₹5,000/month "tests" so often fail — not because Meta doesn't work, but because the system never got enough data to work with.

Percent-of-Revenue Benchmarks

As a sanity check: businesses in growth mode typically invest 7–15% of revenue in marketing, with paid media taking half or more of that. A ₹10L/month business spending ₹30–₹80K on Meta is in normal range. Early-stage businesses buying their first customers often run higher percentages deliberately — that's an investment decision, and it's fine as long as unit economics are tracked.

Splitting the Budget Inside Meta

A durable default: 60–70% to cold prospecting (broad audiences, your best creative), 20–30% to retargeting (site visitors, engagers, video viewers), 10% to testing new angles. Retargeting will show your prettiest ROAS — resist funding it beyond audience size; it can only harvest what prospecting plants.

When to Increase — and How Fast

Scale when three things hold for 2+ weeks: CPL at or under target, lead quality confirmed by sales, and frequency under ~2.5. Then raise budgets 20–30% every 3–4 days rather than doubling overnight — sharp jumps reset learning and spike costs. If CPL climbs more than ~30% after a raise, hold; the auction is telling you creative or audience needs refreshing before money does.

When to Cut

Cut when leads stop turning into conversations (a quality problem money won't fix), when frequency passes 3–4 with declining CTR (fatigue), or when follow-up capacity is saturated — more leads you can't call fast are wasted budget wearing a good ROAS costume.