There's a number people love to throw around, usually somewhere between 10 and 20 percent of revenue. It's a comfortable rule of thumb and it's mostly useless. Here's why, and what to use instead.
The percentage depends on your margin, not a rule
If your gross margin is 60 percent, you can afford to spend a lot more on acquisition than a brand at 30 percent. The only ceiling that matters is your break-even. Your ad spend as a percentage of revenue can go right up to your gross margin before you lose money on a first order. Above that, every new customer costs you.
The math in one pass. Break-even spend, as a share of revenue, equals your gross margin. If you keep Rs.600 of every Rs.1,000 before ads, you can spend up to Rs.600 acquiring that thousand and still break even. That's 60% of revenue. Any ad budget below that ceiling is first-order profit, and how far below is a choice about how much of the margin you want to spend on growth.
So a brand spending 15 percent of revenue on ads might be printing money at a 65 percent margin and hemorrhaging at 25 percent. The percentage alone tells you nothing without the margin next to it. "We spend 12% of revenue on Google" is a fact about a company, not an answer to whether it's too much.
The number that actually matters
Blended MER, and your new-customer CAC against first-order contribution. If you spend Rs.1,00,000 on Google and it drives Rs.3,00,000 in revenue, your MER is 3x. Now compare that to your break-even MER from your margin. If break-even is 2.5x, 3x is profitable and you can push spend up. If break-even is 3.2x, you're underwater at 3x no matter what the revenue looks like.
Watch the two ratios separately, because they fail differently. Blended MER includes your branded search traffic, which converts superbly and flatters the average. People who google your exact brand name were already coming; Google charges you a toll on a road you mostly built yourself. New-customer CAC strips some of that out and tells you what acquiring a stranger actually costs. When blended MER is 4x but new-customer cost keeps climbing, the mix is shifting toward harvest and the real growth engine is slowing. That matters more than any percentage of revenue.
When you're spending too much
You're over-spending when your marginal ROAS, the return on the last chunk of budget you added, drops below break-even. Not your average, your marginal. The average can look healthy while the newest spend is losing money, because the efficient early spend is propping up the average.
Here's how that hides in practice. Your first Rs.50,000 a month returns 5x, converting existing demand, brand searches, past visitors. You push to Rs.1,00,000 and the added Rs.50,000 returns 2x. Your blended number now reads 3.5x, which looks great against a 2.5x break-even. Meanwhile the second Rs.50,000 is underwater on first-order economics, and only a marginal read would show it.
Add budget in steps. Measure the return on each step. Stop when the next step doesn't clear break-even. That's the entire discipline, and skipping the marginal read is how brands spend their way into a loss with a healthy-looking dashboard.
When you're spending too little
If every increment of Google spend returns above break-even, you're leaving growth on the table. That's the more common mistake, actually, and it's more expensive because it's invisible. Founders cap Google at some arbitrary percentage of revenue when they could profitably spend more, and a competitor takes the share instead.
Test it simply. Raise Google spend 20% for a month. If the added spend holds above break-even, raise again. You'll find the ceiling in two tests, and it'll be some unglamorous number like 22%, not the 12% you'd been running on habit. The time to find your actual ceiling is before a competitor does, because the customer you don't acquire this quarter is a customer someone else bills for the next three years.
The real answer to "what percentage"
Forget the rule of thumb. Calculate break-even MER from your margin, keep a separate eye on new-customer CAC, and scale Google until the marginal return drops below that line.
The percentage of revenue will be whatever it needs to be. A brand at a 70% margin might run 30% of revenue into Google and still print money. A brand at 30% margin might cap at 8% and is still at risk. Both made the right call, because the ceiling came from their margin and their marginal returns, from nothing else.
The conversion that turns it into a number
Rule of thumb aside, here's the conversion that makes "what percentage" answerable in your own account. Break-even spend share equals your gross margin, so take your margin and you have your first ceiling. Suppose your margin is 45%. Anything up to 45% of first-order revenue on acquisition is at worst break-even on the first order. Now decide how much of that room you actually want to spend for growth. If you want a healthy cushion after overhead, you might run Google at 18% of revenue, hold 27% of margin in reserve, and know precisely why the number is 18. It's a chosen number with a reason attached.
And when someone benchmarks you against their brand, run one question back at them: what's your margin, what's your first-order contribution, and does your percentage include your own brand search? Two founders can quote 12% and 25% and both be right. The percentage is an output of the economics, and the economics decide, so the answer to "when is it too much" always has the same shape: when the next rupee stops clearing your break-even.
I'm Hirdesh Matai, founder of HDM GRO-ads, the science of e-commerce advertising.