Ask ten marketing experts how much you should spend on ads and you'll get the same maddening answer ten times: "it depends." It feels like a dodge, like the doctor who won't tell you what's wrong. But it isn't a dodge, and once you see why, the whole question changes shape.
The reason nobody can hand you a number is that the number was never something to hand out. It's not a figure you pick off a chart or copy from a competitor. It's something you calculate, from one piece of information about your own business, and most owners have never been shown the equation. So they guess. They pick a round number that feels brave but not reckless, "I'll try a thousand dollars," and hope it works out. That's not a budget. That's a bet.
Here's the better way to think about it, drawn from how the people who do this well actually reason through it.
The One Number Everything Hangs On
Before you can know what to spend to get a customer, you have to know what a customer is actually worth to you. Not what they pay you once. What they're worth over the whole time they do business with you, after your costs.
That distinction is where most budgeting goes wrong. Say someone buys a $50 product from you and your margin is 60 percent. The sale was $50, but the customer was only worth $30 in actual profit. Now say that same customer comes back and buys four more times over the next two years. Suddenly they're worth $150 in profit, not $30. The person hasn't changed. What changed is how much of their lifetime you're counting.
That lifetime profit number is the single most important figure in this entire conversation, because it sets a hard ceiling. It is the absolute most you can afford to pay to acquire a customer and still come out ahead. Spend less than that to get one, and you make money on every customer you buy. Spend more, and you lose money on every single one, no matter how clever your ad is or how good your targeting looks. A brilliant campaign that acquires customers for more than they're worth is just an efficient way to go broke.
So the first move is not "how much should I spend?" It's "what is a customer worth to me?" Answer that, and you've turned a vague anxiety into an actual boundary you can reason inside of.
Why "A Percentage of Revenue" Is a Ballpark, Not an Answer
At this point someone always brings up the famous rule of thumb: just spend a set percentage of your revenue on marketing. It's a reasonable place to look, so it's worth being precise about what it can and can't do for you.
The real benchmarks are more modest than the big numbers that get thrown around. Recent surveys of what companies actually spend land marketing at roughly 7 to 8 percent of revenue on average, with business-to-business companies often lower, around 2 to 6 percent, and brand-new or fast-growing businesses pushing higher because they're trying to buy growth quickly. So if you're doing $200,000 a year, a typical marketing spend would be somewhere in the neighborhood of $14,000 to $16,000 annually, more if you're in a land-grab phase.
That's genuinely useful for one thing: a sanity check. It tells you whether your instinct is in a normal range or wildly off. What it cannot tell you is whether the spending will actually make you money. You could spend exactly 8 percent of revenue and still lose on every sale, if it costs you more to acquire a customer than that customer is worth. The percentage sizes the budget. It says nothing about whether the math underneath it works. Use it to ballpark, never to decide.
The Catch Nobody Mentions: You Can't Do the Math Yet
Here is where the honest version of this advice diverges from the tidy version. To work backward from your cost to acquire a customer, you need to know your cost to acquire a customer. And when you're starting out, or launching on a new platform, you don't. You have no data. The equation has a blank in it that only real spending can fill.
This is the trap that freezes people. They read that they should calculate everything first, they realize they can't calculate anything yet, and they either stall or they fall back on guessing. Both are avoidable once you accept the thing nobody says out loud: your first ad budget is not an investment. It's tuition.
You are not spending that first chunk of money expecting a profit. You're spending it to buy one specific piece of information you cannot get any other way: what it actually costs you, in your market, with your offer, to turn a stranger into a customer. Reframe it that way and the pressure lifts. A test budget that "loses" money but tells you your real cost per customer didn't fail. It did exactly its job. It bought the number that makes every future dollar smarter.
How Much Tuition? Enough to Lose, Small Enough to Survive, Big Enough to Care
So what's the right size for that first learning budget? The best answer I've heard doesn't come from a formula. It comes from two feelings you're trying to hold at once.
The floor is this: spend an amount that, if you lost every dollar of it and got zero sales in return, would not hurt your business. Not sting, not hurt. Because that's the honest worst case while you're still testing an unproven offer with unproven ads. If losing that money would put you in a bind, it's too much, full stop.
But there's a ceiling on the small side too, and it's the part people miss. The amount also has to be big enough that you actually care. If a seven-figure business "tests" with ten dollars a day, nobody pays attention, the results are too thin to read, and even a winner is too small to matter. You want the number to sting a little, because the sting is what keeps you in the account, watching the data, cutting what's failing, and doubling down on what's working. That engagement is worth more than the money.
Notice what this rule is not. It is not "spend enough to satisfy the platform." Ad platforms will happily tell you that you need a large budget to get out of the "learning phase" and unlock good performance. There's a grain of truth in it, but it mostly benefits them. Plenty of campaigns perform perfectly well on modest budgets while they're being tested. Prove your offer and ads can work at a small, affordable number first. Then, and only then, scale.
Now the Math Is Real, So Scale on Evidence
Once that first budget has run, the blank in the equation gets filled in. Now you know, roughly, what it costs you to acquire a customer. And now the original question finally has a real answer, because you can put your two numbers side by side.
If your cost to acquire a customer is comfortably below what a customer is worth, congratulations: you've found something rare and valuable, a machine that turns money into more money. At that point the answer to "how much should I spend?" becomes almost embarrassingly simple. As much as you can, while the math holds. If every dollar in reliably returns more than a dollar, you don't ration it, you feed it, scaling up in steady steps and watching that the returns hold as you grow.
If instead your cost to acquire a customer came back higher than a customer is worth, the worst thing you can do is spend more. More budget won't fix that. You don't have a budget problem, you have an economics problem, and it lives upstream of the ad spend: the offer isn't compelling enough, the targeting is off, or the page you're sending traffic to isn't converting the clicks you're already paying for. Fix the thing that's actually broken, get your cost per customer under your ceiling, and only then pour fuel on it.
Two Honest Caveats Before You Set a Number
Two things will quietly wreck this if you ignore them.
First, when you're testing, put your budget behind one platform, not several. Splitting a small test budget across Google and Meta at the same time is the fastest way to learn nothing about either, because neither gets enough data to give you a readable answer. Pick the one platform that best matches how your customers actually look for what you sell, prove it out there, and expand later. That single choice matters enough that it's worth its own decision before you spend a cent.
Second, "budget" is not the same as "ad spend." The money you put into the platform is only one input. The creative and the landing page decide whether that money turns into customers or evaporates. A great budget behind a weak offer and a slow, confusing page will lose to a smaller budget behind a sharp offer and a page built to convert, every time. When you plan what to spend, plan for the whole machine, not just the media.
The Number Was Never a Guess
So the next time you catch yourself trying to pick an ad budget out of the air, stop and reach for the equation instead. What is a customer worth to me, over their lifetime, in profit? That's your ceiling. What does it cost me to acquire one? That's the number you don't know yet, and the only way to learn it is to spend a deliberate, affordable amount to find out.
That's the whole method. Calculate the ceiling, buy the missing number with a test budget you can afford to lose but big enough to respect, then scale hard once the two numbers prove the machine works. "It depends" was never a dodge. It was an invitation to go find out what it depends on, which is a far more useful place to stand than staring at a blank field wondering what feels brave. The businesses that win at this aren't the ones that guessed the biggest number. They're the ones that knew their two numbers and let the math tell them what to do.
