Academy › Publish › Get found on Google
The budget that survives the test
You'll walk away with a daily budget calculated from your service's real margin and the test size you need, written down and fixed until the day you evaluate it.
By Yodai · Published September 6, 2026 · Updated October 12, 2026
You open the screen where Google asks how much you want to spend per day, and it suggests numbers: five, ten, twenty dollars. You pick the one that "sounds reasonable" and launch the campaign. That small choice decides whether, two weeks from now, you'll actually know something, or you'll be stuck with the same doubt as always.
That number shouldn't come from whatever's left over this month. It should come from a calculation with two figures you pull from your own numbers: how much margin each customer leaves you, and how much you can afford to pay per click without losing money on it.
If there's no calculation behind the budget, it doesn't matter whether you set five dollars or fifty: at the end of the test, you won't be able to read the result. You'll have a Google bill and no answer.
What sets the number isn't your wallet, it's your margin
Start with the most you can pay per click without the campaign costing more than the customer is worth. Say that's $2 a click, for an emergency call-out with a $40 margin. That number is your ceiling.
Your daily budget isn't that ceiling. It's something else entirely: how many of those clicks you need to buy to find out if the campaign works. And that's where most people get it wrong from the start.
A lot of people set a low budget thinking that's the "safe" way to test the waters. But testing with a handful of clicks isn't caution, it's not testing anything at all. If you only buy fifteen clicks in two weeks, and on average it takes one in twenty to land a customer, chances are you won't land a single one.
And then you're left not knowing whether the campaign doesn't work or you simply didn't give it enough clicks to have a chance. That's the most expensive outcome of all: paying for the test and still having no answer.
Three customers isn't an arbitrary number either. With one or two, you can't tell a lucky streak from a real trend: that one customer might have come in even with a weak campaign, or none might have come in even with a strong one. With three, you've got more than a one-off story, though it's still thin: it's a minimum worth reading into, not final proof.
- 1Your per-click ceiling
Divide the margin a customer leaves you by the clicks it takes to land one: a $40 margin over twenty clicks comes out to $2. Pay more than that, and every customer costs you more than they're worth. - 2How many customers to trust the result
Decide how many customers you need to see come in through the campaign before you trust the result. Fewer than three isn't reliable data, it's luck dressed up as data. - 3The clicks you'll need
Multiply those customers by the average number of clicks it takes to land one. If one customer comes in for every twenty clicks and you want to see three, that's sixty clicks total. - 4The total cost of the test
Multiply those clicks by the maximum you can pay for each one. Sixty clicks at $2 each comes to $120 for the whole test. - 5Spread across the days
Divide that total by the number of days the test will run. $120 over fourteen days comes to just over $8.50 a day.
- Feels more cautious, because you're spending less each day
- It takes weeks to build up the clicks you need to see any result
- By the end of the test, you still don't know whether the campaign fails or whether you never gave it the chance to work
- It comes from a written calculation, not a feeling of caution
- It builds up, within the timeframe you set, the clicks needed to see two or three customers come in
- If it doesn't work, you'll know it from real data, and you can shut off the campaign without hesitation
Which one to pick: Go with the budget based on your margin. Only drop down to the "safe" budget when your service's real margin can't handle paying more per click without losing money, and at that point the problem isn't the budget: it's that the service needs a price review before you go chasing clicks.
- A daily budget based on your real margin and the size of test you need, not a round number picked at a glance
- The math, written down and saved: per-click ceiling, customers needed to trust it, total clicks, daily budget
- The number stays fixed until the day you set to judge the campaign, no touching it early out of impatience
- Coming up in the next lesson: which exact search phrase is worth paying for that click