Ask three agencies how much you should spend on Google Ads and you'll get three confident numbers, none of them derived from your business. The honest answer is that nobody can hand you the right budget, because the right budget is a calculation, not a quote. It works backwards from three things only you know: what a customer is worth to you, how often a lead becomes a customer, and how long your cash flow can wait while the account learns.
This article walks through that calculation. To keep it concrete, we'll follow one worked example the whole way through: a plumber whose average job is worth $400. Every number attached to that plumber is an illustration of the method, not a benchmark. Swap in your own figures as you read.
Step 1: Start with the value of one customer
Most businesses set an ad budget by asking what they can spare each month. Flip the question. Start with what one new customer is worth, because that is the thing your budget is buying.
Our plumber's average job brings in $400. That's revenue, not profit — parts, labor, fuel, and overhead come out of it. Say roughly half survives as margin: about $200 of actual profit per job.
Then ask the question that changes the math: does a customer call once, or come back? The homeowner who calls about a burst pipe may call again about the water heater, and mention you to a neighbor. If your customers tend to return, one new customer is worth several first jobs, and you can afford to pay meaningfully more to win them. If your work is genuinely one-and-done, price the first job only. You don't need a precise lifetime value — you need an honest range you'd defend to your accountant.
Step 2: Back into your allowable cost per lead
Now decide what you're willing to pay for one customer. There is no universal ratio; it's a judgment about how much margin you'll trade for growth. Then remember that ads don't deliver customers — they deliver leads. Calls and form fills, only some of which close. Divide what you'll pay for a customer by the number of leads it takes to win one, and you have your allowable cost per lead.
Keeping about $200 of profit on a $400 job, our plumber decides he'll spend up to $100 in advertising to win one customer — half the first job's margin, easy to justify if some customers return.
He wins about one job for every four leads he answers, so a customer costs him four leads.
$100 per customer ÷ 4 leads = $25 allowable cost per lead
That one figure is the spine of the whole budget. It tells you which keywords you can afford, when a campaign is working, and when it's quietly bleeding. If you don't know your close rate, track it for a few weeks before you spend a dollar: count leads, count wins, divide. A rough number you measured beats a precise number you invented.
Step 3: Fund enough data to learn from
Here is the step most small budgets skip. A budget has two jobs — buying leads, and buying information — and early on, the second job matters more.
A campaign that produces two or three leads a month tells you almost nothing. Was the slow month a bad campaign or bad luck? You can't say, and neither can Google's bidding system, which improves by learning from conversions. Starve it of conversions and it never gets better.
So the floor for your budget isn't "what you can spare." It's enough conversions per month to read the data: enough that one fluke week doesn't look like a trend, and enough that you can compare this month to last month and believe the difference. Work it backwards — decide how many leads per month would let you honestly judge the account, multiply by your allowable cost per lead, and that product is your monthly floor. For our plumber at $25 a lead, every ten leads of monthly signal costs $250 of budget; how many leads it takes to see a pattern depends on how noisy his market is.
If the floor comes out higher than you can sustain for several months, don't shrink the campaign until it fits. A budget too small to generate readable data isn't a small campaign — it's a slower way to lose the same money.
The three budget mistakes we see most
Spreading the budget across too many campaigns
One funded campaign beats five starved ones. Every campaign you add divides your data: each learns slower, and none accumulates enough conversions to optimize. Start with your single best service in your single best area, prove it pays, then add the second campaign.
Quitting inside the learning period
The first weeks of a new campaign are the worst it will ever perform. Bids are uncalibrated, the system is still working out which clicks become leads, and cost per lead runs high before it settles. Businesses that judge the account on those weeks pay full tuition and drop out before the exam. Decide before launch how long you'll run and how much you'll spend before making the call — then keep that promise to yourself.
Scaling before tracking
If you can't tie each call and form fill back to the campaign and keyword that produced it, you don't know your cost per lead — you know your cost per hope. More spend just buys bigger uncertainty. Call tracking and form tracking are cheap and boring, which is why they get skipped, and why setting them up is the highest-return hour in the whole build. Fix tracking first. Scale second.
When not to run Google Ads
Ads amplify what you already are. Sometimes the honest recommendation is: not yet. Hold off if any of these describe you.
- You can't answer the phone. A missed call is a lead you paid for and handed to a competitor. If nobody can pick up during business hours, fix that first — it's free.
- Your website can't convert. If the page is slow, confusing on a phone, or missing an obvious way to call or book, you'll pay for clicks that go nowhere. Land the ads on a page built to convert, or don't buy the clicks.
- You don't know your numbers. If you can't estimate average job value and close rate even roughly, the framework above has no inputs. Measure for a month, then budget.
- The budget is next month's rent. If the plan only works when the first month pays for itself, the plan is fragile. Ads should be funded by money that can survive the learning period.
And sometimes the issue is sequencing, not the channel. If your buyers research slowly or you're playing a long game in one territory, building organic visibility first can be the smarter order of operations — we've written up how to decide in SEO vs. paid: which comes first.
The framework in three lines
- Value. Decide what one customer is honestly worth, first job and repeat business included.
- Ceiling. Divide what you'll pay for a customer by leads-per-customer. That's your allowable cost per lead.
- Floor. Fund enough leads per month to read the data — or wait until you can.
Working through those three lines with real numbers is the first thing we do on every engagement — it's how our pay-per-click advertising team builds every budget. If you'd like a second pair of eyes on yours, bring your numbers and we'll run the framework together.