How much should you spend on Google Ads?

The most common Google Ads budget is a number that felt affordable, and that is how accounts end up spending enough to cost real money but not enough to learn anything. The budget question has an actual structure, and working through it takes twenty minutes with your own numbers.

Start from the sale, not the spend

Three numbers you should know before any budget conversation:

1. What is a customer worth? Not one transaction: the realistic value of a new customer over whatever horizon you manage to. A $6,000 install with occasional service visits is not a $6,000 customer.
2. What can you afford to pay for one? A share of that value that leaves acceptable margin. This is your target cost per acquisition, and it is a business decision, not a marketing one.
3. What does a click cost in your market? Keyword Planner ranges, or better, a few weeks of real auction data. This one you cannot choose; the market sets it.

Now the arithmetic. If clicks cost $8 and your landing page converts 5 percent of visitors into enquiries, an enquiry costs about $160. If a third of enquiries become customers, a customer costs about $480. If that number fits under your target from step two, the channel can work. If it never can, no budget size fixes it, and you just saved yourself a year of disappointment.

The floor: enough data to steer

The budget also has to buy enough events to make decisions. A campaign producing three conversions a month cannot be optimised; every change drowns in noise, and automated bidding has nothing to learn from.

A workable rule: budget for at least 20 to 30 conversions a month in the core campaign. At the $160-per-enquiry example, that is $3,000 to $5,000 a month. In a market where clicks cost $2, the same confidence costs a few hundred. This is why “what should a small business spend” has no universal answer: the floor is set by your market’s click prices, not by the size of your business.

If the floor is genuinely out of reach, narrow until it is not: one service, one city, one campaign. A tight campaign at sufficient density beats a sprawling account spread too thin to learn, every time.

The mistakes that waste the first six months

Spreading the budget across everything. Ten campaigns sharing $2,000 means nothing gets enough data. Start with the highest-margin, highest-intent slice and expand from proof.

Judging the whole channel on week two. Early data is dominated by learning: bad queries not yet excluded, bids uncalibrated, the platform’s models untrained. Budget for a genuine test window, usually two to three months, and pre-commit to what you will measure at the end.

Scaling on a broken measurement layer. Increasing spend into an account whose conversions double-count or track the wrong events just buys wrong answers faster. The tracking audit costs less than one month of a mis-steered budget, which is why we do it first, always.

Confusing the media budget with the total cost. Management, landing pages and tracking work all sit on top of media spend. A $1,500 media budget with $1,500 of management on top has different math than $5,000 with the same management fee, which is one reason very small budgets often make more sense in other channels entirely.

When the answer is “less” or “none”

Paid search buys demand that already exists. If almost nobody searches for what you sell, the budget question is moot and the money belongs in channels that create demand rather than harvest it. And if you are at capacity, the right budget this quarter might be lower: pace down, keep the account’s history warm, and keep only the highest-margin campaigns live rather than going dark entirely.

If you want the honest version of this arithmetic run against your actual market before you commit anything, that is precisely what a free assessment is for: we pull the auction data, run your numbers, and tell you the floor, including when the floor says do not start.