How much should you actually spend on ads?

Ask ten agencies and you will get ten numbers. Better Leads puts most Australian small businesses starting out at $1,000 to $3,000 a month in Google Ads spend (Better Leads). AdNabu says most businesses spend $1,000 to $10,000 a month (AdNabu). Precision Digital puts Google Ads spend anywhere from $500 to $10,000 or more (Precision Digital).

Those ranges are honest. They also answer a different question: what do other businesses spend? Yours is what your business can afford, and what the platform needs in order to work.

The better guides already point in the right direction. Thinkify and Better Leads both work backwards from the customers you need rather than forwards from what feels comfortable. Riseo goes further and turns the question into how many customers you can profitably acquire. Prolific You and AdNabu add sensible daily starting points. We agree with almost all of it.

But working backwards gives you one number. You need two, and the useful part is what happens when they collide.

Why this matters

The ceiling: what your margin can pay

The first number is the most you can pay for a customer and still make money. The trap is working it out from revenue.

AdNabu's formula is revenue divided by target ROAS (AdNabu). Precision Digital's worked example counts a 12:1 return on revenue (Precision Digital). Both are fine as arithmetic. Neither tells you whether you made money, because revenue does not pay for ads. Margin does.

A 4x return on ad spend sounds healthy. On a product with a 25% margin it means every dollar of ads brought back one dollar of gross profit, before shipping, staff or rent. That is a lot of effort to break even.

Riseo's version has the right shape: lifetime value, multiplied by gross margin, multiplied by roughly a third to leave room for overheads and profit (Riseo). Their plumber example lands at a maximum of $385 per new customer, and shows how a $150 cost per lead converting at 30% produces a real acquisition cost of $500 (Riseo). In the dashboard that account looks busy. In the bank it loses money on every job.

So the ceiling is what a customer is worth over their lifetime in margin, not revenue, with room left for the rest of the business.

The floor: what the platform needs to learn

The second number has nothing to do with your finances. It is what the algorithm needs before it can do its job.

On Meta, an ad set needs roughly 50 conversion events a week to leave the learning phase and stabilise (Thinkify). Thinkify puts the practical minimum at $30 to $50 a day, and calls $5 to $10 a day not cautious but costly (Thinkify). On Google, automated bidding typically needs around 30 conversions a month, and below about 10 clicks a day the system never gathers enough signal (Better Leads).

Below the floor the ads still run. They just never learn, so you pay for noise.

Where the two collide

Put them side by side and you get the real answer.

Take a service business with a $60 cost per lead, as an illustration. To clear Meta's learning floor, one ad set needs about 50 leads a week. That is $3,000 a week, or roughly $13,000 a month, for a single ad set.

Now check it against the ceiling. If one in four leads becomes a customer, the real acquisition cost is $240, and those 50 weekly leads become more than 50 new customers a month. Two questions decide whether that budget works. Is $240 under what your margin can pay? And can the business actually take on that many new customers? If your ceiling is $200, the same setup loses money on every customer it finds.

This is the conversation most budget guides skip. When the floor sits above the ceiling, the answer is not a smaller budget, because a smaller budget simply drops you below the floor. The answer is a different structure: fewer ad sets so conversions pool in one place, an optimisation event that happens more often, or manual bidding on a tight set of keywords until volume justifies automation (Better Leads). Sometimes it is a different channel. Better Leads is unusually candid that businesses with an average sale under about $150 and no repeat purchase struggle to make paid search work at all (Better Leads).

The budget is rarely the lever people think it is. On one APAC lead generation account, we delivered 641 leads at a 28% lower cost per lead on 21% less spend (Dadek Digital). The budget went down, and the cost of each lead went down with it.

What people overestimate

Percentage-of-revenue rules

Riseo lists 5 to 10% of revenue for established businesses and 10 to 20% for growth-stage ones (Riseo). Digitlc suggests 10 to 15% of a marketing budget for social ads aimed at leads, and 3 to 5% for awareness (Digitlc). These tell you what is typical. They cannot tell you what is profitable, because they never touch your margin or your cost of winning a customer. Riseo says as much and recommends the acquisition-cost method instead (Riseo).

The platform's version of your acquisition cost

Every formula above divides by a cost per customer. Make sure it is your number, not the platform's. The platform reports leads. Your business pays for customers. If you close one in five, your real acquisition cost is five times your cost per lead (Thinkify). And platforms count conversions generously, which is why we always check them against the CRM or the bank. Feed a flattering number into a budget formula and you get a confident, wrong budget.

That a daily budget is a daily budget

Google treats a daily budget as an average. It can spend up to twice that on a busy day, and caps the month at roughly 30.4 times the daily figure (Better Leads). That cap applies per campaign, so three campaigns at $30 a day is closer to $2,736 a month than $900 (Better Leads). Australian advertisers also pay GST on ad spend, so a $2,000 budget is $2,200 out of the account, even if it is claimed back later (Better Leads).

That scaling is a straight line

Once you sit between the floor and the ceiling, raise spend in steps. Thinkify suggests increases of about 20% once cost per lead has held steady for two weeks (Thinkify), and Prolific You suggests 10 to 20% once a campaign has stabilised (Prolific You). Then watch what the next dollar returns rather than the average, because the average stays reassuring long after the extra spend has stopped paying for itself.

Work out your two numbers this week

Before you approve another month of spend, write down three figures. The most you can pay for a customer, from margin and lifetime value. The least the platform needs, from your cost per conversion and the learning thresholds. And your actual acquisition cost, from your CRM or your bank, not from the ad dashboard.

If your actual cost sits between the floor and the ceiling, the budget question has largely answered itself and the job is to scale carefully. If the floor sits above the ceiling, stop adjusting the number and change the setup.

Dadek Digital works the budget out from margin and real acquisition cost first, then builds the account so it can clear the learning floor without breaking the ceiling. If you want those numbers worked out for your business before you commit next quarter's spend, a free audit is the place to start.