Guides / Paid media / 8 min read / Updated 1 August 2026

How much should you spend on Google Ads?

Everybody wants a number. The number is not a secret being withheld from you, it is an output of your own margins — and the arithmetic takes about ten minutes.

The question underneath the question

Nobody actually wants to know what Google Ads costs. Google Ads costs whatever you tell it to. What people want to know is the smallest amount that will produce a real answer, because the fear is spending for three months and still not knowing whether the channel works.

That is a good fear and it points at the right problem. A budget too small to produce conversions does not produce a cheap answer — it produces no answer, over a longer period, for the same total money.

Work backwards from one sale

Start with what a customer is worth, not what a click costs. Four numbers, in this order:

1. Average order value. What a first purchase is, on average.
2. Gross margin. What is left after the cost of the goods and the cost of fulfilling them. This is the number that pays for advertising, not revenue.
3. Site conversion rate. Of the people who land on the page, how many buy. If you do not know this, you are not ready to buy traffic — you will not be able to read the result.
4. What you are willing to pay for a sale. Some share of the gross margin. Not all of it, unless you are deliberately buying customers at break-even and know why.

An example with round numbers, so you can substitute yours. A $120 order at 40% gross margin leaves $48. If you want half of that to survive as profit, your maximum cost per acquisition is $24. At a 2% site conversion rate it takes 50 clicks to make a sale, so your maximum cost per click is about $0.48.

Now go and look at what a click actually costs in your category. If the going rate is $3, that arithmetic has just told you something important before you spent anything: at a 2% conversion rate this does not work, and the fix is upstream — a better page, a higher order value, a narrower keyword set with buying intent — not a bigger budget.

Then work forwards to a monthly floor

Once you have a target cost per sale, the monthly floor is that number multiplied by the conversions you need in order to believe the result.

That second figure is the one people skip. Automated bidding — which is now most of how Google Ads is bought — is a model fitted to your conversion data. With four conversions a month it has nothing to fit. Google's own documentation has historically pointed at roughly 30 conversions in 30 days as the point where a target-CPA strategy has enough to work with, and while the exact figure moves, the shape of the advice does not.

So: $24 target cost per sale × 30 conversions = about $720 a month as a floor for one bidding unit. Not a recommendation, an arithmetic consequence. Yours will be different and you can compute it in a minute.

If the floor is more than you have

This is the common case and it has a real answer. Do not spread a small budget across five campaigns so that each of them starves. Narrow until one campaign can clear the floor.

• Fewer keywords, higher intent. The bottom of the funnel first. Somebody searching your exact product plus "buy" converts at a multiple of somebody searching the category.
• One geography. Not the whole country because the whole country is available.
• One campaign, one bidding strategy. Concentration is how a small account gets to statistical significance at all.
• Manual or maximise-clicks while the data is thin, then move to a conversion-based strategy once there is something to learn from.

A campaign that clears the floor in one city beats five campaigns that clear it nowhere.

What the platform will happily let you do

Google will take $200 a month indefinitely and never mention that the account cannot produce a readable result at that level. The interface will keep showing impressions, clicks and a green recommendations score. None of that is dishonest; it is just not answering your question.

The same applies to the recommendations tab, which optimises for spend and for adopting Google's newer products. Some of it is genuinely useful. All of it should be read as advice from a counterparty, and auto-apply should be off.

Signs you are spending too little

• Campaigns sit in "Limited by budget" every day and the impression share lost to budget is high — you are being throttled out of the auctions you already qualified for.
• Conversions per month are in single digits and the cost per conversion swings wildly between weeks. That swing is noise, not performance, and reacting to it makes things worse.
• You are changing something every few days because there is never enough data to wait for.

Signs you are spending too much, too early

• Spend is going to broad-match terms you would never have chosen, and the search terms report is full of research queries rather than buying ones.
• You are scaling a campaign whose conversion tracking has never been checked against your back office. That is the expensive mistake, and it is its own guide.
• Budget went up more than about 20% in a day. Large jumps re-open the learning period and you pay for the same education twice.

The management fee question, since it changes the answer

A management fee charged as a percentage of ad spend rewards the manager for spending more, and everybody in this industry knows it. It is not automatically wrong — large accounts genuinely take more work — but on a small account it puts your interests and theirs on opposite sides of every budget conversation.

A flat monthly fee has the opposite bias: it rewards keeping you as a client, which is the incentive you actually want. Whichever way it is priced, ask the question out loud before you sign, because the answer tells you what advice you are going to get later.