How to decide a Google Ads budget

The question arrives in the same form nearly every time: what should we be spending on Google Ads? The honest answer is that the number is derivable from four things you already know or can find out, and that no external benchmark can substitute for them, because published benchmarks average across businesses with wildly different margins, sales cycles, and definitions of a conversion.

Start from what a customer is worth

You need two figures: what a customer produces in gross profit, not revenue, and over what period you are willing to count it. Revenue is the wrong input because it ignores what it costs you to deliver, and a campaign optimised against revenue on a low-margin product loses money enthusiastically.

If you have repeat purchases, decide deliberately how much of the lifetime value you will spend against. Spending against three years of projected value is a financing decision as much as a marketing one, because you pay for the acquisition today and collect over three years.

Decide what you can afford to pay

Your allowable cost per acquisition is the share of that gross profit you are willing to give up to win the customer. That is a business decision about growth appetite, not a marketing formula. A company funding growth from cash flow will pick a lower share than one deliberately buying market share.

If your sales process converts leads to customers rather than selling directly, carry the conversion rate through. An allowable cost per customer divided by your lead-to-customer rate gives your allowable cost per lead, which is the number the ad account can actually be steered by.

The account cannot optimise toward a number nobody has calculated. Most accounts running without a derived target are simply running at whatever they happened to reach.

Check the market can absorb it

A target cost per acquisition you can afford is only useful if enough people are searching at a price that meets it. This is where the budget question becomes a market question, and where the answer is sometimes that the channel cannot take your money at the price you need.

• Look at search volume for the terms with genuine purchase intent, not the broad category terms. Category volume is large and mostly not for sale to you.
• Look at the cost per click those terms actually clear at, and multiply by a realistic landing page conversion rate to get an implied cost per acquisition.
• Compare that to your allowable figure. If the implied cost is higher, the fix is upstream: a better landing page, a stronger offer, or a different set of terms.
• Check impression share on the terms that already work. If you are already winning most of the available impressions, more budget will buy worse traffic rather than more of the same traffic.

Start smaller than you think and scale on marginal return

The opening budget only has to do one job: generate enough conversions to learn from within a reasonable window. Smart Bidding needs conversion volume to work, so a budget that produces a trickle of conversions leaves the algorithm guessing and produces results that get blamed on the algorithm.

Once it is running, the scaling question is not whether the account is profitable overall. It is whether the next increment of spend is profitable, which is a different and much more useful question. Average return stays comfortable long after marginal return has stopped justifying the increase.

In practice that means raising budget in steps, holding each step long enough to read past the learning period, and watching whether cost per acquisition holds. When it starts climbing consistently at each new increment, you have found the ceiling for the current structure. Getting past it requires new terms, better conversion rates, or a higher allowable cost, not more budget against the same setup.

How to split it across campaign types

Allocation follows intent, and the ordering is more stable than the percentages.

1. Fund high-intent search first. These are people describing the thing you sell, and it is the cheapest genuine demand available.
2. Fund brand defensively and deliberately. It looks efficient because it converts well, and some of that conversion would have happened anyway, so treat it as insurance rather than as your best performing campaign.
3. Fund remarketing next, capped. It is cheap and it flatters attribution, which makes it easy to overfund.
4. Fund prospecting and Performance Max last, from surplus, once the intent-driven campaigns are not budget capped.

Before increasing any of it, check the conversion rate of the pages that traffic lands on, because a better rate lowers cost per acquisition on every campaign at once. That is why we usually run a conversion rate optimization pass before a budget increase, and why our engagement models start with a diagnostic rather than a spend commitment.