How to Calculate Profit After Google Ads

The mechanics of calculating profit after Google spend are the same as for any channel: contribution margin for the period, minus everything the advertising cost. What makes Google different is that a blended account figure hides more than most.

Why the blend misleads on Google

Google accounts routinely contain campaign types with completely different economics running side by side.

Brand search captures people already looking for you. It converts at a high rate and a low cost, and a meaningful share of those buyers would have arrived through organic results if the ad had not been there.

Non-brand search captures people looking for the category rather than for you. Genuinely incremental, more expensive, and the campaigns that grow a business.

Shopping sells whatever the feed promotes, weighted toward whatever wins on price comparison.

Performance Max blends several inventories, which makes separating them harder rather than easier.

Averaged together, cheap brand conversions subsidise expensive acquisition campaigns in the reported figure. The account looks efficient. The part of it doing the actual work often is not.

So the first move is not a calculation at all — it is a split. Brand separate from everything else, Shopping separate from search. Without it, the profit figure is arithmetically correct and strategically useless.

Getting the spend figure right

The campaign view reports media cost. What advertising actually cost the business is larger.

Tax on invoices, where your market charges it. In the example above, $1,175 in one month.

Agency and feed management fees. Shopping in particular tends to involve a feed tool, and it is a cost of running the channel.

Currency conversion on accounts billed in a different currency than you settle in.

Take spend from billing rather than the reporting interface, because billing reflects credits, refunds and tax while the campaign view reflects estimated media cost.

Shopping margin is not store margin

The single most useful check on a Google account, and the most frequently skipped.

Shopping campaigns sell what the feed pushes. Feeds are usually optimised for click-through and price competitiveness, which selects for cheaper products, which in most catalogues means thinner margins. The result is that Shopping revenue can carry a contribution margin several points below the store average.

Because break-even ROAS is one divided by contribution margin rate, a lower margin means a higher required return. Shopping is frequently judged against a store-wide break-even it was never going to meet.

Calculating contribution margin on Shopping orders specifically takes an hour and often changes the budget allocation immediately. The usual finding is not that Shopping fails, but that a subset of the feed does — and excluding those products fixes the channel rather than abandoning it.

The brand search problem, stated honestly

There is no clean answer to how much brand search is incremental, and anyone claiming a precise figure is guessing.

The available evidence comes from pausing it. Turn brand campaigns off for a defined period and watch what happens to total revenue rather than to the campaign’s own numbers. If overall revenue holds, most of that traffic was arriving anyway. If it falls, the campaigns were doing work.

That test is uncomfortable — it involves deliberately not buying traffic — and it is the only method that answers the question. Running it once a year on a two-week window is cheap relative to spending against an assumption indefinitely.

In the meantime, the honest treatment is to report brand and non-brand separately and to judge scaling decisions on the non-brand figure, since that is where additional budget actually goes.

Performance Max, and what it hides

Performance Max complicates all of the above, because it deliberately blends inventories that behave differently on margin.

A single campaign can serve Shopping listings, search results, YouTube, Display and Gmail, allocating budget between them automatically. That is the point of it, and for pure conversion volume it frequently works. What it also does is fold brand search and remarketing — the cheapest, least incremental traffic available — into the same reported figure as genuinely new demand.

The result is a campaign whose return looks strong and whose incrementality is unknowable from the reporting alone. It is not that the numbers are wrong; it is that the thing being measured is a mixture.

Two partial defences exist. Brand exclusions, where the account structure allows them, keep the cheapest conversions out of the blend so the remaining figure means more. And holdout testing — pausing the campaign entirely for a defined window and watching total store revenue rather than campaign revenue — answers the incrementality question directly, at the cost of a period of deliberately unbought traffic.

Neither is comfortable. Both are cheaper than scaling a campaign whose reported return is largely a reflection of demand you already had, which is the failure this section exists to name.

Reading the result

Three numbers make the output actionable, and they are the same ones that work for any channel.

Profit after advertising, monthly. Rising while spend rises means scaling works.

Advertising as a share of contribution margin. Eighty-six percent in the example, which leaves fourteen percent of every marginal dollar to carry the fixed cost base. That ratio is usually more diagnostic than any ROAS figure, because it is denominated in the money that actually has to cover overheads.

Contribution margin per order by campaign type. Where the Shopping finding surfaces, and where a feed exclusion list usually pays for itself within a month.