Google Ads management pricing models: % of spend, flat fee, or hybrid

Every Google Ads agency will tell you their pricing is fair. Almost none will tell you what their pricing model pays them to do.

That second part is the whole game. A pricing model is not just a number on an invoice. It is a set of instructions for how the agency will treat your account for the next twelve months. Pick the wrong model and you can hire a great team that is quietly paid to do the wrong thing.

We are a Google Ads agency ourselves, running accounts for 200+ ecom brands, so read this knowing where we sit. But we will play it straight. If you want the headline cost numbers first, start with what a Google Ads agency costs for ecommerce and come back.

Model 1: percentage of ad spend

This is the default model in ecom, and there is a good reason for it. The work and the responsibility really do grow with spend. A €100k account has more campaigns, more products, more ways to bleed money, and bigger consequences when something breaks. Tying the fee to spend keeps the price in proportion to the job.

But look at what a flat percentage actually pays the agency to do. It pays them to raise your spend. Not your profit. Your spend.

Every euro you spend, profitable or not, grows their invoice. If your account is bleeding on bad search terms, the agency earns a cut of the bleeding. When they recommend "scaling the budget," you can never be fully sure whose growth they mean. Most agencies on this model are not villains. The model just never asks them the hard question: was that extra spend worth it?

The fix is not to ditch the percentage. It is to bend it. More on that below.

Model 2: flat fee, also called the retainer

A fixed amount every month, no matter what happens in the account. Flat retainers commonly run somewhere between a few hundred and a few thousand a month, depending on who you hire and where they are.

The pitch is predictability, and that part is real. You know the cost. It never surprises you.

Here is what the model pays the agency to do: as little as possible. The retainer arrives whether your account doubles or flatlines. The most profitable client on a retainer is the one who never messages, never grows, never needs anything. So the agency's book fills up with parked accounts, each paying the same fee, each getting the leftover hours after the loud clients are handled.

That is why the classic retainer horror story is always the same: great first month, then silence. Nobody is being lazy. The model simply stops paying anyone to care once the setup work is done.

A retainer is defensible in one case only: a fixed budget that will never move, where the job is maintenance, not growth. If you want to scale, a retainer pays your agency to keep you parked.

Model 3: hybrid

A smaller base fee plus a variable part - usually a lower percentage of spend, sometimes a bonus tied to a target. The base covers the agency's fixed costs, the variable part keeps some skin in the game.

Hybrids are not bad. They are just easy to rig. The tell is the ratio. If the base is big and the variable slice is small, you have bought a retainer with a costume on, and the incentive problem from model 2 is back, just partly hidden.

Ask any agency quoting a hybrid one question: how much of your fee do you lose if my account goes nowhere this quarter? If the answer is "not much," you know what the model is really paying for.

Model 4: performance-based

The boldest pitch in the industry: "we only get paid when you win." Fees tied to revenue, ROAS, or new sales. No results, no bill. Sounds like the fairest deal on the table.

It is usually the messiest. The problem is one word: attribution.

The agency gets paid on numbers it also gets to measure. So the fight moves from growing your store to defining "performance" generously. Brand search - people who typed your store name and were buying anyway - suddenly counts as agency-driven revenue. Last-click hands Google Ads credit that email and organic earned. ROAS gets propped up by pouring budget into your existing bestsellers while nothing new gets built. The dashboard glows. Your accountant sees something else.

There is a quieter problem too: an agency carrying performance risk prices that risk in. Steep result percentages, a base fee underneath anyway, or a roster of accounts that were going to win regardless. You are rarely getting a bargain. You are paying for insurance and calling it alignment.

If someone offers you pure performance pricing, ask exactly how performance is measured, on which attribution model, and who verifies it. The quality of the answer tells you everything.

The version that actually aligns: a tiered percentage that drops

Here is the structural fix for the flat-percentage problem, and it is the reason we built our own pricing this way.

Keep the percentage - the work really does scale with spend. But make the rate fall as spend grows. Our tiers run from 10% on the first €10k of monthly spend down to 6% above €150k, and each rate applies only to the spend inside its own bracket. No retainer underneath, no setup fee, no lock-in.

Look at what that model pays the agency to do. The only way the agency earns meaningfully more is by scaling you into higher brackets - and it can only keep you there month to month, with no contract holding you, if the scaling is actually profitable. Reckless spend inflation gets punished twice: the marginal rate on the new spend is lower, and an unprofitable client on a monthly deal walks.

And the falling rate answers the fairness question a flat percentage dodges. Managing €150k a month is not fifteen times the work of managing €10k. The heavy lifting - feed work, structure, tracking - is built once and maintained. A tiered model hands that efficiency back to you instead of pocketing it.

Which model fits which store size

Honest answers, including the ones that do not send business our way.

Under roughly €3-5k a month in ad spend. No agency model fits well. Any meaningful fee is huge next to your media budget, and a fee small enough to be proportionate buys almost no senior time. Run it yourself or use software, and hire once the spend justifies real hands.

Roughly €5-10k a month. The retainer trap zone. Flat-fee agencies hunt here because the fixed fee looks small and reasonable. A percentage of spend serves you better - it keeps the fee honest relative to the account size.

€10k to €80k a month. Percentage of spend, tiered, no long contract. This is where the model differences bite hardest, because the account is big enough that incentives show up in behaviour. A flat retainer here nearly guarantees under-servicing as you grow.

€80k a month and up. Tiered percentage or nothing. At this size a flat percentage becomes a growth tax, and a retainer big enough to cover the real workload loses its one advantage, predictability, anyway.

Any size, considering performance-based. Only with attribution terms you fully understand and can independently verify. If the measurement lives in the agency's dashboard, so does your fee.

The one question that cuts through every pitch

You do not need to memorise four models. You need one question: when my account grows profitably, does your income grow - and when it stalls, does yours stall too?

A flat retainer fails it. A big-base hybrid fails it. Performance-based passes on paper and fails in the attribution fine print. A flat percentage half-passes: the agency wins when you grow, but also when you merely spend. A tiered, dropping percentage with no lock-in is the cleanest pass we know of - which is why it is what we charge.

Once you have picked the model, the next job is checking what is inside the fee itself - who touches the account, what is included, and what quietly costs extra. We broke that down in Google Ads management fees explained.