If your agency is paid a percentage of your ad spend, read this
The standard way to charge for ad management in this country is a percentage of what you spend. Usually somewhere between ten and twenty per cent, sometimes with a minimum attached. It's so normal that most business owners have never thought about it as a choice somebody made.
Here's the thing it does, and I'd like to say it plainly without suggesting anybody is a crook.
The person you're asking "should I spend more?" gets paid more if the answer is yes.
That's it. That's the whole problem. Nobody has to be dishonest for it to bite. You just have to be human, and to be asked a genuinely uncertain question by someone whose fee moves in one direction.
Why it's normal, and the fair case for it
The model exists for a decent reason. A bigger account is more work. More budget usually means more campaigns, more keywords, more creative, more reporting, more meetings. Charging a percentage is a rough way of pricing effort without having to negotiate every time something changes, and it means a small business isn't paying the same as a national advertiser.
It also aligns one thing correctly: if the account does well and you want to scale it, the agency shares in that. There's a version of this where everybody is happy and nothing goes wrong, and plenty of accounts run that way for years.
I'm not pretending otherwise. What I'd like you to notice is the specific question where it stops working.
The question it can't answer honestly
Every ad account eventually hits the same moment. You're spending an amount, it's producing something, and someone asks whether to increase it.
There are three honest answers. Spend more, because there's demand you're not capturing. Keep it the same, because you're at the efficient point. Or spend less, or nothing at all, because the money is going somewhere it shouldn't.
Under a percentage, two of those three answers cost the person answering. Not a little. Directly, that month, on their own invoice.
And the honest answer is very often the second or third one. Search demand is finite. There are only so many people in your area typing the thing you sell in a given month, and once you're capturing most of them, extra budget doesn't find more of them, it just pays more for the same ones or reaches people further from buying. 
Why ads stop working for trade businesses goes into what that actually looks like in an account.
So the ceiling is real, it arrives in almost every local account, and the standard fee structure means the person best placed to spot it has a reason not to mention it.
What it does to the advice, not the ethics
I want to be careful here, because the version of this argument that accuses people of fraud is both unfair and wrong about how it actually goes.
It's rarely a decision. It's a thumb on a scale across a hundred small judgements. Whether to test a broader keyword set or tighten what's there. Whether to add a second campaign or fix the one you've got. Whether "we need more data" means genuinely more data or just more spend. Whether to raise the recommendation from two thousand a month to three when the honest answer is that two is working fine.
Every one of those is a defensible call in isolation. All of them lean the same way. And nobody involved feels like they're doing anything other than their job, which is exactly why it's worth knowing about the structure rather than looking for villains.
The measurement problem underneath it
There's a second thing a percentage encourages, and it's the more expensive one.
If the fee is attached to spend, then spend becomes the thing that gets managed. Reporting fills up with spend-shaped numbers: impressions, clicks, cost per click, cost per lead. All real, all easy to produce, none of them telling you whether the work was worth having.
What actually matters is which jobs came in, and that lives in your business, not in the ad platform. It only gets back into the account if somebody deliberately wires it there, which is what Teaching Google what a good customer looks like is about. Until that happens the platform is optimising for form fills, because form fills are all it can see, and the same trap catches lead-count guarantees for the same reason. A lead-count guarantee is a different product, not a worse one
The platforms themselves are explicit that they optimise toward whatever target you set them. Google's own documentation on choosing a bid strategy and on Target ROAS says as much: you tell it what you're aiming at and it goes after that, relentlessly. So pointing it at the wrong thing isn't a small error. It's a machine working perfectly on the wrong goal.
What we do instead, and what it costs us
We charge a flat fee for the operation and you pay the ad platforms directly, with your own card, on your own account. The money never passes through us.
Two consequences, and the second is the point.
You can see exactly what was spent, because it's your account and your statement. There's no markup to find, because there's nothing in between.
And when you ask me whether to spend more, my answer costs me nothing either way. That's the only reason you can trust it. I'm not a better person than someone charging a percentage. I've just arranged things so I've got no reason to talk you into anything, and I'd rather remove the temptation than ask you to take my word for it.
What that costs us is straightforward. When an account scales up, we don't scale with it. A client who triples their spend pays us the same, and we do more work. That's a genuinely worse deal for us on the good accounts, and it's the trade for being able to say "stop, you're at the ceiling" and have it mean something.
There's a related version of this in What is your monthly actually for?, which is about what a recurring fee contains rather than how it's calculated.
What to ask, whoever you're paying
Four questions, and none of them are aggressive.
Is your fee a percentage, a flat amount, or both? Some are both, with a percentage above a threshold. Worth knowing where the threshold is.
Whose account is the advertising running in, mine or yours? If it's theirs, you can't see the raw spend and you don't keep the account's history when you leave. That history is worth real money, because a Google Ads account that's been learning for two years does not perform like a fresh one.
What would make you tell me to spend less? The best answer names a specific condition: a cost per booked job above a number, or search impression share already high enough that more budget buys nothing. A vague answer isn't dishonest, it just means nobody's defined the ceiling, so nobody will notice it.
Are you marking up the media? Unusual in small accounts here, common enough elsewhere to be worth asking once.
One more thing worth knowing rather than asking. If anyone puts a revenue figure or a guaranteed return in the ad that brought you to them, the ACCC's guidance on false or misleading claims is the standard those statements are held to, and it applies whether the claim is in the headline or on the call. Checking what a promise actually commits to is not rude. It's the thing the law assumes you'll do.
The honest limit
None of this makes a percentage wrong. If you're running a large account and you want the agency invested in growing it, the incentive genuinely points the right way, and a flat fee can leave a good operator underpaid on an account that's taken off.
The point is smaller and more practical than "this model is bad". It's that you now know which question your arrangement can't answer neutrally, and you can go and get that one answered somewhere else.
Most people in this industry are trying to do a decent job inside the structure they inherited. It's the structure that's worth looking at.