When I write an operator, the section I spend longest on is the one that lists what it will not do.
That sounds backwards. It isn’t. The limits are load-bearing — they’re the part that decides whether the thing is trustworthy six months in, when the novelty has worn off and you’re relying on it.
Here’s the list from the marketing one, more or less as it’s written in the file.
Sponsored · Productive Studio What could your AI be doing while you sleep? Two minutes tells you where you stand — and what to hand over first. Take the quiz →It doesn’t spend without approval. Ever.
No launch, no budget raise, no new ad goes live without an explicit yes. I’ve written separately about the rule that enforces it, so I won’t repeat the reasoning here. The one-line version: it’s a media buyer behind a gate, not an autopilot with a credit card.
It doesn’t guarantee results
This is the one I’d underline.
It runs the engine well — clean campaigns, tested copy, honest tracking, tight targeting. But the result depends on the offer, the budget and the market, and those three things are not inside the operator’s control. No honest marketer promises you a number.
Which leads to the line I put in the file and will happily repeat anywhere:
If someone guarantees you leads, they’re selling something.
I don’t mean that as a cheap shot. I mean it as a practical filter for a business owner who is about to spend money and doesn’t have a way to evaluate the person they’re spending it with. You usually can’t assess someone’s media-buying skill from a sales call. You can notice whether they’re willing to promise an outcome that depends on variables they don’t own. The ones who promise it either don’t understand that, or are counting on you not to.
If someone guarantees you leads, they're selling something.
The operator in this postMaven — Marketing OperatorRuns the paid side — behind a gate on every dollar.
See Maven
It isn’t a fix for a weak offer
Ads pour fuel on whatever’s already there.
If the product, the price or the promise doesn’t land, more spend loses money faster. That’s not a marketing failure, it’s a marketing success at revealing something — and it’s an expensive way to find out.
So when that’s the problem, the operator says so plainly instead of spending to prove it. I wanted that behaviour explicitly, because the commercial incentive for a human agency runs the other way. An agency paid a percentage of spend has no structural reason to tell you the offer is the problem. I’m not suggesting they’re all dishonest. I’m saying the incentive points somewhere, and it’s worth knowing where.
It can’t run on nothing
It needs real ad accounts, a real budget, and working analytics. Without a budget there’s nothing to run. Without tracking you’re spending blind, which is worse than not spending at all, because you’ll draw conclusions from it anyway.
It isn’t the organic social operator
That’s a different one — video, posting, community, organic reach. The marketing operator runs the paid side, plus SEO and reporting. “Post this to Instagram for free and reply to the comments” is not its job, and it will tell you so rather than doing it badly.
I’ve watched people buy one tool expecting it to be the whole department. Being clear about the seam is kinder than being flexible about it.
Pairs well with thisSage — ResearcherTen hours of market research, done before lunch.
See Sage
It needs somewhere capable to run
In a plain chat window it can plan, write and brief — genuinely useful, and quite a lot of the value. But it can’t pull live numbers or push a live change from there; that needs an environment with reach into your accounts.
The important part is that it tells you which mode it’s in rather than pretending. A tool that quietly can’t do the thing it’s describing is worse than one that can’t do it loudly.
Why I publish the limits
Two reasons, and the second one matters more.
The first is expectation management, which is the obvious one. People who know the edges don’t get disappointed at the edges.
The second is that the limits are how you tell whether anyone thought about it. Anything can have a feature list. A feature list costs nothing and proves nothing — it’s the easiest document in the world to write, and everyone’s looks similar. A limits list is harder, because you have to have actually used the thing, found where it stops, and then decided to say so in public rather than hope nobody notices.
When I’m evaluating something now, that’s what I look for first. Not what it claims. What it admits.
If I can’t find a single honest limitation anywhere in the marketing material, I assume one of two things: either nobody has used it hard enough to find the edges, or they found them and decided not to mention it. Neither is a reason to buy.
One more thing about replacing a marketing team, since this is the post where I’m being pedantic about claims. An operator replaces the team you’d pay — the retainer, the salary, the agency fee. It does not replace the ad budget — the money that goes to the platform. Nothing replaces that. Anyone implying otherwise is doing the thing I just told you to watch for.
The Big Five Websites, apps, ads, content and video — the five premium operators, together. See the Big Five →