It’s never been easier to outsource marketing work to automation. Advertising platforms will manage bidding, targeting and creative. CRMs will score leads, trigger workflows, and suggest the next action. You can feed performance data to an AI assistant and get optimization recommendations before your coffee is ready.
None of this is hype. These systems truly work toward the goals you give them, autonomously and continuously, and the suggestions become more and more useful.
But notice what hasn’t changed in that sentence: the targets you’re giving them.
Automation does not set a vague goal. A vague goal given to one person usually produces messy results and unclear direction.
Your customers are searching everywhere. Make sure your brand introduces himself.
The SEO toolkit you know, plus the AI visibility data you need.
@media (maximum width: 768 px) { .headline-responsive { font-size: 30px !important; line-height: 1.3 !important; } }
When automation does exactly what you asked
Give AI a vague goal, and you’ll likely get overconfident results and overconfident direction. The system will find the most efficient path in the wrong direction.
Ask for a higher ROAS and automated bidding will happily rely on branded search, warmer leads, and retargeting. People who were going to buy anyway. The ROAS increases.
Ask for more signups and your campaigns risk filling the funnel with low-intent volume that never activates. Registrations are increasing.
Ask for a lower CAC and the system will quietly narrow your reach to the simplest audience you have. The CAC is falling.
In all cases, the indicators improve, but not necessarily the company. The automation did exactly what you said. It just didn’t meet the needs of the business.
Stop giving direction to automation. Give him a field.
“We need higher ROAS” is not a goal. It’s an orientation. An optimizer will follow one direction forever, until the point where it stops helping the business.
What automation really needs is a playing field. Clear sidelines on both sides. What counts as a win, and just as explicitly, what counts as a loss, even if the metric looks good.
Take a brand that runs paid campaigns with 8x ROAS. Leadership wants growth, so the real goal is not to protect the 8x. It’s about acquiring more new customers. Stated correctly, the goal looks like this: we will accept that ROAS increases from 8x to 5x if the volume of new customers increases with it. Below 5x, we stop and re-evaluate. It’s the ground.
The AI now has room to move. This can expand audiences, seek additional customers, and spend in less efficient territories, because someone has decided in advance how efficiently the company is willing to trade and where the line lies. Without this floor, you get one of two failure modes. Either the team stifles campaigns by protecting a ROAS number that no one actually needs, or the system collapses with no agreed-upon stopping point.
Skill is not just about choosing the metric. This involves defining the two sidelines before the start of the match.
Decide what to turn off before you turn it on
The same thinking applies to new AI features within the platforms themselves, and this is where I see teams skipping homework.
Let’s take the example of an advertiser in a regulated sector, such as insurance, who activates Google’s AI Max. The default posture is to enable everything and let the system optimize. But for this advertiser, the loss conditions include things that no dashboard will ever report. The AI rewrite carefully reviewed the ad copy to make it into something compliant that was never approved. Branded terms are integrated into broader correspondence where they do not belong.
So the right decision is to decide what to disable before enabling anything. Text personalization disabled. Brand excluded. Then let the AI work hard in what’s left.
This is not distrust of technology. It’s the opposite. This gives the system a field on which it can sprint. The guardrails are what make the stand-alone safe enough to actually use.
Automating a guess is still a guess
One more version, from the CRM side, because it’s easy to assume that the problem only lies with paid media.
It’s very easy to create elaborate automated workflows. Trigger that email, assign that task, drive the customer to that action. The question that is rarely asked is whether there is data showing that customers who take this action actually retain better. Many workflows are an engineering effort layered on top of a hypothesis. The automation works. The hypothesis has never been tested.
If you don’t have anyone do the task manually because you can’t tell what it improves, automating it doesn’t make it smarter. This simply allows guessing on a schedule.
Where do humans go?
None of this is an argument for less automation. The tools are good and getting better, and refusing to use them is a risk in itself at this point.
It is a question of where human judgment now finds its place. Do not approve every bid change or review every suggested action. Human work is about owning the definition of the field: the floors, the exclusions, the compromises the company will accept, and the victories that don’t matter. Then watch for times when the system wins the metric but loses the game.
Automation will achieve whatever goal you set for it. This is exactly why the goal deserves more thought than it receives.
So here’s a question worth asking: If every automated system you use hit their numbers this quarter, how many of those wins would move the business forward?
The position Automation doesn’t eliminate vague goals appeared first on MarTech.
