She changed where the money went... then the numbers started moving...
CPR Down 50%... ROAS Up 50%... Labor Day Target Crushed
The interesting part wasn’t the sale. It was what she had started doing weeks earlier to cut spend, move money and quietly improve performance before Labor Day ever arrived...
Her goal was to beat last year’s Labor Day sales by 15%.
She finished 26% over.
And when the team looked at what had happened inside her campaigns...
CPR was down 50%.
ROAS was up 50%.
Spend was down.
Revenue was up.
Of course, it was a sale.
She knew that.
In fact, when she saw the numbers, she was the first to point it out:
“Obviously we had a sale... kind of helping to drive that.”
But the performance improvement hadn’t started on Labor Day.
It had started weeks earlier.
She had begun changing how she made budget decisions.
Instead of staring at each platform and trying to decide what deserved more money...
She began using Blueprint’s AI decision model to help identify where spend should be cut, where money should move, and where the next dollar had the best chance of producing another result.
Then she started acting on those recommendations.
Cutting spend.
Moving money.
Changing campaigns.
Optimizing around the results the business actually cared about.
And almost immediately, the account started behaving differently.
Performance improved roughly 30–40%.
All before the Labor Day results came in.
That’s what makes this story interesting.
Because most marketers think of AI in marketing as a way to make things.
Write another ad.
Generate another image.
Create another report.
But there may be a much more valuable use hiding in plain sight.
Helping you decide what to do next.
Because when you’re managing dozens, hundreds, or thousands of ads, the hard part isn’t finding another button you could push.
It’s knowing which button deserves to be pushed now.
Which campaign has more room.
Which one is quietly wasting money.
Where another $1,000 could create more results.
And where spending another dollar is likely to do almost nothing.
That’s a very different AI problem.
It’s not:
“Can AI make my ads?”
It’s:
“Can AI see something in my performance data that changes where I put my money tomorrow?”
That was the shift happening here.
The Labor Day promotion still mattered.
The creative still mattered.
The offer still mattered.
Nobody is claiming an AI model magically produced every point of that 26% growth.
But she entered Labor Day with something she hadn’t had a few weeks earlier:
A system helping her continuously decide where the money should go next.
And when the sale ended?
She told the team:
“I feel like we’re in a good spot.”
That may be the more interesting AI story for marketers.
Not AI replacing your marketing.
AI helping you make the dozens of allocation decisions that quietly determine how well your marketing performs.
And in her case...
Those decisions were followed by a 30–40% jump in performance...
Then CPR falling 50%...
ROAS climbing 50%...
And a Labor Day target that was supposed to be +15%...
Finishing at +26%.
The question isn’t whether AI can write your next ad.
It’s whether it can answer something far more valuable:
Where should your next dollar go?