CPR Down 50%... ROAS Up 50%... Labor Day Target Crushed

“I’m feeling good about everything.”

There was good reason.

Going into Labor Day, one marketing leader had a clear target:

Beat last year’s sales by 15%.

She finished 26% ahead.

CPR was down 50%.

ROAS was up 50%.

Spend was down.

Revenue was up.

But the interesting part of the story had started weeks earlier.

Because before Labor Day ever arrived, she had changed the way she was making day-to-day optimization decisions.

And performance had already started improving.

The numbers were no longer just telling her what happened

Like most experienced marketers, she already knew how to optimize campaigns.

The problem was everything that had to happen before she could confidently make the decision.

Pull the numbers.

Check the dashboards.

Figure out which performance data to trust.

Dig into campaigns.

Find what changed.

Decide whether something was actually underperforming or just looked that way.

Then decide where the money should move.

In fact, this particular marketing team had already dealt with a measurement problem that made those decisions even harder.

Large bulk purchases that weren't driven by advertising were being credited back to campaigns.

So the reported performance could look better than the actual ad performance.

The marketing leader had been manually pulling those orders out to calculate a more accurate blended ROAS.

Blueprint helped automate that process so those bulk orders could be excluded from GA4 and she could get what she called a “true read” on which ads, campaigns and platforms were actually driving sales.

That mattered because optimization is only as useful as the information underneath it.

If the wrong sales are being credited to an ad, you don't just have a reporting problem.

You can make the wrong decision with real money.

Then Blueprint started surfacing what to do next

Once the underlying data was in a place she trusted, the next question became more interesting:

What should she actually do with it?

This is where her use of Blueprint started changing.

Instead of another report she had to dig through, she began getting daily optimization recommendations surfaced directly to her.

Her words:

“It’s just going to help me make decisions faster. I think that’s huge.”

The recommendations could flag underperforming budgets, identify top and bottom creative, surface wasted spend and show opportunities to shift budget.

And this wasn't AI taking over her campaigns.

She was still the marketer making the decision.

Blueprint might recommend increasing a budget here.

Decreasing one there.

Turning off an ad.

Moving additional spend toward another opportunity.

But she could open the recommendation and see why Blueprint was suggesting it.

Then, as she explained it:

“I can say, all right, I agree with that, or I don’t agree with that.”

If she agreed, she could implement the action.

If she didn't, she didn't.

That distinction matters.

Because the value wasn't handing control of the media budget to an algorithm.

It was shortening the distance between:

Something changed in my marketing...

and:

I know what I want to do about it.

And the performance started moving before Labor Day

As she began using those recommendations to cut spend, move budget and optimize campaigns, the team saw performance improve roughly 30–40% ahead of Labor Day.

That was the part that made what happened next particularly interesting.

Because the marketing was already moving in the right direction.

Then Labor Day arrived.

And the numbers got even better.

The company had gone into the holiday trying to beat the previous year's sales by 15%.

Instead:

Sales finished 26% above the previous year.

At the same time, the team's post-Labor Day review showed CPR down 50% and ROAS up 50%.

The marketing leader summed up where things stood pretty simply:

“It looks much better.”

And then:

“I feel like we’re in a good spot.”

This may be a much more valuable use of AI for marketers

Most conversations about AI in marketing still start with creation.

Can it write the ad?

Make the image?

Produce the video?

Build the report?

Those things can be useful.

But for a marketer controlling a meaningful media budget, there's another question that can be worth considerably more:

Can AI help me make a better decision about where the money goes next?

Because you don't need another 50 pieces of creative if you're putting another $50,000 behind the wrong campaign.

You don't need another dashboard if the useful insight is still buried somewhere inside it.

And you don't need AI to replace the experienced marketer who understands the promotion, customer, creative and business context.

You need it to make that marketer faster.

That's what makes this client's description of Blueprint so important.

She didn't describe the value as having more data.

She described wanting the recommendations “surface[d] for us, not buried in a report.”

And ultimately:

“It’s just going to help me make decisions faster.”

Because the real optimization problem isn't always a lack of data

Sometimes it's the amount of work between the data and the decision.

The numbers are already there.

The marketer still has to determine:

What's actually working?

What's wasting spend?

Which creative is starting to weaken?

Where does budget have room to grow?

What should I change today?

And where should I leave things alone?

Blueprint's approach is to bring the relevant performance data together, get the measurement into a state the marketer can trust, and then use that information to surface specific actions for the marketer to consider.

That changes the job of AI.

Instead of simply telling you what happened yesterday, it can help answer the question that matters tomorrow:

What should I do next?

For this marketing leader, that's exactly where she wanted Blueprint to go.

As the platform learned from more accurate data, she wanted it recommending the changes she should be making, including where to increase budget and where to decrease it.

And her team's longer-term goal was even simpler:

To know “where exactly we do want to spend our money” with greater certainty.

Labor Day gave them one particularly good glimpse of what that can look like.

The target was +15%.

They finished +26%.

CPR fell 50%.

ROAS climbed 50%.

And the marketer responsible for making those decisions?

“I feel like we’re in a good spot.”

Then she added five words that may say even more:

“I’m just going to keep going.”

The interesting question isn't whether AI can make another ad for you.

It's whether your existing performance data contains opportunities you haven't acted on yet.

FAQ

How does Blueprint help marketers decide where to move budget?

Blueprint can surface daily recommendations based on performance data, including where budget may be underperforming, where there may be room to spend more, which creative is performing best or worst, and where a budget increase or decrease should be considered.

Does Blueprint automatically make optimization decisions?

The marketer remains in control. In this client's workflow, Blueprint explains the recommended action and why it's recommending it. She can investigate the recommendation, agree or disagree, and choose whether to implement it.

Why does accurate attribution matter for AI optimization?

Recommendations depend on the data underneath them. In this client's case, bulk purchases that weren't driven by advertising were being credited to campaigns. Blueprint helped remove those orders from GA4 automatically so the team could get a cleaner view of actual ad performance.

What kinds of marketing decisions can Blueprint surface?

The client described recommendations involving budget increases and decreases, turning off ads, finding additional areas to spend, identifying top and bottom creative, and spotting creative fatigue or wasted spend.

What is the benefit of AI-assisted marketing optimization?

For this client, the clearest benefit was speed. Rather than having useful information buried in a report, she wanted optimization recommendations surfaced directly so she could evaluate them and make decisions faster.