Two Months, 3x ROI: Can You Market with AI Without Guessing?

A while back, I came across a marketing case study from 2022.

A bit dated. But the more I turn it over, the more interesting it gets.

The star of the story is a North American entertainment group. Hotels, concert venues, live shows, resort destinations—a whole portfolio of brands, each running its own show. Back then, they used AI to rebuild their content marketing from the ground up.

The result?

In two months, they turned their marketing spend into a 3x return.

Break down the numbers: engagement rates up 197%; 6.8 million impressions and 754,000 interactions; 397 pieces of content produced across the campaign, with the performance of every single one reviewed after the fact and fed back into the model.

Impressive.

You might be wondering: it's just content marketing—does it really have to be this elaborate?

It does. Because look closely at what they did, and you'll find they got exactly one thing right:

they stopped guessing their way through creative decisions.

The creative ideas stayed the same. What changed was how decisions got made.

What Makes Entertainment Marketing So Hard?

Think about it: how many different kinds of people walk through the doors of an entertainment group?

The people staying in the hotels are one crowd. The ones coming specifically for a show are another. The families hauling kids off to a resort destination are a third. And each crowd's tastes are completely different.

There's a name for this: micro-communities. Behind every brand, every business line, stands a small crowd of discerning, hard-to-please fans.

So these brands' goals were clear: build buzz for these North American entertainment experiences; keep the content authentic and reusable across every channel; guarantee exposure and engagement; and build up data they can act on to guide the next round of decisions.

How did they get there? They leaned heavily on creators. The content bloggers and influencers posted on social platforms brought them exposure, engagement, and business.

Which raises the question.

How do you shape the creative? Who do you partner with? What do you publish?

The traditional way: hold a meeting, make the call, and rely on the experience of senior people.

Sounds reasonable. In essence, it's guessing.

That's nobody's fault. The human brain genuinely cannot process that much information.

But here's the thing: the data has been there all along.

Creators' posts, blogs, emails, web pages... how many places does a brand run into its users every day? Every touchpoint leaves behind data about preferences.

The data isn't what's missing. What's missing is the ability to turn data into judgment.

Creativity can run on inspiration. Judgment cannot run on guesses.

What Did That AI System Actually Do?

The company that handled this for them is called RAD AI. Its CEO is Jeremy Barnett. Strip the system down and its playbook comes in three steps.

Step one: pre-campaign analysis.

Pull the data. Post-level data from Instagram, Reddit, TikTok, and YouTube, drawn from more than 600 data sources, layered with ten years of historical data.

Looking for what? The creators who genuinely care about your venue, your show, your hotel.

Note the phrase "genuinely care." Not the biggest follower count, not the friendliest quote, but a creator whose content and interests actually line up with your business.

Once this step is done, creative concepts, partner picks, audience profiles, and the overall strategy all have evidence behind them.

Step two: distribution.

Content isn't a fire-and-forget exercise. Those 397 pieces went out one by one to the matching micro-communities: hotel lovers saw the hotels, show fans saw the shows.

Step three: the review.

How each piece performed gets analyzed and fed back into the model. The next round runs sharper.

What's it like? A navigation app. Whichever route gets you there fastest, it remembers—and the next person benefits. The more it's used, the smarter it gets.

The traditional playbook? Firing a rocket. Once it's lit, all you can do is pray.

Why Is Guessing Working Less and Less?

Some will say experience still has value, that a master's intuition can be dead-on.

True. It does.

But entertainment moves far too fast. Trends arrive quickly and vanish faster. Using last month's intuition to make this month's content is, more often than not, a mismatch.

And guessing your way through decisions carries an even bigger cost: risk.

If you don't know why something worked, you can't repeat it; if you don't know why something flopped, you'll step on the same rake next time.

Go data-driven instead—strategy built on real behavior—and every dollar of budget goes out on solid ground.

That's exactly what gave this entertainment group the nerve to chase 3x ROI in two months.

No—"nerve" isn't quite the right word.

That's not nerve. That's certainty.

A Final Word

This case is from 2022—not exactly new.

But the lesson hasn't aged a day—if anything, it cuts deeper today: the more widespread AI becomes, the higher the price of guessing. When everyone else is making calls with data and you're still going on feel, it's no longer a question of doing well or badly. It's a question of whether you get to keep playing at all.

Next time you plan a content campaign, don't rush to ask "does this creative look good?"

Ask one question first: does this judgment have data underneath it?

Creativity sets your ceiling. Judgment decides whether you survive.

Here's wishing that at your next content planning meeting, there's a little less "I think" and a lot more "the data shows."