Why AI’s shiny ads don’t matter (and what really does)

Everyone’s losing their minds about AI’s creative powers in advertising — slick images, endless ad variants, branded video spots at the push of a button. But I created an experiment to prove this attention is misguided, and misses the real trick at the heart of successful campaigns: saying the right things to the right people.

I’ve long thought that too much time and effort goes into making things look nice at the expense of figuring out what you actually need to say (or do) to sell something. And with the marketing world increasingly split between “AI-slop is killing our beautiful craft” and “Bro, this one prompt replaced my entire ad agency”, I wanted to see if AI could create an ugly, low-effort campaign that beat benchmarks without me having to do the boring legwork.

So I set myself a challenge: build a bare-bones recruitment campaign for a local choir, use AI to handle the grunt work, and see if “ugly” could actually sell better than “slick”. 

Spoiler: it did. And the reason why says a lot about what AI can, and can’t, really do for marketers.

If you’re generating AI ad creative while thinking it’s a game-changer, this story is a warning for you.

A big idea for a small campaign

A local choir was struggling to get recruits, so I decided to help them out with a free recruitment campaign in exchange for being able to write about the results. 

The challenge:

• Acquire 5-10 new choir members

• In 4 weeks (or less)

• By creating a landing page and a $500 Facebook lead campaign

• Using free/low-cost no-code tools

• And as much AI help as possible

To start, I made a pretty ugly landing page:

👉 https://burlingtonwelshchoir.carrd.co

It’s built in Carrd, a gloriously minimalist no-code tool that makes you feel like a web designer from 2007 (in a good way). No animations, logos, or sexy brand fonts. Just a bunch of text, a few images, and a big red button that scrolls to a basic lead form.

Instead of focusing just on the singing, I wanted to highlight the social aspects of the choir as much as possible - I had a feeling that this was the real reason most men joined.

I used AI (ChatGPT) a lot to work faster and avoid wasting time on stuff I couldn’t be arsed to do, like creating a comprehensive FAQ section, or coming up with different headlines for each section. You can read more about the process I went through creating the landing page, including some of the actual prompts, here. 

Then it was time to create the campaign.

The ugliest campaign ever

I kept it simple:

• One Facebook campaign, one ad set

• A landing page with the bare minimum of polish

• Two ads: one aimed at lapsed singers, one focused on the social side of joining a choir

I wasn't joking when I said they were ugly. I also threw some spelling mistakes and formatting errors for good measure. This was the highest performing ad.

AI handled the grunt work of drafting ad angles, suggesting headlines, and outlining copy. My job was to sort the good from the bad, tighten the message, and make sure the ads spoke directly to the right audience.

So, what happened?

The “social” ad flopped. High CTR, but no conversions.

The “lapsed singers” ad crushed it. CTR of nearly 7%, CPC of $0.17, and 26 leads at a $10 CPL.

The campaign worked so well that I had to stop it early after the choir president called me in a panic, worried that too many new recruits would turn up on the first night.

KPI	Campaign Result	Benchmark Range	Evaluation
Cost per Lead (CPL)	$10.47	$5 – $20	Excellent — comfortably within range, on the low side for a niche campaign.
Click-Through Rate (CTR)	6.99%	1% – 3%	Outstanding — more than double the typical benchmark, showing highly engaging ads.
Cost per Click (CPC)	$0.17	$0.75 – $2.00	Exceptional — extremely efficient at driving traffic compared to norm.
Landing Page View Rate	91%	70%+	Excellent — indicates page is relevant and loads quickly, minimal drop-off.
Total Leads Generated	26	10–20 (expected for local niche)	Strong — exceeded realistic volume expectations for a small, hyper-local campaign.

6 new people showed up to the first practice, and a few more said they’d be joining future ones. So while I ended up slightly short of the 10 new members I’d hoped for, it wasn’t a bad result for a choir who hadn’t attracted a single new member in nearly a year. 

But what does this mean for the rest of us?

Design is optional, copy isn’t

No-one cared that the campaign was ugly. They cared that the ad spoke directly to them, and the landing page clearly spoke to every aspiration and concern they’d have about joining a choir.

If you’re spending time obsessing over pixel-perfect design and ignoring the words, you’ve got it backwards.

And if you’re dismissing AI because you think it only makes slop, you’re missing the point. It didn’t replace me. It handled the boring parts - drafting versions, iterating faster - so I could focus on choosing the right audience and the right message.

The big takeaway

I didn’t need 20 different videos, or a deep well of slick images. Just a single message targeted to a tight audience.

That doesn’t mean that you need to make things ugly or simple - only that having slick creative is not the key to success. 

But this campaign wasn’t really about a choir. It was a test of where marketing is heading.

AI can churn out endless images and videos, and design polish is cheaper than ever. That stuff isn’t the advantage anymore. The only edge left is taste — knowing what to keep, what to throw away, and what to double down on.

That’s the Rick Rubin approach: strip it back, focus on what matters, trust your gut. The opposite is the old way: tweaking buttons, over-producing, and flooding the platforms with ads while hoping that something will stick.

The marketers who win will be the ones who use AI like a strategic intern, not a creative director. 

So stop wasting your time generating a mountain of AI creative, and start using it to get back to basics: figuring out who you should be talking to, and what you need to say to get them to buy more stuff.

Get in loser, we’re all vibe marketing now.