AI: Probably (not) the best creative in the world

A recent catch up with arguably one of the most creative b2b leaders I know reminded me of an old Carlsberg campaign from Saatchi & Saatchi, which was new to him.

Back in 2008, during an FA Trophy match between AFC Wimbledon and Torquay United, someone in the crowd heard an announcement over the tannoy. A supporter’s wife was apparently calling to say she’d given birth to a healthy baby boy. Lovely news, obviously. But then came the small print: she was happy to name the child after Wimbledon players, would arrange an au pair from Sweden, and would get a taxi home from the hospital on the following Tuesday so her husband wouldn’t have to miss Wimbledon’s next home game against Bromley.

At which point, the announcement became rather less believable, and considerably more entertaining.

It was a set-up, of course, created for Carlsberg by Saatchi & Saatchi. Carlsberg were sponsoring the FA Trophy that season, so the match provided a natural setting for the campaign. The announcement was captured on video and subsequently found its way onto YouTube, allowing the idea to travel well beyond the supporters who were there that day.

If you haven’t seen it, it’s worth a look. It’s a lovely bit of work, and a reminder of something I think we could all do with a little more of in modern marketing.

It is of its time, it was nearly 20 years ago, so perhaps this tone wouldn’t work in 2026, but the principle that holds is the creative use of the medium for the message.

Someone looked at a football ground and saw an advertising channel that most people wouldn't have given a second thought.

The tannoy wasn’t new technology. It wasn’t an exciting new media format that needed a six-month strategy, a specialist agency and a deck containing the words “disruptive”, “immersive” and “ecosystem”. It was already there, doing what tannoys do: announcing substitutions, reminding people not to leave their cars in the car park overnight, and occasionally telling someone that their Ford Fiesta lights were on.

Saatchi simply asked a different question. What if, instead of using the tannoy to interrupt the experience, we made the announcement itself the experience?

That’s lateral thinking. Not necessarily complicated, not particularly expensive, but clever because it connected a brand, an audience, a setting and a human insight in a way that felt entirely natural. The supporters weren’t being asked to stop watching the football and pay attention to an advert. They were watching the football when something unexpected happened, right in the middle of their afternoon.

And this was 2008. YouTube existed, of course, but the original audience for the idea was the few thousand people at the ground on the day. The video capture gave it another life, but the original thinking wasn’t dependent on an algorithm finding it an audience. It had already found one.

That distinction feels increasingly important in a marketing world where we seem to have become very good at choosing our channels and rather less adventurous about challenging what a channel might actually be.

We have social platforms, paid social, organic social, influencers, video, podcasts, newsletters and an ever-growing list of places where marketing can be distributed. We have sophisticated tools to target people, measure their behaviour and optimise our campaigns. We can tell you the best time to post, which format is likely to generate the most engagement, and how many different versions of a creative asset we ought to test.

All useful things. I’m not suggesting we return to the days of throwing a poster at a wall and hoping for the best.

But there’s a danger that we end up treating the available channels as a menu of finite options. We decide we need a campaign, open the familiar media-plan spreadsheet, and start filling in the boxes. LinkedIn? Tick. Paid social? Tick. Video? Tick. Perhaps a webinar if we're feeling particularly adventurous.

Somewhere along the way, we risk forgetting to ask whether there's a more interesting way to reach people in the first place.

Now add AI to the mix.

AI can help us research an audience, develop messaging, produce content, generate variations and get things out of the door considerably faster. Used well, it can be a genuinely useful addition to the process. I use it, and I’m not particularly interested in pretending we can put the genie back in the bottle.

But AI systems learn patterns from existing material and generate outputs based on those patterns. They can combine things in interesting ways, and the results can sometimes be surprising or novel. So I’m not going to make the lazy argument that AI can never produce anything new. My concern is more about what happens to our own thinking when the easiest thing to do is ask a machine what we should do next.

Particularly when everyone else has access to much the same machine.

The Carlsberg idea didn’t need a machine to tell it that football supporters attend matches, that grounds have public-address systems, or that people respond to unexpected announcements. Those facts were available to anyone. The clever bit was making the connection between them, then having the nerve to turn it into something real.

Perhaps an AI tool could help someone develop that idea today. Perhaps it could even suggest something similar. But I wonder whether, as the tools become more capable and convenient, we might become a little less inclined to go looking for inspiration in the odd places where some of the best ideas have traditionally been found.

What if the best channel for your campaign isn’t a social platform at all? What if it’s a queue, a football ground, a reception desk, a customer service interaction, a product label, a delivery box, or some other overlooked moment in somebody’s day? What if the opportunity isn’t to create another piece of content, but to do something that people will want to talk about afterwards?

And perhaps just as importantly, are we still prepared to entertain the ideas that initially sound completely ridiculous?

Because I can imagine someone pitching the Carlsberg idea in a meeting today. “Right, we're sponsoring the FA Trophy, so we're going to make an announcement about a fictional wife having a baby over the tannoy at Wimbledon.”

I can also imagine someone asking how many impressions it will generate, whether it can be repurposed into six social assets, what the projected engagement rate is and whether we can run an A/B test.

All perfectly reasonable questions, up to a point. But sometimes a good idea needs a little room to be ridiculous before we start measuring it.

The answer isn’t to abandon evidence or stop using AI. It’s to make sure we haven’t become so focused on the efficiency of producing and distributing marketing that we forget to leave room for the unexpected.

I suspect the best marketers of the next few years won’t necessarily be those who are best at prompting AI, or who can produce the most content in the shortest time. They’ll be the ones who know when to use it, when to challenge it, and when to put the laptop down because the interesting thought is somewhere else entirely.

And I think there's a lesson here for the people who commission marketing, too. If every brief demands a predictable outcome, every idea has to fit an established channel, and every recommendation needs to be supported by evidence from something that's already worked, we shouldn't be surprised when the work feels rather familiar.

Sometimes we need to give people permission to suggest something that sounds daft.

Because Carlsberg’s campaign wasn’t memorable because it used the latest technology. It was memorable because someone looked at an ordinary situation and imagined something unexpected happening in it. YouTube helped the idea travel, but it didn’t create the idea in the first place.

In a world where AI can help us make more things, more quickly, perhaps one of the most valuable things we can do is spend a little more time wondering whether we should be making something else entirely.

Or putting it somewhere nobody has thought to look.

And if the suggestion sounds completely bonkers, perhaps don't dismiss it quite so quickly.

After all, someone once suggested advertising beer over a football ground tannoy by pretending a bloke's wife had just given birth.

And it was probably one of the best ideas in the world.