Seriously, Did You Just Run a Focus Group?

Seriously?

You ran a focus group?

In 2026?

You found eight people.

Put them in a room.

Gave them some sandwiches.

Showed them your campaign.

And asked:

“Would this make you more likely to consider purchasing our product?”

Jesus Christ.

Did you fax the results back to Head Office?

Look, we love talking to customers.

Please talk to your customers.

Call them. Meet them. Watch them. Buy them lunch. Ask why they love you. Ask why they hate you. Ask why they abandoned their cart at 11:47pm after spending 25 minutes choosing a colour.

Human beings are fascinating.

But pretending a small group of people sitting in a research facility discussing an imaginary future purchase is some kind of crystal ball?

We have questions.

Mainly:

Have you seen how much data you already have?

Your Customers Have Been Talking to You for Years

They just haven’t been doing it in a focus group.

They’ve been doing it on social.

Every day.

They watch.

Scroll.

Skip.

Share.

Save.

Comment.

Click.

Buy.

Unfollow.

Come back.

Send something to a friend at 1:14am with the message:

“HAHAHAHA this is literally you.”

That behaviour is telling you something.

Over the last 11 months, maybe you tested 100 different hooks.

Some died immediately.

Some held attention.

Maybe one absolutely flew.

You tried different faces.

Different creators.

Different messages.

Different jokes.

Different formats.

Different lengths.

Different levels of weird.

Your audience has been voting on your creative every single day.

Not by raising their hand.

By actually doing something.

That distinction feels rather important.

KBD’s philosophy has always been that social data should feed creative decisions, not just sit inside a report after the campaign is over.

What People Say and What People Do Are Two Different Things

Imagine showing someone an ad in a focus group.

“What did you think?”

“Yeah, I liked it.”

Great.

“What specifically did you like?”

“It felt authentic.”

Excellent.

“Would you buy the product?”

“Potentially.”

Fuck me.

Get the media agency on the phone.

We’ve got a potentially.

But you know the one thing people almost never do in a focus group?

Get their credit card out.

Nobody says:

“Would you buy this?”

And then slides an EFTPOS machine across the table.

“Cool. Buy it.”

Because suddenly the hypothetical becomes real.

Would you actually spend $89 on it?

Would you choose it over the thing you normally buy?

Would you click?

Would you subscribe?

Would you put your email address in?

Would you hand over actual money that you worked actual hours of your finite human life to earn?

Or did you just think the ad was quite nice?

There’s a canyon between “I would buy that” and buying that.

The problem isn’t that people are lying.

Humans are just terrible little narrators of their own behaviour.

We say we want to eat healthier.

Then Uber Eats arrives.

We say advertising doesn’t influence us.

Then buy the shoes.

We say we hate stupid videos.

Then watch a golden retriever fall into a swimming pool 11 times.

What people say matters.

But what they actually do is pretty interesting too.

And social gives us a gigantic window into the second one.

They don’t have to tell us whether the hook would get their attention.

We can see if it did.

They don’t have to tell us whether they’d watch the video.

We can see if they watched it.

They don’t have to tell us whether they’d click.

We can see if they clicked.

And if you’ve connected your data properly, they don’t even have to tell us whether they’d buy.

We can see if they got the credit card out.

That feels slightly more useful than potentially.

Social Media Is One Giant, Slightly Unhinged Experiment

Every piece of content teaches you something.

That doesn’t mean:

“THIS POST GOT 300,000 VIEWS. MAKE EVERYTHING EXACTLY LIKE THIS FOREVER.”

Please don't.

That’s how brands accidentally turn one successful joke into a three-year content strategy.

But over time, patterns emerge.

Certain openings hold attention.

Certain talent consistently performs.

Certain topics make people talk.

Certain creative structures get shared.

Certain messages drive action.

Certain beautifully art-directed posts that everyone internally adored get absolutely murdered in public.

Useful.

Not emotionally.

But strategically.

You can begin developing creative from evidence instead of everybody sitting around a boardroom table saying:

“Personally, I prefer Option B.”

Thank you, Colin.

Very useful.

And Now Things Are Getting Weirder

Because social data isn’t the only thing changing how we research ideas.

AI is entering the room.

Not ChatGPT writing 47 customer personas called Marketing Mary.

We deserve better than Marketing Mary.

We’re talking about synthetic audiences.

Models built using real research, behavioural data, customer information and other inputs that can simulate how different audience groups might respond to questions, propositions and creative ideas.

Instead of recruiting a new group every single time you want to explore something, you can interrogate a synthetic audience.

Ask another question.

Change the proposition.

Test another angle.

Ask why.

Push deeper.

Do it again.

At 2am if you really want to.

The synthetic customer does not need a sandwich.

Does That Mean We Fire All the Humans?

No.

Calm down.

We like humans.

Some of our best friends are humans.

Synthetic research isn’t valuable because AI has suddenly become an infallible oracle that can tell you exactly what customers will do.

It can’t.

And treating synthetic people as perfect replacements for actual people would be just as stupid as treating eight people behind one-way glass as perfect representatives of Australia.

The interesting bit is what happens when you combine things.

Actual behavioural data from social.

What people really watched, clicked, shared and ignored.

Real customer research.

What people tell you, in their own words, about their needs, frustrations and lives.

Synthetic audiences.

A way to explore hypotheses, pressure-test thinking and ask far more questions, far faster.

Now research becomes less like a single ceremonial event before the campaign begins.

And more like something creative teams can use constantly.

That’s exciting.

Research at the Speed of Ideas

This might be the biggest change.

Traditional research can be slow.

Recruit.

Schedule.

Prepare.

Run the session.

Analyse.

Present.

Debate the presentation.

Schedule another meeting to discuss the meeting where you debated the presentation.

By then, someone on TikTok has invented an entirely new language.

Creative doesn’t work like that.

Ideas move quickly.

One question creates another.

“What if we changed the opening?”

“What if the creator said it?”

“What if we removed the product entirely?”

“What if we made the thing everyone internally hates?”

Historically, you couldn’t commission a new research project every time someone in the creative department said “what if?”

Now?

We can get closer.

That means data doesn’t have to arrive at the beginning and end of creativity.

It can sit beside it.

Not driving.

Definitely not driving.

But sitting in the passenger seat, occasionally saying:

“Hey, that thing you just assumed about 25-year-old men might be complete bullshit.”

Useful passenger.

Data Should Make Creative Braver

This is where we’ll probably upset someone.

Because the point of better research isn’t to remove risk.

It’s to let us take better risks.

There’s a nightmare version of all this where brands use AI, social data and predictive tools to optimise every creative decision until advertising becomes one enormous algorithmically generated beige sludge.

Every joke pre-tested.

Every rough edge sanded down.

Every idea statistically likely to be acceptable.

Kill us now.

Data shouldn’t tell creatives what to make.

It should give them a better understanding of the world they’re making it for.

There’s a difference.

If 11 months of social data tells us your audience consistently responds to a certain creator, that’s interesting.

If synthetic research suggests an assumption behind the brief is wrong, that’s interesting.

If customer interviews reveal something nobody inside the business has considered, very interesting.

Then creative people get to do what creative people are supposed to do:

Make the leap.

Our job isn’t to optimise our way towards the safest possible answer. KBD exists precisely because safe, committee-approved, default-setting marketing tends to produce the same beige work as everybody else.

Maybe the Focus Group Isn’t Dead

There are things a room full of actual humans can tell you that a dashboard never will.

There are things behavioural data can reveal that humans themselves can’t articulate.

And there are questions synthetic audiences may let us explore at a speed and scale that would previously have been ridiculous.

So perhaps this isn’t:

OLD RESEARCH BAD. AI GOOD.

That would be boring.

It’s about having more tools.

Better signals.

Faster feedback.

More ways to understand people before spending millions trying to get their attention.

And maybe fewer occasions where six weeks of research culminates in a PowerPoint slide informing us:

“Consumers value authenticity.”

Groundbreaking.

Your Audience Is Already Giving You Answers

So before your next campaign, look around.

Look at the last year of social.

What actually held attention?

Who did people watch?

What did they share?

What did they ignore?

What surprised you?

Look at your customer conversations.

Look at your search data.

Look at your comments.

Look at what people actually do.

Build synthetic audiences carefully and interrogate the assumptions you’re making.

Talk to real humans when real humans are what you need.

Then take all of that beautiful, messy evidence…

And make something nobody could have predicted in a focus group.

Because the future of creative research shouldn’t be about finding increasingly sophisticated ways to ask permission.

It should help us understand people well enough to do something interesting.

And if your research still ends with:

“The focus group preferred the blue one.”

Maybe keep that to yourself.