Question time? We're getting better at getting it wrong
To get the right answer, you have to ask the right question and, as the old adage goes, 'to assume can make an ass out of u and me'. So, asks art director Syed Shuja Hussain, if AI can make an assumption look like certainty, who’s brave enough to reopen the brief?
)
I’m one small part of this industry, working on the art side. Much of my career has unfolded across global agency networks, alongside people who taught me to see things differently. From my corner of the table, I’ve noticed something amusing.
We’re exceptionally good at finding answers. Give us a business problem and we’ll return with consumer insights, cultural tensions, behavioural patterns and a presentation explaining why everything makes perfect sense. Usually around slide 27.
AI can make the wrong question look brilliantly answered. The uncomfortable part is what happens next.
Someone calls it a breakthrough. Someone else asks whether we can make the headline bigger. The client nods. The agency nods. Everyone leaves feeling considerably more intelligent. Whether anything changes is another conversation.
Now, AI can help us research, explore creative directions and assemble polished strategies before the coffee gets cold. Cool. I’m in. Except AI can make the wrong question look brilliantly answered. The uncomfortable part is what happens next, because a convincing answer can it harder to challenge.
Above: "Give us a business problem and we’ll return with consumer insights, cultural tensions, behavioural patterns and a presentation explaining why everything makes perfect sense. Usually around slide 27."
Imagine a client asking us to make its brand more appealing to younger consumers. AI helps analyse competitors, develop audience profiles and suggest creative territories. By lunchtime, we could have a campaign. By afternoon, possibly a case-study film. But perhaps younger people aren’t rejecting the original advertising. Perhaps they’re rejecting a product that takes seven steps to buy, or an experience that still treats them like their parents.
A campaign might be the right intervention, but it might also be a beautifully produced invitation to help discover the same frustration. Yet, the more convincing the proposed campaign becomes, the less welcome an investigation can feel. That’s the danger: confidence in the answer can become confidence in the assumption that the question beneath it is the right question to ask.
The more convincing the proposed campaign becomes, the less welcome an investigation can feel.
Working on the art side makes this easy to recognise. Composition, colour and craft can make an argument feel convincing before we’ve proved it. AI brings that persuasive power to the thinking itself. The finish can conceal the fault, and a beautifully finished fault is surprisingly difficult to send back. Someone has approved the strategy. Someone has developed the idea. Someone has booked the shoot. Revisiting the original question now means reopening decisions that everyone thought were settled.
Yoda might suggest unlearning what we’ve learned. Procurement would ask whether it’s included in the retainer.
Above: The wisdom of Yoda would suggest we unlearn what we've already learned.
There’s a debate about whether today’s transformation is technological or psychological. I suspect the two are rather busy changing each other. Technology changes expectations; expectations change how technology gets used. Either way, explaining that shift doesn’t tell us whether we’ve chosen a useful response to it.
Our agency processes don’t always make that decision easy to revisit. Research produces findings. Strategy develops the proposition. Creativity creates the idea. Production makes it happen. Effectiveness eventually tells everyone whether it worked. A beautifully organised relay race.
Each department inherits the confidence of the department before it. What began as “we think” quietly becomes “we know”.
Except a question raised during research can become an unquestioned instruction by the time it reaches production. Each department inherits the confidence of the department before it. What began as “we think” quietly becomes “we know”. Sweet. The assumption now has a budget.
This isn’t an argument for endless debate. Ideas need decisions, deadlines and people willing to commit. But commitment should come with a clear understanding of what we know, what we’re assuming and what would make us reconsider. Otherwise, momentum becomes its own justification. The brief must be right because look how much work we’ve done.
Above: You can't have endless debate. Ideas need decisions, but decisions need to be made with a clear understanding of what's being asked.
I’ve been fortunate to learn from a few Jedi Masters. Aunindo Sen, Executive Creative Director at Publicis Middle East, is one of them. He taught me that great thinking begins with recognising the questions that deserve challenging. An approved brief deserves that attention, too. Vicky Kriplani, Business Lead at Leo Burnett Dubai, helped me appreciate something equally important: brilliant creative thinking must create value beyond the creative department.
Together, those lessons give me a useful standard: challenge the starting point, then make the resulting idea matter outside the meeting room. That means being able to say, “This is our strongest direction, and here’s what could change our minds.” I’d trust that more than a presentation in which every slide somehow proves the previous slide was correct. Obi-Wan would recognise the technique. “These aren’t the assumptions you’re looking for.”
Challenge the starting point, then make the resulting idea matter outside the meeting room.
AI can help us examine assumptions. Alongside developing a proposal, we can ask it for competing explanations. What else might account for the behaviour? Which claim depends on information we haven’t collected? What would we expect to see if our assumption were wrong? The useful output is a sharper investigation.
A plausible response from an AI-generated customer remains a prediction. Before it becomes “consumer understanding” in our deck, we should be clear about where it came from. Back with our younger audience, I’d want to see where people abandon the buying process, hear what frustrates them and explore whether changing the experience alters their response. The scale of that investigation should suit the decision.
Above: Creativity still needs room for surprise.
Creativity still needs room for surprise. We’re trying to understand the risk we’re taking, while leaving space for an idea nobody could have predicted in a research session. Piyush Pandey put it plainly: “No campaign is worth its salt unless the people on the street love it.” That gives advertising an external judge. We can approve the work. We cannot approve the public’s response. People are wonderfully inconvenient like that.
There’s a commercial consequence here, too. As competent work becomes easier to produce, clients have less reason to pay a premium for more options, more presentations and more polished executions. It’s easy to mistake the number of options for the breadth of the thinking. Ten creative territories can still share one unexamined premise. Feed AI conventional assumptions and it can help us become conventionally excellent. Which is a slightly terrifying ambition for people paid to find a way nobody expected.
AI hasn’t made thinking cheap. It’s made predictable thinking harder to charge for.
AI hasn’t made thinking cheap. It’s made predictable thinking harder to charge for. Our value may increasingly lie in the decision we help a client change: the campaign they reconsider, the experience they improve, the opportunity they hadn’t noticed. That can create an awkward conversation about how agencies get paid. Discovering that a client needs a simpler buying process instead of a large campaign may reduce the work we expected to produce. It may also be the most valuable thing we discover.
If our business model rewards defending the original brief, we shouldn’t be surprised when that brief becomes remarkably difficult to challenge. A team that spots a faulty assumption early has made progress. A strategist who changes their recommendation after new evidence has exercised judgment. A creative who exposes a weakness before production has protected the idea.
We should make those actions easier to admit and easier to reward, because the alternative is an industry getting exceptionally good at making its first answer look inevitable. The question I’d like agencies to get comfortable with is simple: “What would make us change our minds?” Ask it while the answer can still change the work.
Next time someone asks whether we can make the headline bigger, I’ll probably say yes. I’m on the art side, after all. But first, let’s check we’re not making the mistake bigger.
