Most marketing departments run on one mental model: advertising persuades people to buy, so you should measure whether it drove a sale. In this interview, conducted between our co-founder Coen Olde Olthof andJames Hurman, author of The Case for Creativity and Future Demand and co-founder of brand tracking platform Tracksuit, Hurman explains why that model creates a performance trap. He argues that only some advertising converts current demand, and the rest must create future demand among people who are not in the market yet. He also covers why the 50/50 split keeps turning up in the data, why retail media is brilliant but incomplete, and why AI agents won't replace brand choice.

Key Takeaways:

• The performance trap: If you believe advertising works by persuading people to buy, you move money to whatever shows immediate sales. Base sales then decline, and overall marketing ROI falls.
• Two jobs, not one: Converting current demand (people in market now) and creating future demand (people who'll buy next year) are equally important, and need different tactics and different metrics.
• The 50/50 pattern: Studies from JWT (1992), Binet and Field, Google, Analytic Partners and BCG keep landing on a rough balance of brand and performance. Whether 60/40 or 53/47 matters less than the balance.
• Creativity is for the not-yet-in-market: Functional ads work for people ready to buy, while creativity earns the attention of everyone else.
• AI automates performance, not brand: AI can churn out functional ads and targeting, but brand work should remain human.

What is future demand?

Future demand is the demand from people who are not yet in the market for your category but will be later. Creating it means making your brand familiar and emotionally positive to them now, so that they are biased toward you when they finally buy. It sits alongside current demand, the people ready to buy today, whom performance marketing converts.

From proving creativity pays to building future demand

James Hurman's first book set out to answer a question clients had every right to ask: is creative work actually more effective? Fifteen years later, his second book asks what creativity is for. We spoke to him from Auckland, which is about as far from Amsterdam as you can get …

In 2011 you set out to prove creativity pays. What changed by the time you wrote Future Demand?

"When I started the project, probably around 2007 or 2008, we had a group of clients who really believed in creativity and a bigger group who weren't sure. I thought that was a fair question. If you're spending millions with an agency and it suggests highly creative work, is that the right thing to do? Some clients were taken by the rhetoric, and others just needed numbers. So I collected every study I could find, from academic institutions and industry alike. All of them said that more creative work tends to be more effective. I didn't find a single study that said the reverse."

Future Demand grew out of a different audience. Five or six years ago, working with startups and venture capital firms, Hurman met people who did not believe in brand at all. "Future demand was a way of explaining why brand matters to non-marketers. There are lots of people who aren't in market yet but will be your customers next year. If you want to win them, start marketing to them now, in a more creative, more emotional way than you do your performance. That biases them towards you when they enter the market."

The Performance Trap

What is the performance trap, and what does it look like from inside a marketing department?

"If you believe marketing is a persuasion mechanism, you believe the way to measure it is whether it drove a sale, more or less immediately. Then you look across your activities. Some show immediate sales, some don't. If you're a logical person, you put the money into what appears to be working and take it away from the rest. But that other stuff was doing a different job, one that was harder to count but really did count."

"Your base sales go down, even though your performance marketing looks great. The more you lean into short-term results, the lower your overall marketing ROI. It's a very logical trap for anyone who thinks advertising exists to drive sales immediately."

How do you pitch brand to a VC who needs traction in two and a half years?

"I love that it's a challenging environment. It often requires reframing. Their ultimate goal is an exit, not 50% growth next month. It doesn't matter if you grow 50% a month for two years and then fail. So first, you're playing a long game. Over that time, lots of customers will enter your market and choose you or a competitor. If your competitor has made their brand familiar and warm, their performance marketing will work harder than yours. It's about market share, and how much of the future value of the category you need to win to get that big exit."

"It starts like most things in business, by painting a picture of where the most money is. You work back from how to build the most valuable company, rather than forward from how to make an amazing ad."

Retail media: Genius or the Next Trap?

Retail media is growing fast. Is it the next performance trap?

"It's a trap if viewed in the wrong way. I actually think retail media is genius. If you're browsing an aisle, there's a high chance you're in the market that day. Converting current demand is as important as creating future demand, and retail media does it better than anything, giving people the facts: what it costs, what it does, why it's better."

"But it's only one side of the coin. If another brand has built theirs, their retail media works harder than yours. They'll nick the sale from you in the aisle. The biggest conversion rates are in retail media because you're speaking exclusively to people in the category. That rate will go down and down if you forget to invest in your brand."

Here, Coen compared it to last-click attribution in physical form, while Hurman ran with it: "When I walk to the beer aisle I always walk past the oranges. Some clever person could say the oranges caused the beer. That's what last-click attribution is."

Where the 50/50 Split Comes From

How confident are you in the 50/50 split?

"More and more. Les Binet and Peter Field's 60/40 was an excellent piece of work and a revelation for the industry. But the pattern shows up elsewhere. In 1992, Peter Kim at JWT used a large US data set to show that brands spending about half on brand advertising had the best ROI. Last year, Google studied e-commerce retailers, Analytic Partners studied global brands, and BCG studied a bunch of brands, and all found the same thing."

"In marketing science we talk about generalizable patterns, ones that show up in different data sets, markets and periods. This is one. Whether it's 60/40, 50/50 or 53/47 doesn't matter much. These two things help each other, and when held in balance you get the best return. If you only do performance, you leave tons of money on the table."

Creativity's Job: Earning Attention from People Who Aren't Buying

Is the link between creativity and effectiveness as strong as when you wrote The Case for Creativity?

"My view hasn't changed, but it's more precise. Then, I thought of advertising as one thing. Now I think of it as two. For converting current demand, you almost don't need creativity. Pretty functional, well-branded information in front of people in the market works. Adding creativity, more media spend or more channels doesn't lift it much."

"Creativity is critical for people not yet in the market. If you're a phone brand talking to someone who won't buy until next year, they won't even notice your advertising. Our brains filter it out. Unless it earns your attention, there's no chance you look at it. That's why all of us, who will never buy a commercial truck, watched Jean-Claude Van Damme do the splits between two reversing trucks. The creativity earned our interest. That's my case for creativity now: it's how you earn the attention of people who will buy from you next year."

AI, Agencies and the Agentic Shopper

What should agencies stop doing, and what should they do more of?

"AI is going to be fantastic at automating pretty much everything we think of as performance marketing: churning out functional ads, placing them, targeting them. We should let AI do all of it. On the brand side, AI will automate almost none of what we think of as brand marketing, because it's about creativity and eliciting emotion, a much more human domain. If I led a creative agency, I'd automate the performance work, not even let my creatives see it, and focus on very, very good brand work."

"The one thing agencies have not been good at is educating clients. At Colenso, half my job was making sure clients walked in wanting original, engaging ideas and knowing why it was good for their business. It's time-consuming and annoying, but if you've got clients who walk in wanting fantastic work, it's much easier to deliver it."

Will AI agents make our buying decisions, so that brands must optimise for "model availability"?

"At Tracksuit we acquired an AI visibility company, so we can show customers how they perform among people and among models. It matters. But the idea that agents will make our purchase decisions for us is nonsense, for four reasons."

"One: lots of people enjoy shopping, and we won't hand that to an AI. Two: even the most aggressive consultancy forecasts say roughly 10 to 14% of purchases will be agentic, so around 86 to 90% of sales next year will happen the old-fashioned way. Three: look at our own behaviour. We use an LLM for research, then Google the options ourselves. Four: when people are surveyed, they say they'd use agents for utility bills and grocery shopping. Those are brand choices already made, and the agent just repeats them."

"We'll still default to brands we're familiar with and feel good about. That's how the human brain evolved to decide. It won't change in our lifetimes."

Measuring Future Demand for the CFO

What argument works with a CFO, and how do you measure this?

"Start with the framework: some people are in market today, and a lot of people will come in later. Performance metrics, conversion rate and cost of acquisition, are excellent for current demand and useless for future demand, because you're measuring people who aren't buying yet. So park them and establish what they're good for and terrible for."

"For future demand, you need brand tracking of awareness, consideration and preference. To get to return on investment, you need either long-running incrementality testing or a good marketing mix model. Once there's agreement on that split, the conversation about measuring brand over the long term gets simpler."

What's the most common mistake in brand tracking?

"People don't know what to do with the data. They try to correlate a 1% uplift in awareness with sales. It's not about that. It's about diagnosis: how strong is your brand compared to your competitors, where is it strong, where is it weak, and are you converting consideration to preference at the same rate as your category? That's where we're heading at Tracksuit, and alongside making it about a tenth of the traditional cost and always on, like performance dashboards."

Lightning Round

If a busy Dutch CMO reads only part of the book, which chapters? "Can I have three? I didn't want to write one of those business books where it's all in the first chapter and then repeats itself for twelve more. The first three chapters take about fifteen minutes."

One belief the industry should drop by 2030? "That advertising works by persuading people to buy. Performance ads don't persuade, they divert sales that are already happening toward you. Brand marketing doesn't cause sales either, it builds a bias toward your brand. AIDA was invented by a salesman to explain how sales work, not advertising. It's the wrong mental model."

Favourite effective campaign of recent years? "Heinz Ketchup from Rethink in Toronto: a great long-term platform, incredible creative, amazing business results. And AXA's 'Three Words' campaign in France on domestic violence. A beautiful, smart idea."

Most underrated metric? "Pricing power."

Which is harder, brand or performance? "Brand, by far. Performance is easy. We've become really good at it. Great brand work takes more effort, talent, time and money, but the payoff is much bigger."

‍

Watch the full interview here: