The next marketing advantage: Engineering demand, not just capturing it.

I think marketing has become a little too good at optimizing the bottom of the funnel.

Over the past decade, we’ve built incredibly sophisticated capabilities to identify high-intent audiences, personalize communications, optimize media and convert customers. But somewhere along the way, the ability to measure short-term performance started disproportionately influencing where we invest, often at the expense of creating future demand.

This isn’t a new argument. Binet and Field have demonstrated the importance of balancing long-term brand building with short-term activation for years. More recently, the Profit Ability 2 study found that 58% of advertising-generated profit occurred after the first 13 weeks, reinforcing how much value we miss when we focus primarily on immediate returns.

58%

of advertising-generated profit occurred after the first 13 weeks

You can only optimize existing demand so far before you need to create more of it.

And two developments in AI are about to make that distinction considerably more important.

The platforms are taking over performance marketing

Look at what’s happening with Google and Meta. Both are rapidly integrating AI into their advertising platforms, automating targeting, bidding, creative development and campaign optimization.

There are real advantages. AI can identify patterns and optimize campaigns at a scale no human team could realistically manage. But as more decisions move inside these platforms, brands risk losing visibility into where their money goes, which audiences they’re reaching and what’s actually generating incremental business.

There’s also a competitive question worth asking: If everyone has access to increasingly similar optimization technology, how much sustainable advantage can any one brand build through performance execution alone?

Performance marketing isn’t going away. It will become more effective, but also increasingly standardized and controlled by platforms that brands don’t own.

What happens when AI becomes the shopper?

The second development could be even more disruptive.

Agentic commerce has the potential to change how consumers discover, evaluate and purchase products. Rather than navigating search results, advertisements and comparison sites themselves, consumers may (err will) increasingly delegate parts of that process to AI.

Google’s introduction of its Universal Commerce Protocol in January 2026 is one indication of where this is heading. The infrastructure is developing quickly, even if widespread adoption takes longer.

Imagine someone asking an AI agent to find the best laptop for their needs. The agent could compare specifications, reviews, pricing and availability, potentially bypassing several opportunities for brands to influence the decision through traditional lower-funnel marketing.

But what if that person already has a strong preference for Apple? That preference enters the conversation before the agent begins evaluating options. The agent might still recommend something else, but the brand has already influenced the decision.

Of course, consumers may also develop preferences through conversations with AI itself, creating another challenge for marketers: understanding not only how people perceive their brands, but how AI systems interpret and recommend them.

The more purchasing decisions are mediated by AI, the more valuable it will become to establish preference before the transaction begins.

Engineering demand creation

Put these developments together and I think we’re approaching an important shift.

Technology platforms are gaining greater control over the mechanics of demand capture, while AI agents could increasingly influence how consumers evaluate and purchase. Both developments have the potential to reduce the control brands traditionally exercised at the bottom of the funnel.

That makes demand creation strategically more important. But I don’t think the answer is simply spending more on brand advertising.

A brand is shaped by its product, service, experience, reputation and communications. Creating demand means deliberately building the conditions that make customers more likely to choose you. Capturing demand means being present, relevant and easy to buy from when they’re ready.

Both matter, but they’re often managed by different teams, measured against different objectives and competing for the same investment.

As AI reshapes marketing and commerce, organizations need to connect these activities more deliberately. Understand what creates incremental demand, invest in the experiences and associations that build preference, and ensure the business can capture the value when customers are ready to act.

We’ve spent years engineering increasingly sophisticated ways to capture demand. Perhaps it’s time we applied the same discipline to creating it.