Why Generative AI Makes The Brand-vs.-Performance Divide Even More Pointless

The rivalry between brand and performance has always been absurd. It is even more so in the age of AI search and GEO (Generative Engine Optimization).

Brand is not the opposite of performance. It is the cognitive and emotional infrastructure that makes performance possible. When the two are aligned, brand plants the seed, while performance nurtures it and harvests the results.

The mistake, I believe, comes from a basic confusion: We call “performance” what can be measured immediately, rather than everything that creates value, whatever the time horizon.

As Les Binet and Peter Field, two leading British experts in this field, have shown¹, we need to distinguish between two complementary dynamics:

• Sales activation targets people who are close to buying. It produces fast, measurable effects, but they are often short-lived.
• Brand building works on awareness, familiarity, preference, and mental associations. Its effects are slower, but they last longer and play a more structural role in business growth.

According to their research, the most effective brands devote, on average, around 60% of their efforts to brand building and around 40% to sales activation. Naturally, the exact split depends on the industry, the maturity of the brand, the length of the buying cycle, and other more or less structural factors.

The most effective strategies combine short- and long-term effects instead of setting them against each other. Without brand, performance eventually ends up optimizing a shrinking pool of demand. Measuring only what moves quickly is therefore not the same thing as creating lasting value.

In that regard, Nielsen estimates, based on its own experience base, that a one-point gain in brand metrics such as awareness or consideration is associated, on average, with a 1% increase in sales. Even though it is intangible, brand is therefore an asset in the same way, for example, that a distribution network is an asset. It influences sales volumes, pricing power, and customer acquisition costs.

Brand building plays an even more important role in the age of generative AI. It feeds not only human preference, but also the algorithmic interpretation of generative search engines. Brand becomes an interface between two forms of cognition: Human cognition, rooted in memory, emotion, and trust; and algorithmic “cognition,”² fed by explanations, evidence, and recurrence.

With generative search, we are not simply witnessing a change in human behavior. We are also seeing some of that behavior delegated to algorithms. As a result, the change in shopping is no longer only about where commerce happens: First in physical stores, then on websites, then in apps. It is now about who, or what, performs the act of buying, as artificial intelligence can act on behalf of Internet users.

In traditional marketing, the funnel represents a sequential journey: A person discovers a brand, starts considering it, becomes interested, compares options, forms a purchase intent, buys, and may then become loyal to the brand or even recommend it to others.

With generative search, users no longer necessarily move through these stages one by one. In a single prompt, they can ask to identify brands, understand the criteria that matter, compare available options, and receive help making a decision. What used to be spread across several stages of the marketing funnel can now be handled in a single answer from a generative search tool. The funnel is being compressed.

This means that the stages of the funnel need to be translated into content families that AI can access and surface in one or several answers. In this context, the challenge for brands is no longer simply to produce content for each stage of a supposedly linear customer journey. It is to build a coherent content ecosystem capable of answering all information needs at once: Understanding, comparison, choice, purchase, and usage.

In this new model, each content family plays a specific role:

• Identification content helps AI understand what makes a brand distinctive, what it offers, and which category it belongs to.
• Credibility content provides proof: Customer reviews, references, use cases, rankings, comparisons, and third-party mentions, including media coverage.
• Explanatory content answers users’ practical questions.
• Decision-support content makes it easier to compare several options.
• Transactional content makes purchase possible.
• Support content extends the relationship after purchase.
• Recommendation content makes a brand easier for third parties to cite, legitimize, and amplify.

Where the marketing funnel relied on persuading prospects step by step, the new model requires all content families to be available at the same time, because AI may need them all in a single response to a user’s prompt.

Incidentally, this brings communications closer to revenue generation. Not because communications becomes a sales function, but because its content can now directly influence the answers that shape purchase decisions.

For years, companies have opposed brand building and performance marketing. That opposition has always been sterile. After all, if performance measures the decisive act – the purchase -, brand is often what creates the intent behind it. It is the intangible asset that prevents performance from becoming a permanent auction.

Generative AI does not merely add a new channel to marketing and communications. It changes how brands are discovered, compared, and bought.

In this context, setting brand and performance against each other is not only absurd. It is dangerous.

—

¹ In their seminal book “The Long and the Short of It.”

² I am well aware that this phrasing anthropomorphizes artificial intelligence.