The following is an excerpt from soon-to-be-published "Competing at Machine Speed," a new book by Interpublic alum Jeriad Zoghby,
Roald van Wyk, and Joan Smith about what the agentic era portends for modern marketing.
The Last Free Click
Just as the fragmentation became unmanageable,
generative AI search arrived and went after the one thing brands still got for free: organic traffic.
Consider how discovery used to work for a brand with a decent digital presence.
A consumer types “best protein powder” into a search engine. The engine returns a page of results: organic listings for the brand’s site, its product pages on various retailers,
reviews and articles that mention it. The brand invested in search optimization to be in those results, and over time that investment paid out in free traffic, visitors who found the brand without the
brand paying for each click.
advertisement
advertisement
Generative AI changes the mechanic. The model reads the top results, writes an answer, and hands it back inside the interface. The consumer reads the
answer and moves on. No click. No visit. No attribution. An answer without a destination, a recommendation without a referral.
This is the zero-click journey, and it is accelerating.
When the model builds that answer, it reads what it can parse: product descriptions, ingredient lists, review summaries, attribute tables, comparison data. The content that earns a place in the answer
is not the most beautifully written. It is the most machine-readable. Complete structured data. Clear attributes. Specifications organized well enough that a model can cite them with confidence.
So the brands that poured their investment into polished, human-facing creative, the photography, the narrative copy, the aspirational imagery, find that the model cannot read most of
what they built. It needs data, not poetry. Structured attributes, not emotional prose. The content that won the award show is not the content that wins the machine.
That is a change
in what brand content is for. And it does not stop at the search box.
The Buy Button Has No Budget
If there is one place where all three
disruptions converge, it is the Product Detail Page (PDP), where the consumer has split shopping from buying, the retailer has become a media company, and the path has shattered into walled
gardens.
The PDP is where the algorithmic journey ends with the click of the buy button. It is where organic results lead, where sponsored placements link, where creator content
sends the click, where the review summary points. For a consumer who has crossed from awareness to intent, it is the moment of conversion. It is also, routinely, the most neglected asset in the
portfolio.
If you follow the spending, often the largest share of the media budget goes to traditional media, such as television production and the big-idea campaign, much of it
agency fees. A reasonable share goes to digital advertising, usually a separate agency and a separate creative cost. Another share goes to the media agencies handling search bidding, social
placements, and programmatic. Finally, a small portion goes to an e-commerce agency or in-house team that deploys assets, usually built for other channels, to the PDP.
The page that
all of that spending exists to drive traffic to, which decides whether the consumer converts or not, with the most important button in the history of commerce on it, is the last page to be attended
to. The same page is often not refreshed for months, if not years, while the media agency reports strong return on ad spend (ROAS) and impressive click-through rates on their campaigns.
And nothing about how the work is organized is built to close that gap. It is split across separate agencies and internal teams, each working to its own brief, its own metrics, its own
calendar, none of them accountable for what the page actually earns. So the page stays under-invested, under-maintained, and the media spend climbs every year against an operation that cannot keep
up.
And the human is no longer the only one reading the page. Retailer algorithms rank it on completeness. Agentic shopping tools, working on a consumer’s behalf, read its
attributes mathematically. A page that performs beautifully for a person, rich imagery, a compelling story, can be close to invisible to a model building a recommendation, because the structured data
underneath it is thin. The page is now the interface between the brand and both the consumer and the machines acting for the consumer.
The most consequential asset in the portfolio
is the last one to be prioritized.
I watched a CEO discover that in real time. He ran a shelf-stable food brand with decades of heritage and a fiercely loyal following. In our
conversation about how his products were surfacing online, I pulled up the actual search results. In a store, a shopper knew his products were food for people because of where they sat. Online, there
was no aisle to say so, so his products were turning up beside pet food and missing entirely from some of the searches real customers were typing.
The fix was almost insultingly
small. His team had to add one word to the page, “human,” so the engine would understand the product was food for people and surface it where shoppers were actually looking. Without that
small clarity written on the page, he would not appear at all. What unsettled him was not the word. It was that a page he had treated as routine content was quietly deciding whether he made the sale
at all. Smaller competitors had already worked this out and were catching the searches he was missing. They were taking share from him not on product or brand but on understanding how people shop when
the aisle is gone. He was leaving for a board off-site that week and told me afterward it was the only thing the board discussed.
It is tempting to call this a content challenge.
Produce better content for more channels, and the problem goes away. It does not. The real challenge is that each of these commerce media ecosystems comes with its own algorithm and requirements, each
changing without notice, which already exceeds what most brands can manage cost effectively.
The gap between what the landscape demands and what the current operating models can
deliver widens every quarter.