Who Benefits?
A Case for Leaning In and Building Relationships Again
The brand-customer relationship is the only compounding asset you have. You’ve spent twenty years allowing it to be disintermediated and to who’s benefit?
What gets built in the act of a consumer choosing something — taste, identity, the sense that you know what you want and why — doesn’t survive delegation. An agent can find what you asked for. It cannot build the conviction that made you ask. Neuroscience tells us the desire for agency in our own decisions is biological, not something that should be optimized away. People value what they participated in choosing. They remember what cost them effort. They form identity through the things they seek out, not things that simply arrive. The entire direction of commerce to remove friction, compress choice, and delegate the last mile of judgment is building toward a future the brain would rather reject.1
Agentic commerce isn’t a new disease. It’s the terminal stage of an old one: the platform finally cutting out the brand-customer relationship entirely, doing what platforms always do, on a faster timeline. The velocity is different but the pattern is the pattern.
The Thing About Patterns Is, They Repeat
Hotels gave their customer relationship to Booking and Expedia and now pay 15-30% per booking forever. Musicians gave theirs to Spotify and now earn fractions of a cent per stream while the platform decides who gets seen. Restaurants gave theirs to Uber Eats and watched the platform earn more per delivery than the kitchen that made the food the platform is supposed to “support.” Twenty years of consumer brands surrendered theirs to Meta and Google and built businesses that collapsed the moment the attribution stopped working.
DTC rose to become the darling of business models. The whole movement was sold to investors and operators and ourselves as building a direct relationship with the consumer, cutting out the middleman, owning the customer. What most companies actually built was a Shopify storefront acquiring traffic from Meta, retaining it through Klaviyo, fulfilling through a 3PL, and calling the result a relationship. When the paid acquisition tap got expensive, the relationship turned out to be exactly as deep as the next ad impression. Saying DTC had anything to do with a good consumer experience and a tighter relationship with customers is malpractice at this point.
Serious people inside rooms with smart slides and confident projections. Different industries. Different models. Same bad decisions. Each choice looked rational, and was certainly defencable, there is some strength in numbers. Now they are a case study (and not the good kind).
A Dirty Little Secret
The people inside most companies don’t have time to develop a point of view of their own about what’s happening to commerce. Twenty years of defending brand against a system that keeps finding new ways to devalue it doesn’t give one a lot of time to think deeply and even less political capital to leverage. The turf wars are constant, everything has the urgency of an ongoing freight train, and the last three rounds of churn, swirl and fights over the dashboard make looking for the answers in a neat little deck feel like the right answer.
The problem is, most of the people paid to develop a point of view for brands — the agencies, the consultancies, the platform partners with the readiness audit — sell the same one to everyone else in the category (and often beyond). The deck on your desk this week is a sales solution not a strategy. And, your competitors have the same playbook. So does the brand two categories over. The differences are cosmetic.
In one of my operator roles, the fights over who “owns the narrative” included product managers, product designers, product marketing, brand marketing, product research, marketing research, and comms. Add an overlay of a business line, a segment, or a market and it gets tribal quick. Everyone owning a different channel, each measured on a different dashboard, each defending a different forecast. We were, as we say in tech, shipping the org chart. The customer was meaningfully absent. She appeared in the deck we showed the board, and in the press release, and did not qualify in the room where the actual decisions got made.
The Hard Work
What comes next is not a technology question and not a marketing question. It is a question about the customer’s experience of you — every part of it, from the moment she first hears your name to the moment she decides whether to come back. If the platform stack of the last twenty years removed the customer from most of those discussions and moments and replaced her with a conversion event. The work now is putting the consumer back in the frame. Not because it’s nice, because it is the compounding asset available to you. For now.
If that is the question, what is the answer:
First, become findable on terms she set, not the agent. Most encounters with brands have been outsourced to algorithms like search, social feed, recommendation engines, and now agents. The smart investment is in (re)building the surfaces people choose to engage with. The catalogue good enough to show your friend. The retail moment where wandering is welcome. None of these scale the way a paid channel does, and that’s the point. They produce a customer who came looking and that is the only kind the agent cannot intercept.
Second, make learning who you are part of the experience. Customers don’t encounter your brand once and decide. They learn you over time through the product, the packaging, the way you handle a problem, the things you stand for and refuse. Most consumer brands have collapsed that learning into a 6-second ad and a “optimized” checkout flow, leaving no on-ramp. The brands consumers come back to give people a place to go deeper whether that’s a story that unfolds, a magazine that teaches, a founder who is worthy. The customer learning you is the experience.
Third, stop renting the customer relationship. Wherever a platform sits between you and the person buying your product, the platform owns the compounding asset and you own the unit sale. That trade was tolerable when the platform was one of several. It is not tolerable now that the agent is collapsing all of them into a single intermediary you cannot negotiate with. Direct infrastructure including owned channels, first-party data, and the ability to actually talk to your customers without a layer above you is no longer a nice-to-have. It is a prerequisite for existance. If you do not own the relationship, you cannot improve it. You can only pay more to access it.
Fourth, design for what she will remember. Kahneman’s peak-end rule is the most underused finding in consumer research. We do not remember experiences as averages. We remember peaks and endings. The post-purchase moment when the customer opens the box, when the product arrives at her door, when she calls customer service and gets a real person are the ending the brand actually controls, and most brands have defunded these moments. Most brands don’t send a follow-up that isn’t a discount code. We need to stop seeing these as nostalgic gestures. They are the moments she will use to decide whether you are a brand or a transaction. An agent can calculate price and utility. It cannot calculate the memory of a vellum insert. Pick three peak-or-end moments. Fund them properly. Measure them on a horizon longer than a quarter.
Fifth, build trust into what the customer encounters. Most consumer brands market trust they have not built into the product. Customers can tell. The agent definitely can. Costco spends almost nothing on traditional advertising and renews 93% of its memberships because trust is structural embedded in the $4.99 rotisserie chicken that's been a loss leader for decades, the return policy, the Kirkland Signature discipline, every operational decision signals benefit not extraction. The brand does not have trust; trust is what the brand is. Every interaction confirms it. The work is auditing every touch a customer has with you and asking whether it actually delivers on what your marketing claims, because if it doesn't, you are spending money to remind people you can't be trusted.
Sixth, find the brands you should have been working with all along. Your customer is not yours alone. They move between brands, trust certain industries more than others, certain certifications more than logos. The will trust certain coalitions more than individual claims. The brands that compound past the agent are may need to build shared infrastructure their competitors had to participate in. The agent does not respect your brand boundary; it respects the trust layer. Most of your competitors will not do this work because it requires sitting in rooms with companies they think of as rivals. Do it anyway because its actually the agent that’s the rival and one you can’t outspend. With infrastructure, you have to participate in shaping it or you will just be shaped by it.
The Hardest Work
Nothing is possible without conviction, and conviction is the thing the system has spent twenty years selecting against. Binet and Field’s 60/40 has been the most cited and least followed finding in marketing for a decade. The reason isn’t that nobody believes it. It’s that brand investment doesn’t show up this quarter and performance supposedly does. The people deciding the budget are measured in three month sprints. The system selects for the shorter horizon every time and people keep making the call the system rewards them for making. Overriding that requires the CEO and the CFO and the board to agree, in advance, that the next year or two matter more than the next earnings call. There is no other way to get there. The CMO cannot make this decision alone. She has tried.
The companies that survive the agent will be the ones whose customers came looking. That is the only outcome the platform stack cannot route around. They came looking because somewhere along the way you gave them a reason to. Something they encountered, something they learned, something they remembered, something they trusted. Not because you optimized for them but because you paid attention to them.
Twelve essays, one argument: discovery isn’t finding, and the brands worth finding are the ones who know the difference.
Footnotes
Leotti, Iyengar & Ochsner (2010) established the desire for control over one’s environment as a biological imperative. Norton, Mochon & Ariely’s work on the IKEA effect demonstrated that labour increases valuation. Kahneman’s peak-end rule — that we judge experiences by their most intense moment and their ending, not their average — remains the most underused finding in consumer strategy.