When the Customer is a Machine

The Agentic Marketing Revolution briefing mapped the structural shift. This analysis addresses the question the briefing deliberately did not resolve. If an AI agent selects against a matrix of structured criteria, what survives of brand equity, storytelling, and emotional resonance?

The short answer is that brand survives. What it has to become is something else entirely. Under the logic of machine-mediated purchasing, brand is no longer a creative asset expressed through narrative and emotional territory. It is a set of verified signals that lower the agent's uncertainty score during selection. The change is not cosmetic. It is a re-architecture of what brand equity is measuring and where it is stored.

For a C-suite reading the quarterly P&L, the question is no longer whether to invest in brand. It is whether the brand investment currently on the marketing line is being deposited into assets that an agent can actually read.

How a machine actually buys

The traditional purchase funnel assumed a human decision-maker exposed to a branded sequence. Awareness, consideration, decision. The machine-mediated journey operates on a different logic entirely. A consumer delegates a task. An agent interprets the instruction, queries available data, evaluates options against user-defined constraints, and executes. The brand has one opportunity to be selected, and that opportunity is almost entirely determined before the agent begins reasoning.

The implication is that the majority of what classical brand marketing was designed to influence, namely the emotional stumble between awareness and purchase, does not occur. There is no serendipity to optimise. There is no impulse to capture. There is only the agent's prior, formed from what it can find, parse, and compare.

Agents do not stumble

Human buyers make emotional mistakes. They encounter a brand at a moment of vulnerability, form an attachment against their own stated preferences, and rationalise the decision afterwards. An entire discipline of marketing was built around engineering those moments. The adjacency, the atmosphere, the unexpected encounter that shifts consideration.

Agents do not stumble. They evaluate. The moment a consumer's preference is encoded as a constraint in a prompt, the agent's reasoning becomes mechanical. If the instruction is "book a flight, prefer Airline A unless more than one thousand dollars costlier," the agent does not develop an attachment to Airline A. It treats the preference as a variable and resolves it against cost. Brand preference still matters, but only to the extent the consumer has chosen to encode it. Beyond that encoding, the agent's logic governs the outcome.

Storytelling migrates into a different substrate

Lindsay Boyajian Hagan, Vice President of Marketing and Co-Head of Revenue at Conductor, has spent the last three years rebuilding an enterprise AEO platform around this shift. Her framing is that storytelling is not disappearing. It is migrating into a different substrate.

Her operating principle is that brand narrative must now be encoded as topical authority. The brand that is cited by AI engines on a given subject is the brand the agent will treat as the default expert. That citation preference is built through depth, specificity, and genuine subject-matter mastery published at scale. Generic AI-generated filler, what the industry has started calling slop, does not earn citations. Machine-readability is necessary but not sufficient. The citation has to be earned.

Topical authority is the new storytelling. The brand's equity is no longer held primarily in recall and emotional association. It is held in the density and credibility of its machine-readable expertise across every topic the brand wants to own. The content marketer is no longer the drafter. The content marketer is the curator of the brand's verified subject-matter signal.

The commercial consequence is immediate. The brand investment that produces machine-readable authority is additive to enterprise value because it compounds across every agent query that follows. The brand investment that produces emotional territory without structured translation is a depreciating asset the moment the journey becomes machine-mediated.

Marketing to the agents themselves

Tony Marlow, Chief Marketing Officer of LG Ad Solutions, frames the selection logic directly.

Optimisation for machine consumption is not an extension of search engine optimisation. It is a new discipline with higher stakes and narrower consideration sets. Traditional SEO competed for ten blue links on page one. Agent-mediated selection competes for inclusion in a consideration set the buyer never sees, generated from the agent's training data and real-time crawl. A brand absent from that training data, or present but lacking the structured information an agent needs to compare it against alternatives, is excluded by default. Not deprioritised. Excluded.

Andrew Bialecki, Co-CEO and Co-Founder of Klaviyo, extends the logic to its architectural conclusion. His vision of the autonomous CRM envisages a near-term state in which AI search engines do not crawl a brand's static website. They query the brand's own agent directly, expecting a parseable, real-time response tailored to the query and the consumer behind it.

Every brand will need an addressable agent that can answer questions, represent the business to both humans and other AI, and produce a consistent, up-to-date response at machine speed. The website becomes a legacy artefact. The brand's real competitive surface moves to the agent-to-agent interface, and the readiness of the data layer beneath it becomes the single most consequential investment a CMO can protect.

The one asset that cannot be replicated overnight

This is a structural argument, not a romantic one. The agent evaluates against constraints the consumer has encoded. Those constraints are not neutral. They are shaped by the prior relationship the consumer has formed with brands before the prompt is written. Trust, belief in the company's direction, and community affiliation are the upstream inputs that determine which brands make it into the prompt as preferences and which are left to be evaluated purely on specification.

The implication is not that emotional brand equity is dead. The implication is that emotional brand equity now has exactly one job. To survive long enough in the consumer's mind to be written into the instruction. After that, the agent takes over.

Human assets and their machine equivalents

Every line item on the current brand P&L should be evaluated against a single test. Does this asset produce a signal an agent can read, parse, and prefer? The translation below is not a marketing exercise. It is a capital reallocation question.

Assets that fail the test are not worthless. They are upstream inputs to the consumer prompt, which is a narrower and more specific role than they have historically been asked to play.

Stories or signals

A brand held primarily as stories is a brand whose equity depreciates every quarter that agent-mediated purchasing expands. A brand held as verified signals, topical authority, structured metadata, citation density, addressable agents, is a brand whose equity compounds with every agent query that resolves in its favour. The delta between those two trajectories is the real measure of whether the brand budget is a capital investment or a liquidating position.

The CMO who cannot produce a signal-level audit of their brand assets by the next board meeting is presenting a risk the CFO has not yet been told about. Invisibility to machines is not a marketing problem. It is an existential one, and the quarter-over-quarter cost of remaining invisible compounds silently until it shows up in the top line.

The brands that will be preferred by agents five years from now are being encoded into the training data right now. The CMOs who understand this are treating the next twelve months as the infrastructure window. The ones who do not are assuming they will catch up later, at a cost that will not be available to them when they attempt to pay it.

Contributing Practitioners

The voices behind this piece

This analysis is distilled from long-form interviews conducted on The Business of Marketing podcast with four senior operators shaping how brands show up in machine-mediated commerce. Companion analysis to The Agentic Marketing Revolution executive briefing.

Tony Marlow

Chief Marketing Officer

LG Ad Solutions

Lindsay Boyajian Hagan

VP Marketing & Co-Head of Revenue

Conductor

Andrew Bialecki

Co-CEO & Co-Founder

Klaviyo

Ruslan Tovbulatov

Chief Marketing Officer

Gloat