Marketing Stack

Chapters

Chapter 1: The Death of the Funnel

Chapter 1: The Death of the Funnel

The marketing funnel — awareness, consideration, decision, purchase — was a useful model for sixty years. It described, with reasonable accuracy, the process by which a potential customer moved from ignorance to transaction. The funnel assumed that the marketer controlled the top (awareness), influenced the middle (consideration), and closed at the bottom (decision). The customer moved through the funnel in one direction: down.

The funnel is dead.

It has been dying for a decade, and the cause of death is the same cause that has disrupted every linear model in the information age: the customer no longer moves in one direction. The customer enters the process at any point. They become aware of a product through a peer recommendation in a Slack channel, skip consideration entirely, and purchase on impulse because a colleague mentioned it in a meeting that morning. They research for months, consult twelve sources, and then purchase a competitor’s product because the checkout flow was faster. They purchase, regret, return, repurchase, and become an evangelist — a cycle the funnel does not describe and cannot accommodate.

The funnel was a model for a world in which the marketer controlled the information environment. That world is gone. The information environment is now controlled by the customer, or more precisely, by the algorithms and agents that mediate between the customer and the information. These mediators do not respect the funnel’s linear logic.


What Killed It

What Killed It

Three forces converged to make the funnel unworkable.

The first force is information abundance. When a prospect can access every piece of information about your product, your competitors, and your category within minutes, the “awareness” and “consideration” stages collapse into a single event. The prospect does not move through stages. The prospect arrives already informed, or becomes informed so rapidly that the stages are indistinguishable.

The second force is channel fragmentation. A B2B buyer in 2020 encountered your brand through perhaps five channels: your website, LinkedIn, email, an industry event, and a peer recommendation. A B2B buyer in 2027 encounters your brand through thirty or forty touchpoints, many of which you do not control and cannot track. The fragmentation makes the funnel’s sequential logic impossible to apply — the buyer is not “in” any stage, because the buyer is simultaneously in all stages across different channels.

The third force is AI mediation. The buyer increasingly encounters your product not through your marketing, but through AI systems that have processed your marketing and synthesized it into a recommendation, a comparison, or an answer. The AI system does not move the buyer through a funnel. The AI system delivers the buyer directly to a decision, having compressed the funnel into a single inference.

These three forces — abundance, fragmentation, and mediation — have not weakened the funnel. They have made it structurally incoherent. You cannot run a linear model on a non-linear system.


What Replaces It

What Replaces It

The replacement is not a new shape. The marketing industry has proposed many shapes: the flywheel, the loop, the infinity sign, the helix. These are geometric metaphors that provide the satisfying feeling of having solved the problem without actually solving it. The problem is not the shape. The problem is the assumption that the buyer’s journey has a shape at all — that it is a path with a beginning and an end, traversed in a describable order.

The replacement is a system: a set of autonomous agents, operating continuously, adapting in real time, that meet the buyer wherever they are and serve them whatever they need. The system does not move the buyer through stages. The system responds to the buyer’s current state and adjusts its behavior accordingly.

This distinction matters more than it appears to. A stage-based model tells the marketer: “identify which stage the buyer is in, and deliver the appropriate content for that stage.” A state-based model tells the marketer: “observe the buyer’s current signals, and let the system respond in real time to the pattern those signals describe.”

The stage-based model requires the marketer to classify the buyer. The state-based model requires the system to read the buyer. The difference is the difference between a filing cabinet and a conversation.


The Implications

The Implications

If the funnel is dead, then everything built on the funnel must be rebuilt. This includes:

Marketing organization design. Most marketing teams are organized around funnel stages: a brand team (top of funnel), a demand generation team (middle of funnel), and a sales enablement team (bottom of funnel). This organization is a fossil. It creates handoff points that the buyer does not experience and does not respect. The buyer’s experience is continuous. The marketing organization must be continuous.

Measurement and reporting. Funnel metrics — MQLs, SQLs, conversion rates between stages — are the language of marketing accountability. They are also increasingly fictional. An MQL is a lead that has been classified as “in the middle of the funnel,” but if the funnel does not exist, then an MQL is a lead that has been classified into a category that does not correspond to any real state. This is not a semantic objection. It is an operational one. If your measurement system is measuring a fiction, your decisions are based on a fiction.

Technology architecture. The marketing stack of the funnel era was organized around the funnel: top-of-funnel tools (ad platforms, social media management), mid-funnel tools (marketing automation, nurture systems), and bottom-of-funnel tools (CRM, sales engagement). These categories persist in the analyst reports. They no longer describe how the technology actually operates — or should operate.

Budget allocation. Funnel-based budgeting allocates resources to stages: X percent to awareness, Y percent to consideration, Z percent to conversion. If the stages are fictional, the allocation is arbitrary. It may accidentally correlate with effective spending. It may not. The correlation is not earned — it is luck.


The Transition

The Transition

The transition from funnel-based to system-based marketing is not instant. It does not require burning down the existing stack and starting over. It requires a gradual shift in three areas:

First, a shift in mental model: from “moving buyers through stages” to “responding to buyer states.” This shift costs nothing and changes everything, because it changes what questions the team asks.

Second, a shift in technology: from stage-organized tools to agent-organized systems. This shift is expensive but increasingly inevitable, as the tool vendors themselves are moving toward agent-based architectures.

Third, a shift in measurement: from stage-conversion metrics to state-change metrics. This shift is the hardest, because it requires the marketing team to give up the clean, single-number metrics that the C-suite understands and replace them with richer, more accurate, and more difficult-to-explain measurements.

The rest of this book describes the system that replaces the funnel: its agents, its channels, its measurement, its logic, and its requirements. The funnel is dead. What replaces it is more complex, more capable, and more interesting.

That is not a consolation. It is an opportunity.