From Customer Journeys to Customer States

How AI Is Rewriting the Logic of Digital Marketing

For decades, marketing and experience design have depended on the same conceptual model of the consumer journey. A linear journey of consumer awareness, interest, desire, conversation, and loyalty as basic stages with slight variations in different models.

These models have shaped how organisations plan campaigns, design consumer experiences, allocate budgets, and measure success. Entire operating models—from funnel reporting to martech stacks—have been built around it.

However, marketers have always understood that human behaviour is not linear. It is situational, inconsistent, and highly sensitive to context.  People's interactions are influenced by their moods, levels of urgency, beliefs, social input, willingness to take risks, etc. And these factors change all the time.

Journey models stayed around not because they were right, but because they were useful.

Artificial intelligence changes this dynamic. 

Why journey models worked anyway.

There is a lot of writing on the problems with linear journey models in both academic and professional journals.

Herbert Simon's theory of bounded rationality says that human decision-making is limited by our finite cognitive abilities, incomplete information, and time constraints. We often choose "good enough" (satisfying) options over optimal ones because our minds use mental shortcuts to deal with complexity.  

Marketing experts are aware that customers rarely follow the paths that companies plan for them. But until now, they did not have many options. It was not possible to design experiences that could adjust in real time to so many contextual variables. Data was broken up, the signals came in slow, and decision-making was within a rigid framework that companies followed for many years. So, businesses avoided complexity by segmenting users, looking at their average behaviour and planning the "most likely" paths. 

The goal was to simplify a structure in which execution was practical.

Journey frameworks worked not because they correctly described behaviours but because they aligned organisations.

Journey maps are coordination tools that help teams agree on priorities, responsibilities, and investments, even when they oversimplify customer reality.

From stages to states: a different unit of design

People do not enter digital experiences in “consideration”. They enter in states.

They may be:

• uncertain, but curious.

• informed but sceptical

• rushed and task-focused

• confident and ready to act

These states are shaped by signals across time and channels like browsing habits, recency, repeat tasks, language patterns, social influence, device context, and past outcomes.

Until recently, these states were largely invisible to systems.

AI changes that.

What AI actually makes possible

AI doesn't really "get" users in the way that people do. It makes a more specific and important contribution. 

AI systems can: 

• Aggregate signals across channels and time

• Figure out likely states based on patterns of user behaviour and keep those guesses up-to- date. 

• Continuously choose a response on the fly instead of choosing the next stage on the linear funnel.

The difference is not where the users are in a journey, but how the system interprets their state.

Consider a simple example. Two people arrive on the same product page. Both are technically at the same point in a traditional journey.

One has visited the site multiple times, read detailed documentation, and returned late in the evening. The system infers confidence and urgency and responds by surfacing pricing, implementation details, and a clear next step.

The other arrives for the first time from a comparison article, spends time scrolling, and hesitates before engaging. The system infers uncertainty and risk, and responds with customer proof, explanations, and reassurance.

	Visitor 1	Visitor 2
Entry point	Returns directly to the product page	Arrives from a comparison article
Prior behaviour	Multiple previous visits	First visit
Content consumed	Documentation, feature details	Scrolls overview, hesitates
Timing & context	Late evening	Midday
Inferred state	Confident, high intent, time-bound.	Uncertain, risk-aware, exploratory
What the system prioritises	Speed and clarity	Reassurance and explanation
Response surfaced	Pricing, implementation steps, clear CTA	Customer proof, comparisons, FAQs
Journey outcome	Accelerated decision	Exploration
Designed journey	Same	Same
Experienced journey	Different	Different

No predefined journey has changed. What changes is the response.

The implication is subtle but profound: customers no longer need to be moved through journeys. Journeys emerge from how systems interpret signals and respond to human states in real time.

When Journeys are just a side effect.

The journey is no longer the goal of design in a state-based model.

The goal of design is now to: 

• Detect State 

• Response Logic 

• Decision support

This is the main point. Therefore, I will clarify the difference between rational journey thinking and outcome-driven thinking, even if it may seem like an oversimplification to some readers.

Traditional (input-driven) thinking	State-based (outcome-driven) thinking
You start by defining: Step 1: Awareness Step 2: Consideration Step 3: Purchase This path is an input: "This is the journey customers should take." Everything else is built to enforce that sequence.	You start by defining: signals you care about states you want to infer responses that are appropriate for each state You do not define the path: "The path is simply the sum of how the system responded over time." That’s why the path is an outcome.

Imagine Cooking with a recipe (input). 

You follow steps in order: Chop -> Fry -> Simmer The sequence is fixed.

Cooking by taste (outcome) 

You taste continuously and adjust: add salt, lower heat, wait longer. The final dish emerges from many small adjustments, not a fixed sequence.

Companies don't plan the road their customers should take; instead, they plan how the system should react. The path takes shape on its own.

Evidence from practice

Consumer platforms

A useful example is Netflix. Netflix does not segment users into awareness or consideration. It continuously infers states based on viewing behaviour, time of day, recency, and device context—and adjusts recommendations and presentation accordingly. The experience changed not because the user “progressed”, but because their inferred state did.

Amazon also adapts experiences based on signals that identify people in a hurry, people sensitive to price, or people who like to compare prices. Some users are presented reassurance (reviews, guarantees), others are given speed (one-click buy), and still others are provided alternatives. The system responds to context, not stage.

B2B Environments

In B2B environments, this pattern is emerging more quietly.

Leading products now change how they onboard, guide, and message users based on things like setup speed, feature exploration, time gaps between sessions, support interactions, etc. The person who isn't sure is reassured and given more information. A person who is sure of themselves gets acceleration. The "journey" is different without being planned out in detail.

How this affects digital marketing teams

The shift from journeys to states is not just a conceptual one. It changes what digital marketing teams are actually accountable for.

If customer paths emerge from how systems interpret context and respond in real time, then the primary task of digital marketing is no longer sequencing touchpoints. It is designing responses.

That means deciding:

• which signals matter.

• Which customer states are worth detecting

• which responses reduce friction, risk, or hesitation

• and how systems should behave when signals are ambiguous or conflicting

Journeys do not disappear in this model. They still exist—but as outcomes rather than plans. This approach raises a practical question that most marketing organisations are not yet equipped to answer:

If journeys are no longer the thing we design, what replaces the funnel as an operating model?

Answering that question requires moving beyond theory and into practice—rethinking how teams structure campaigns, content, measurement, and decisioning around customer states rather than stages.

That is the focus of Part 2.