For the last twenty years, digital marketing has been organised around a fairly simple idea: either people know what they want, or they do not. If they know what they want, we call it intent. They search. They compare. They click. They land on a product page. They convert. If they do not know what they want, we call it inspiration. They scroll. They browse. They watch. They save things for later. They form taste, desire, familiarity and trust.
This distinction became so embedded that we built entire marketing disciplines around it. Search ads, SEO and performance marketing were built around pull: capture existing demand at the moment someone expresses it. Social, display, influencer, brand advertising and content were built around push: create or shape demand before the customer can articulate it.
But this split was not just a marketing truth. It was a product constraint. The old web lacked the intelligence to understand intent as it formed, and the interfaces to serve it dynamically.
A static page could inspire someone, explain a category, tell a brand story, or show a carefully merchandised edit. A search box could help someone find a thing they already knew how to name. A product grid could narrow the catalogue by category, price, rating, brand or size. A product page could persuade someone to buy one item. But no single interface could really move with the customer as their intent formed.
So marketing had to compensate. We pushed inspiration into one part of the system, and intent capture into another. Brand advertising was asked to make people care before they knew what to look for. Search was asked to catch them once they did. SEO was asked to create thousands of static pages for all the things people might type. Product pages were asked to close the deal. Retargeting was asked to stitch the journey back together after it had already fragmented.
That is the old web’s hidden architecture: not just pages and channels, but a forced separation between different stages of human decision-making. And it is starting to break.
People do not actually move from inspiration to intent in a straight line
The problem is that real people do not behave like marketing funnels. A person does not wake up neatly in “awareness”, progress cleanly to “consideration”, and then arrive politely at “conversion”. They drift. They half-know what they want. They change their mind. They discover a new constraint. They compare options. They need reassurance. They want taste, expertise, social proof, price, evidence and emotional confidence all mixed together.
Take a wellness retail example. There is a huge difference between a customer searching “Vitamin D3 4000 IU K2 capsules” and a customer thinking: “I’m a female founder in my early 40s, I feel like I’m doing a lot of the right things but I’m tired, my HRV has changed, I care about longevity, skin, energy and sleep, and I want to understand what kind of supplement stack might actually make sense for me.”
The first query is intent. The second is not just inspiration, and it is not just search. It is an unresolved decision. The customer is trying to understand the category, understand themselves, narrow the problem, learn what matters, avoid wasting money, and eventually choose products.
On the old web, those two needs are served by completely different systems. The exact Vitamin D query goes to site search, Google Shopping, SEO pages, product grids and PDPs. The broader longevity-stack question goes to blog content, TikTok, Reddit, podcasts, newsletters, influencers, maybe a quiz, maybe ChatGPT, maybe ten open tabs. The retailer’s site might have all the relevant products. It might even have good content. But the interface cannot easily hold the journey. So the customer leaves the commercial environment to do the real thinking somewhere else. That is the opportunity.
Marketing has been shaped around what interfaces could serve
We often talk about marketing channels as though they emerged naturally from user behaviour. But a lot of them emerged from interface limitations.
Search advertising works because a search results page is good at serving already-expressed intent. It can match a query to a set of results. It can rank. It can sort-of filter. It can take someone from “I need X” to “here are the best available Xs”. Social advertising works because feeds are good at creating appetite before intent exists. They are visual, passive, emotional and low commitment. They are not asking the user to define their need. They are showing them possibilities.
SEO works because static pages can be made for repeatable query patterns: “best magnesium for sleep”, “collagen for skin elasticity”, “creatine for women”, “vitamin D dose UK”, “NMN vs NR”, “electrolytes for fasting”. Brand marketing works because people need meaning, trust and memory before they choose, and product grids are bad at creating those things.
The marketing stack reflects the UX stack. Push exists because the customer cannot yet search. Pull exists because the customer finally can. But the middle is where most decisions actually form.
That middle is awkwardly served today. It is split across blog posts, buying guides, quizzes, comparison pages, product filters, customer reviews, social proof, creator content, Reddit threads and search results. Each asset answers one small part of the decision. None of them adapts to the person’s evolving intent.
The old interface types are too rigid
The classic ecommerce journey has a few dominant shapes. There is the static content page: an article, guide, landing page or campaign page. It is good for explanation and storytelling, but it is usually the same for everyone. At best, it is loosely personalised to a segment or category. There is the search results page: good when the user knows what to type, weaker when the user’s need is vague, emotional, multi-factor or educational. There is the product grid: good for scanning inventory, bad for making sense of it. There is the product page: good for validating a specific item, bad for helping a customer understand whether that item is the right choice in the broader context of their life. There is the quiz: useful in some categories, but usually brittle. It asks the brand’s pre-written questions, not the customer’s real question.
None of these are wrong. They are just incomplete. They assume the user’s state is fixed. The article assumes the user is learning. The grid assumes the user is choosing. The PDP assumes the user is validating. The checkout assumes the user is done. But modern customers are often doing all of these at once.
A wellness customer might need to know whether magnesium glycinate is meaningfully different from magnesium threonate, whether a supplement conflicts with something else they take, whether it is appropriate for women in their 40s, whether the dose is sensible, whether the price is good, whether the brand is credible, whether the reviews mention sleep or stress, and whether this should be bought alongside Vitamin D, creatine, omega-3 or collagen.
That is not a product grid problem. It is not a blog problem either. It is an interface problem.
AI collapses the boundary between inspiration and intent
AI changes the shape of the web because it can interpret vague human intent before that intent has hardened into keywords. A user no longer has to begin with the perfect query. They can start with: “I’m in my early 40s, very busy, trying to optimise energy, sleep and longevity without taking 25 things. Where should I start?” Or: “I keep seeing creatine, collagen, magnesium and NMN everywhere. Which of these are actually relevant to me?” Or: “I want a simple supplement routine under £100/month that supports energy, sleep and healthy ageing. I already take Vitamin D and I’m worried about cholesterol.”
These are not traditional ecommerce searches. They are conversations around need, context, trade-offs and decision-making. The experience that answers them should not be a static article followed by a product grid. It should be built around the customer’s intent as it forms.
It should educate, but only as much as needed. It should inspire, but not float away from the catalogue. It should compare, but not reduce everything to a table. It should recommend, but explain why. It should let the user move from broad curiosity to a concrete decision without changing channel.
That is the big shift: the future interface is not inspiration-first or intent-first. It is intent-forming. AI-native experiences can finally hold the messy middle: inspiration, education, clarification, comparison and decision-making in one adaptive flow.
The new journey is generated around the user’s question
Imagine a wellness retailer experience that starts with the broader question: “What kind of longevity supplement stack should I consider as a female founder in my early 40s?” The old web would struggle with this. A search box might return blog posts. A product grid might keyword-match “longevity”. A quiz might force the customer into generic health goals. A buying guide might be useful but static.
An AI-native experience can do something different. It can recognise that the user is not ready for a product grid yet. It can create a structured journey around the decision.
First, it might clarify the goal: energy, sleep, stress, skin, metabolic health, healthy ageing, or all of the above. Then it might explain the main supplement categories in plain language: foundational nutrients, sleep and stress support, performance and muscle, skin and collagen support, longevity-positioned products, and products where evidence or suitability needs more care. Then it might build a simple “starter stack”, “performance stack” and “more advanced longevity stack”, each with rationale, price range, dose considerations and product options.
Then it might let the user ask follow-ups: keep this under £80/month; avoid anything too experimental; I already take Vitamin D; show me options with the fewest capsules; what would I take in the morning vs evening; which of these are most relevant if sleep is my biggest issue?
At some point, the customer’s intent crystallises. They are no longer researching “wellness”. They are deciding between three specific products, understanding the dose, checking the monthly price, and deciding whether to add them to basket. The same interface should be able to travel with them all the way there.
That is the difference between a website that contains answers and an experience that helps someone decide.
This changes what brands need to build
The commercial opportunity is not simply “better search”. Better search matters, especially for large catalogues. But this is bigger than search. Search assumes the user can name what they want. The next generation of interfaces will help the user work out what they want.
That means brands need experiences that combine four things that used to live separately: inspiration, education, intent crystallisation and decision-making. In other words: what are the possibilities, what should I understand before choosing, which of these possibilities actually fits me, and what should I buy, book, read, try or do next?
Today, those moments are usually spread across different channels, pages and teams. The future experience brings them into one adaptive journey.
For retail, that might mean moving from vague need to curated basket. For travel, from “special anniversary trip” to three personalised itineraries. For publishing, from “I want to understand this topic” to a guided reading path. For property, from “family home near good schools and a commute to the City” to a shortlist with trade-offs made visible. For wellness, from “I want to feel better and age well” to a sensible, explainable, brand-safe product routine.
The interface becomes the bridge between human uncertainty and commercial action.
The brands that win will not just be findable
For years, the goal was to be discoverable. Rank on Google. Show up in search. Win the auction. Get into the feed. Retarget the visitor. Optimise the landing page. That still matters. But it is no longer enough.
As AI becomes a bigger part of discovery, brands will face a new problem. Customers may arrive with more context, more comparison and less patience for static journeys. They may have already asked an AI assistant to summarise the category. They may already have a shortlist. Or they may expect the brand’s own site to behave with the same intelligence.
In that world, the brand’s website cannot just be a product catalogue with some content wrapped around it. It needs to become a decision environment. One that can handle vague intent. One that can educate without overwhelming. One that can personalise around the customer without flattening the brand. One that can move from inspiration to intent to conversion in a single flow.
The old web split those things apart because it had to. The AI-native web will put them back together.