For nearly twenty years, marketers — myself included — operated with one quiet assumption behind every strategy and forecast:

Eventually, people land on your website.

The journey might be messy. The touchpoints might be chaotic. But sooner or later, the click brought everything together. That moment tied the story back to the funnel we all trusted.

Then 2025 happened.

The assumption didn’t wobble.It broke.

And in 2026, the pace of that shift isn’t slowing. It’s speeding up.

I’m not writing this as a futurist or an analyst looking down on the industry. I’m writing as a marketing executive trying to make sense of this while actually running a business, with revenue targets, dashboards, and a team relying on clarity that suddenly isn’t there.

Every operator I talk to is feeling the same pressure.

Branded search behaves unpredictably.Organic traffic doesn’t map cleanly to demand.Buyers show up already educated in ways analytics can’t explain.Attribution behaves more like a dimmer than an on/off switch.Traffic patterns drift further away from anything we used to call normal.

Everyone senses the shift.But very few people have had the language for it.

So this playbook is not a claim of certainty.It’s my attempt to map what’s happening based on the patterns that are now impossible to ignore.

Here’s the simplest frame:

**The funnel isn’t dead.

It moved.**

It moved upstream, into AI systems that now handle discovery, education, comparison, and vendor selection long before someone ever opens a browser tab or creates a measurable session.

We’ve entered a world where:

  • journeys are nonlinear

  • visibility is partial

  • attribution captures only fragments

  • and most of the buyer’s thinking happens outside our analytics

And if you’re wired like I am — someone who loves numbers, precision, and tidy explanations — here’s the uncomfortable truth:

We have to rely more on directional signal than perfect attribution.Perfect attribution wasn’t ever as perfect as it looked (there, I said it).

Across nearly every category, the same trendline shows up:

Organic click-through rates are down.AI-native search is rising fast.Zero-click outcomes keep climbing.Total discovery expands while website sessions shrink.

The funnel didn’t vanish.It slipped into places our dashboards were never designed to measure.

AI answers, conversational interfaces, and OS-level assistants now handle a huge portion of what we used to call “top-of-funnel.”

Buyers are forming their understanding inside:

  • ChatGPT

  • Perplexity

  • Claude

  • Gemini

  • Siri

  • browser-native answer layers

And here’s the twist that most people overlook:

Your website still matters.In a lot of ways, it matters more than ever.It’s just no longer the first stop.

Your site is the validation layer…the place buyers go to make sure the AI-generated picture is accurate, trustworthy, and aligned with what they need.

If I had to compress the 2026 buyer journey into one line:

Problem → education → shortlist → pricing → alternatives → recommendation → maybe one click

Or, in the words of Bob Dylan (special nod to those of us still spinning vinyl)

“The times they are a-changin’.”

And the funnel changed right along with them.

What follows is what acquisition actually looks like now.

01. AI Answer Presence: Your New Homepage

For twenty years, Google was the front door.Today, AI systems are the lobby, the guide, and the first conversation.

The first impression of your brand is no longer your homepage.It’s the answer someone sees before they ever think to click anything.

AI Answer Presence = how often your brand appears in AI-generated responses to the highest-intent tasks in your category.

These moments matter most:

  • “best X for Y”

  • alternatives and competitor lists

  • vendor comparisons

  • pricing context

  • category explainers

  • how-to workflows

  • implementation guidance

Unlike traditional SEO, answer presence is:

  • cross-platform

  • entity-driven

  • conversational

  • and often intentionally zero-click

The part most teams underestimate

A single AI prompt does not trigger a single lookup.

LLMs routinely fire two or three separate retrieval paths to cross-check information across the open web. Each micro-query samples different sources the model trusts.

This means your brand isn’t evaluated once.It’s evaluated multiple times, in rapid succession.

If your reviews, pricing narrative, entity structure, and offsite footprint don’t align across those retrievals?

You don’t lose ranking.You disappear from the answer layer.

New industry data reinforces this

This week (yesterday, to be exact), James Cadwallader (Co-Founder at Profound) shared research showing that 89 percent of ChatGPT prompts trigger two to three web searches by the underlying User Agents.

One question becomes several retrieval events…and your brand must survive all of them.

James put it plainly:

“How your website appears in those retrieval moments determines whether your content is even eligible for the final answer.”

This is the new visibility rule:

**AI doesn’t reward the “best” answer.

It rewards the most consistently corroborated one.**

Why it matters

LLMs sample across:

  • reviews

  • press articles

  • G2 / Capterra

  • Reddit threads

  • YouTube summaries

  • comparison pages

  • product documentation

  • directory listings

If your story fractures across these surfaces, you’re filtered out before the buyer ever sees you.

This is why offsite corroboration has become one of the strongest drivers of answer-layer visibility.

If you’re not in the answer layer, the buyer won’t skip your website.They’ll skip your brand altogether.

02. LLM Referrals and Brand Lift: The Only Clicks That Matter

LLMs don’t behave like Google, but they influence demand through two categories of signals.

A. Direct signals (rare but highly valuable)

When they appear, direct referrers show:

  • the click originated inside an AI environment

  • your brand appeared in the generated answer

  • intent was high enough to break the zero-click flow

B. Indirect signals (the overwhelming majority)

Most AI-influenced users never click within the answer.Instead, they:

  • Google your brand

  • type your domain directly

  • arrive through “direct load”

  • convert through unattributed paths

  • behave as though they already know you

This manifests as:

  • branded search lift

  • direct traffic spikes

  • unattributed conversions

  • higher-quality downstream actions

The twist most teams misinterpret

AI exposure increases branded demand but decreases branded clicks.

Why?Because AI answers resolve shallow branded queries upstream, leaving only deeper-intent actions for humans.

Clicks fall.Intent rises.

This confuses teams who haven’t yet recognized the shift.

C. Where the real insight comes from

Put together:

  • LLM referral logs

  • branded search behavior

  • multi-LLM answer presence

And you begin to understand:

  • which models influence your category

  • which types of answers correlate with pipeline

  • whether AI-exposed users convert differently

  • whether your brand’s visibility is improving or collapsing

This isn’t perfect attribution.It’s the attribution that reflects how buyers actually behave in 2026.

And yes…it’s uncomfortable when you like clean answers (I feel this deeply).But discomfort doesn’t change reality.

03. AEO: The New Ranking System

In a zero-click world, AI doesn’t rank keywords.It ranks entities.

Models care about:

  • who you are

  • what you offer

  • how consistently you’re described

  • what trusted third parties say about you

  • whether your narrative aligns across the web

Entity SEO existed before.AI made it the dominant ranking force.

To strengthen your entity:

  • clarify product and company entities

  • align naming across every owned surface

  • structure About / Pricing / Features to be machine-legible

  • ensure schema is accurate

  • establish credible authorship and expertise

  • invest in offsite corroboration across reviews and comparisons

LLMs fire multiple retrieval queries per prompt, so consistency across surfaces becomes the ranking signal.

If your entity graph is unclear or inconsistent, your visibility is capped.It doesn’t matter how good your content is.

Models evaluate you like Wikipedia, not a keyword list.

If you want the deeper mechanics — retrieval patterns, entity construction, and measurement — I unpack it fully in my 2026 AEO Guide, which became the manual I wish I’d had when this shift first showed up in our numbers.

04. Conversational Commerce: The New Conversion Path

Buyers aren’t browsing the way they used to.They’re asking their way through the category.

“What’s the best option for…?”“How does pricing compare?”“What should I consider?”

These aren’t research steps.They’re the funnel.

The new journey looks like:

Ask → get a recommendation → clarify → compare options → validate → decide → maybe click once

Your job becomes:

  • be included in the answer

  • be recommended

  • be trusted

  • be the assumed default

  • be the consensus pick

You don’t win because someone visits your website.You win because AI already decided you belong in the conversation.

05. Measuring Traffic When Attribution Blurs

Attribution isn’t broken.It’s evolving.

The 2026 measurement stack:

  • LLM referrer logs

  • branded search lift

  • multi-LLM answer presence

  • consensus gap analysis across competitors

  • entity health scoring

  • post-click activation

This is the first meaningful evolution since multi-touch attribution — not more precision, but more truth.

06. AI Holds the Homepage

Here’s the part most teams still underestimate:

Users may land on your homepage less often.But AI reads it constantly.

LLMs repeatedly ingest:

  • pricing

  • docs

  • features

  • changelogs

  • help content

  • reviews

  • comparisons

Your homepage now has two audiences:

  1. AI systems deciding whether to recommend you

  2. Buyers arriving after AI already shaped their perception

So the real strategic question becomes:

What does ChatGPT think my homepage says?

Because if AI misinterprets your positioning, your funnel collapses long before a human ever touches your site.

Your website still matters…not as the first touch, but as the final source of truth.

THE PUNCHLINE

The Zero-Click Funnel isn’t coming.It already rewrote the buyer journey.

The marketers who win in 2026 won’t be the ones squeezing incremental gains out of traditional SEO or last-click analysis.

They’ll be the ones who understand:

  • visibility happens in answers

  • authority is built through entities

  • traffic is indirect

  • conversion is conversational

  • relevance is decided corpus-wide

  • AI reads your site before the buyer does

  • and presence matters more than ranking

Your job isn’t to rank higher.It’s to:

Be the default answer.Be the trusted recommendation.Be the consensus pick across LLMs.

Because if you don’t control the answer layer,you don’t control the funnel.

And as Dylan said — summing up every era where the ground shifted beneath our feet:

“You better start swimmin’ or you’ll sink like a stone.”

The times aren’t going to change.They already did.

About the Author

Josh Grant is VP of Growth at Webflow, where he leads Growth Marketing, Demand Generation, Lifecycle Marketing, Partner & Ecosystem Marketing, Marketing Ops, Analytics, and Webflow’s AI Search & AEO strategy. He architected Webflow’s growth engine and AI search motion, pioneering LLM traffic intelligence, enterprise-scale structured content systems, and one of the industry’s most widely adopted AEO analysis and optimization frameworks.

Josh is widely regarded as one of the most influential operators at the intersection of AI and growth. He was named PLG 2025 Leader of the Year for his contributions to modern growth systems and his work redefining how companies acquire and convert users in an AI-first world.

His frameworks and insights have been referenced across major industry channels — from repeated call-outs on Lenny’s Podcast to widespread engagement on LinkedIn, where his content has reached over 1,000,000 impressions and 400,000+ operators and executives in the past 90 days, fueled by collaborations with leaders like and .

He publishes long-form strategy breakdowns on his Substack, Stacked GTM, where he writes about AI growth, AEO, narrative design, and modern GTM systems, and he advises multiple startups on GTM, AI-first marketing, and category leadership.

Follow Josh on:

All opinions are my own and do not represent any employer, past or present.

Keep Reading