Never Always, Never NeverNever Always,
Never Never
The BookAI CreativeAI ProjectsChatNewsletter
Buy the Book
Never Always,
Never Never.

Strategic Marketing in an AI World.
By Patrick Gilbert.

Explore

  • The Book
  • The Author
  • Free Chapter
  • Buy
  • Resources

Connect

  • Newsletter
  • Learn
  • AI Projects
  • Blog
  • Chat with the Book
  • Subscribe

Stay Connected

Get updates, bonus frameworks, and new AI project showcases.

© 2026 Patrick Gilbert. All rights reserved.

AdVenture MediaContact
Strategy7 min readNovember 17, 2026

The Marketing Funnel Was Always a Diagram. Never a Map.

Patrick Gilbert

Patrick Gilbert

CEO of AdVenture Media. Author of Never Always, Never Never.

The marketing funnel has never accurately described how people buy. That was never its job. It was a budgeting diagram dressed up as a behavioral model, and confusing the two has cost marketers more wasted spend than almost any other strategic error.

Funnel thinking is not dead. The fantasy that it was ever a literal map is.

The Original Sin: Mistaking a Diagram for a Route

Traditional funnel logic told a tidy story. Consumers see your ad. They become curious. They research. They buy. Each stage hands off cleanly to the next, like a relay race with predictable legs.

That story was always a simplification. It just happened to be a useful simplification when the actual purchase environment enforced the sequence. When a consumer saw a TV ad in 1987, they genuinely could not buy immediately. The friction in the system made the funnel look accurate. What looked like a law of consumer behavior was actually a function of infrastructure limitations.

Remove those limitations and the sequence collapses. The funnel didn't die when smartphones arrived. We just finally got evidence that it had never been real.

Patrick Gilbert covers this directly in Never Always, Never Never, drawing on Google's landmark research into what actually happens between a consumer's first trigger and their eventual purchase decision. Clean linear stages gave way, in Google's framing, to something far more honest: a messy middle where consumers loop repeatedly between exploration and evaluation, influenced by memory, social proof, availability, competitor messaging, and cognitive shortcuts, often in no particular order.

What Google's Research Actually Found

In Decoding Decisions: Making Sense of the Messy Middle, Google's research team simulated purchase scenarios across a wide range of product categories. The finding that should permanently retire the idea of a tidy funnel: a significant share of participants switched their stated brand preference after being shown basic competing information. In categories like car insurance and broadband, the switching was even more pronounced.

Think about what that means. These weren't uninformed consumers who hadn't heard of the brands. They had already named their preferences. Then they encountered normal competitive signals and changed their minds anyway.

Brand loyalty isn't a gate consumers pass through once and lock behind them. It's a position that must be actively maintained across the entire messy middle. A brand that keeps appearing during exploration and evaluation doesn't win because it's better. It wins because it's present.

Google calls this the power of showing up. For anyone running a bottom-funnel-only strategy, the practical implication is straightforward and uncomfortable: a brand that sits out the exploration and evaluation loops because it only buys conversion-stage placements will lose customers it technically already had. Competitors don't need to build a better product. They just need to be more visible during the decision process.

Mental availability matters so much for exactly this reason. It isn't about being top of mind in some vague awareness sense. It's about being retrievable at the specific moments when consumers are looping through options, which can happen multiple times across a single purchase journey.

The Cognitive Architecture Behind the Loop

Google's research identified six specific cognitive biases that shape how consumers navigate the messy middle. Category heuristics give buyers simple rules of thumb to cut through complexity. The power of now makes immediate availability more persuasive than it should be rationally. Social proof, scarcity, authority, and the word "free" all move decisions in ways that have nothing to do with the product's objective merits.

In Google's simulations, these tactics produced meaningful results. Offering a free checked bag with a competing airline caused a notable spike in participants switching from their preferred carrier, including among consumers who hadn't planned to check a bag. The perceived value of free shifted behavior even when the customer wouldn't use the benefit.

A direct implication follows for how marketers think about the funnel stages. If a consumer at the apparent "consideration" stage can be redirected by a single social proof signal or a free shipping offer, the idea that they've progressed past awareness and are now in a separate, manageable bucket is a fiction. They're still in a live, competitive space where any input can restart the loop.

Addressing this directly, the category entry points framework argues that brands need to be present across the full range of mental triggers that can activate a category, not just at the moment a consumer declares intent.

Social Commerce Has Collapsed What Remained of the Sequence

Even if you accepted the messy middle as a reality for search and research-heavy categories, you might have argued that social advertising still roughly followed a top-down logic: reach with video, retarget with consideration content, convert with product ads. That sequence has fractured.

The Drum's analysis makes the point clearly: social platforms have fused discovery and evaluation into a single environment. A consumer can encounter a product through a creator post, assess credibility through comments, watch a demonstration, see social proof from people they follow, and complete a transaction without leaving the feed. The stages don't just overlap. They happen simultaneously.

This makes "top of funnel," "middle of funnel," and "bottom of funnel" increasingly useless labels for social content planning. A single post can create demand, demonstrate the product, answer objections, provide social proof, and generate a conversion in one interaction. Treating that post as an "awareness" unit and measuring it only on reach or brand lift will misread its actual contribution.

A better unit of analysis is what some practitioners now call an influence point: the specific moment or interaction that changes a buyer's confidence, preference, or intent. That could be a creator video. A comparison search. A peer recommendation. A competitor's ad. These don't arrive in a predictable order, and they don't sort neatly into funnel stages.

Where the Funnel Still Earns Its Place

None of this means you should delete the funnel from your planning documents. SAS's 2026 marketing predictions argue that customers don't move neatly through stages but loop, stall, switch channels, and return based on social influence and changing needs. Their proposed alternative is an adaptive ecosystem that uses AI to recognize behavioral patterns rather than force customers into predefined buckets. That's the right direction.

But the funnel retains genuine value as an organizational framework. It still makes sense to distinguish between demand creation and demand capture. Les Binet and Peter Field's research on marketing effectiveness draws exactly this line, and it holds. Separating brand-building investment from performance investment, thinking carefully about which audiences are in-market versus out-of-market, and not treating every marketing dollar as a conversion play is real strategic thinking. The 60/40 principle that Binet and Peter Field documented through the IPA DataBank doesn't depend on consumers moving through the funnel in sequence. It depends on recognizing that most of your future customers aren't ready to buy right now, and that you need to be in their memory when they eventually are.

As a budget allocation guide, the funnel is defensible. As a description of consumer movement, it is not.

That distinction matters because it determines what you measure. If you believe consumers move linearly through stages, you build attribution models that try to tag each touchpoint to its stage and credit it accordingly. This produces the illusion of a clean path. It also produces chronic underinvestment in brand-building activity, because brand work operates over months and years while conversion lag in attribution models is usually measured in days. We've written about this measurement problem in more detail here.

AI Is Making the Journey More Recursive, Not Less

If social commerce fractured the remaining sequence, generative AI is fracturing it further. A buyer can now ask an AI assistant for recommendations, validate those recommendations through social discussion, watch product demonstrations, compare alternatives, return to the AI tool with follow-up questions, and only then visit a brand's website. The journey can begin entirely outside any property the brand owns or tracks.

Google's 2025 guidance on AI-enabled media planning explicitly recommends moving away from treating the customer journey as a linear path tied to specific media formats. It promotes influence maps as a planning tool, designed to identify where different interventions can affect decisions across the full range of consumer behavior. Google Ads' 2025 product updates reinforced this direction, emphasizing measurement systems and AI-powered campaign tools intended to reach consumers from discovery through decision-making as a single continuous challenge, not as separate funnel stages.

Practical consequences follow for how you think about AI search as a marketing channel. When a consumer asks an AI assistant which product to buy, they may never see your paid search ad or your retargeting pixel. The influence point exists before your attribution system can observe it. A brand that gets recommended is one that has built enough presence across enough sources that AI training data reflects it positively. That's a brand-building problem, not a conversion optimization problem.

At AdVenture Media, this shift from linear funnel thinking to influence-point thinking is one of the more consequential strategic adjustments happening in how paid media campaigns are planned and evaluated.

The Actionable Conclusion

Stop asking whether the marketing funnel is dead. That question keeps sending you back to the wrong debate. Funnel stages, awareness, consideration, conversion, retention, advocacy, are still useful categories for organizing objectives and budgets. They are not useful as a prediction of how any individual customer will actually behave.

Here's what to do instead:

  • Map influence points, not stages. Identify the specific moments where your buyers gather information, seek social proof, compare options, or reconsider a decision. Build presence at those moments regardless of where they fall in a theoretical sequence.
  • Design content to do multiple jobs at once. A strong piece of content can create demand, demonstrate the product, answer objections, and provide social proof simultaneously. Forcing every asset into a single funnel stage label wastes potential and misallocates creative investment.
  • Measure incrementally, not sequentially. Stage-progression metrics tell you whether your funnel model is working. Incremental measurement tells you whether your marketing is working. These are not the same question. Incrementality testing gives you an honest answer. Attribution models often don't.
  • Build [physical availability](/learn/glossary/physical-availability-glossary) alongside mental availability. Memorable brands that are hard to find lose ground, and so do easy-to-find brands that aren't remembered. Both matter, and neither is the exclusive job of one funnel stage.
  • Don't optimize only for in-market buyers. The 95-5 rule is a useful anchor: at any given moment, roughly 95% of your potential buyers are not actively in-market. Bottom-funnel-only strategies ignore the future customers who will eventually be ready to buy, and who will buy from whatever brand was most present during their eventual exploration loop.

The marketing funnel was always a diagram. A useful one. A diagram that helped teams align budgets, set objectives, and avoid the trap of treating all marketing dollars as conversion plays.

It was never a map of how people buy. And building strategy as though it were has been one of the most expensive collective mistakes in modern marketing.

Patrick GilbertPatrick Gilbert

Patrick Gilbert is the CEO of AdVenture Media and author of Never Always, Never Never and the bestselling Join or Die. He has been ranked among the top 5 PPC experts worldwide and has delivered keynotes at Google events across three continents.

More about Patrick →

Enjoyed this?

Subscribe for more articles on strategy, AI, and what's actually working in marketing.

No spam. Unsubscribe anytime.

Keep reading

Measurement

Marketing Attribution Is Broken. Here's What to Do Instead.

75% of marketers say attribution underperforms. Stop using measurement tools as scoreboards and start treating them like film rooms.

Strategy

Brand vs Performance: The Budget Split Question Has a Wrong Answer

The 60/40 rule isn't dead, but chasing a fixed split is. Here's how to actually allocate your budget between brand and performance marketing.

Strategy

The 95-5 Rule: Most of Your Buyers Aren't Ready to Buy (And That's the Point)

The 95-5 rule says 95% of B2B buyers aren't in-market right now. Here's what that means for your budget, your strategy, and your growth.