Happy Sunday, Everyone!

I hope everyone’s having a good week. Happy (belated) Valentine’s Day to those who celebrate - may your significant other appreciate you as much as Google appreciates your ad budget (which is to say: quite a lot, but mostly when it’s convenient for them).

This week, I want to talk about something timely that’s been on my mind.

Over the past few weeks, Marty Weintraub and the team at AIMCLEAR published a fantastic two-part “Gut Check” series. Part 1 focused on multi-surface organic discovery featuring 15 experts; part 2 on paid performance media with 23 more. I was honored to be included in Part 2, and I’d genuinely encourage you to read both. The summaries are excellent. The trends align very closely to what we’ve seen across the accounts we manage.

What I found most interesting wasn’t any single take. It was the degree of agreement. Nearly 40 of the sharpest minds in marketing are saying remarkably similar things.

I have enormous respect for the experts featured in both articles. Many of them are friends. But when an entire industry reaches consensus, I think it’s worth pressure-testing the assumptions. I’ve continually advocated that everyone should zig when the crowd zags; it would be hypocritical to think that philosophy doesn’t apply simply because I’m included in the majority.

And I think there’s a big piece of nuance getting lost.

The Prevailing Narrative

Here’s the dominant thesis emerging from these conversations:

The click economy is dead.

AI Overviews generate roughly 1% CTRs (far less than the 35% rates for SERPs). Almost 65% of traditional searches end without a click. Put those two data points together, and you’ll likely reach the conclusion that the old playbook is a liability and the future belongs to brands that optimize for visibility, citation authority, and LLM perception, not clicks, not traffic, not conversions in the traditional sense.

If you take it a step further: the relative import of traditional ranking is declining, too. After all, being “quoted first” rarely matters in the context of a detailed, comprehensive answer. It matters even less if the entity reading that citation isn’t human (something that’s increasingly common).

On the paid side, the consensus is equally tidy: automation is inevitable, so layer intelligence on top. You can’t out-algo Google or Meta. Keywords are dying. PMax is the future (now with windows in the black box!). Consolidate your account structure. Measure incrementality, not ROAS.

I agree with almost all of this. I’ve written about incrementality (Issue #148), about the dangers of mediocrity (Issue #146), about how the next era of marketing demands depth over breadth, and about the economics of navigating uncertainty in Marketing Through the Tariff Storm.

But there’s a piece of the narrative that I think deserves a harder look - and it just so happens to be the claim with the most momentum behind it: “The click economy is dead.”

Let’s not bury the lede: that’s wrong.

The click economy is relocating. It’s evolving. The people declaring its death are confusing where people click with whether they click.

The Error at the Center

Before I walk through the evidence, I want to name the structural error in this narrative, because it’s the same error showing up on both the organic and paid sides of the conversation.

The industry is optimizing for systems when it should be optimizing for signal quality.

On the organic side, that error takes the form of treating “visibility” as a terminal goal, as if being cited in an AI answer is the end state, rather than a waypoint to a business outcome. On the paid side, it takes the form of “automation layering” aka building ever-more-sophisticated guardrails around an engine that is increasingly pointed at the wrong objective.

Both are symptoms of the same confusion: mistaking optimization of the mechanism for optimization of the outcome. The brands that will define the next era won’t be the ones that best navigate the systems. They’ll be the ones whose signal is so clear and differentiated that the systems have no choice but to amplify it.

That’s the thread connecting everything that follows.

The Click Didn’t Die. It Fragmented.

65% of traditional Google searches end without a click to a website. That’s a data point I’ve seen compelling evidence (from Rand Fishkin, the Datos team, SEMRush, our own Google Search Console and client site data) is valid. Yes, a lower percentage of Google searches end with a click to the open web than ever before.

But what almost nobody in either AIMCLEAR article addressed is that new click economies are forming inside the AI surfaces that supposedly killed the old one.

I shared data in Issue #148 showing LLM-sourced leads for a B2C business growing from less than 5% to over 20% of total leads in 7 months. These aren’t “visibility impressions.” These aren’t “brand mentions in an AI answer.” These are real leads that convert into real customers who spend real money. What’s equally interesting? Those leads increased while the total number of leads from other sources also trended up. Put another way: AI leads were not previously organic leads; they were net-new. Incremental.

If the click economy was really dying, you’d expect the non-AI segment to decline while the AI segment increases. That’s not what our data shows across hundreds of accounts; it shows the opposite. Both segments are increasing. That’s not the death of the click economy; it’s the birth of a new one.

Think about what’s actually happening: a user asks Gemini or Claude, “What’s the best running shoe for flat feet?” The LLM synthesizes information, recommends a set of products and the user clicks through to buy. The click happened. It just didn’t happen on a SERP. It happened inside Claude or Gemini or ChatGPT.

And as of 5 days ago, that click has a price tag.

OpenAI officially started testing ads inside ChatGPT on February 9th. These are sponsored results that appear below the AI’s answer, matched to the topic of your conversation. So let me get this straight: the industry spent 2025 declaring the click economy dead… and then the biggest AI company on the planet built an ad platform specifically designed to monetize clicks inside the AI answer.

If the click economy is dead, someone forgot to tell OpenAI’s revenue team.

The Click Is Evolving, Not Disappearing

What we’re actually witnessing is a shift in the type of click that matters, and that shift makes the surviving clicks exponentially more valuable.

The user journey used to be linear: Google search → click → website → convert.

Now, when we do actual user testing and target audience observation, it looks more like this: Shorts → ChatGPT → Reddit → YouTube → AI Overview → click → website → Reddit → ChatGPT (again) → website → convert. Or it might skip “website” entirely and convert inside a platform like IG or TikTok Shops. The path is messier, longer, and harder to track. But the click - the moment a human takes action - still happens. It just happens in more places.

More importantly, we’re moving from “information-seeking clicks” (which AI is absolutely absorbing) to “intent-confirming clicks” (which are becoming exponentially more valuable).

You’ve likely exhibited some of this behavior yourself. An illustrative example from my recent retail therapy binge: I’ve been shopping for several new polos. Instead of my usual process, I’ve been using Gemini a lot more - but my search behavior inside Gemini is quite different from Google. I ask 5-10 follow-up questions inside one session versus conducting dozens of individual searches on Google. I refer back to previous exchanges. I gather all the information I need to make a single, informed choice inside Gemini, so when I click on Sunspel’s result in the AI answer, I’m already sold.

The information that “sold” me was retrieved from Sunspel’s website. The photos, sizing guide, material composition, reviews - all of it was theirs. Gemini supplemented it with conversations on Reddit, several GQ articles, write-ups in men’s fashion blogs I’d never heard of (let alone would think to read) and a YouTube video of a guy I’ve never heard of doing a 22-minute review of the brand.

Sum it all up, and my single click was the product of the same amount (if not more) research than the dozens of clicks my old search behavior would have generated, but the quality of that click was infinitely higher.

Here’s the piece almost nobody talks about: AI is simultaneously creating new click demand that didn’t previously exist. AI Overviews are generating commercial queries from informational intent. Google just confirmed this in their Q4 earnings. Gemini, ChatGPT and Claude are creating purchasing pathways that never existed on traditional search. These aren’t replacement clicks; they’re net-new, incremental opportunities. This aligns with the data I shared in Issue #148.

Now factor in that AI Mode queries are 3x longer and increasingly conversational. The surface-level click economy is going away. Of that, we can all be sure. But what’s replacing it is a click economy that’s deeper, more complex and in many cases, more valuable.

That makes the correct framing:“the click economy has gone multi-surface and multi-modal.” And that framing comes with a completely different strategic implication.

The Numbers Don’t Lie

Consider AI Overviews: just 10 days ago, on Google’s Q4 2025 earnings call, Sundar Pichai confirmed that Search saw more usage in Q4 than ever before, with AI continuing to drive what he called “an expansionary moment.” Daily AI Mode queries per user doubled in the US since launch. Sessions are becoming more conversational, with a significant portion of queries leading to follow-up questions. Nearly 1-in-6 AI Mode queries are now non-text, using voice or images (aside: I think the era of voice search is finally here).

Google’s Chief Business Officer, Philipp Schindler, confirmed that AI Overviews and AI Mode are driving growth in overall queries, including commercial queries. Search revenue grew 17% YoY. Ad revenue hit $82.28 billion for Q4 2025. Alphabet’s annual revenue exceeded $400 billion for the first time.

Read that again: the company at the center of the “clicks are dead” narrative just reported record search usage, record ad revenue and confirmed that AI features are generating more commercial activity.

Who Benefits from “Clicks Are Dead”?

Whenever I encounter a narrative with broad industry adoption, I find it useful to ask, “Who benefits from this being true?”

That’s not (entirely) because people are being dishonest, but because incentive structures shape how ideas spread, and understanding those incentives helps you think more clearly. In this case, the “click economy is dead” narrative happens to benefit 2 groups disproportionately:

The platforms. Google, OpenAI, and every AI surface provider benefit from the idea that clicks to your website are less important. If marketers accept that “visibility” and “being cited” are the new KPIs, they become more dependent on the platform’s own ecosystem and less focused on pulling users into their own. Google’s been moving in this direction for years: broader match types, fewer search term insights, audience expansion as a default. I wrote about this in detail in Day Trading in Paid Media: the platforms are systematically creating a more efficient market for themselves. That’s not malicious; it’s just business. But it’s worth being clear-eyed about it. When you accept that “visibility” is the primary goal, you’ve made it a lot easier for the platform to grade its own homework.

The emerging services ecosystem. There’s a rapidly growing industry around “AEO” (Answer Engine Optimization), “GEO” (Generative Engine Optimization), “LLM perception drift monitoring,” and “agentic friction audits.” Some of these concepts can deliver real value. Most — at this stage — are run by the same people that spent the early 2020s telling us that JPEGs were the safest investment vehicles on planet earth. We’ve all seen this before — maybe not in this specific space, but history does rhyme. Services built around fuzzy metrics (visibility, citation, brand salience) are inherently harder to hold accountable than services built around verifiable outcomes (leads, revenue, profit). That’s not a reason to dismiss them outright, but it is a reason to maintain healthy skepticism about how much weight their proponents should carry as you plan your strategy.

I’m not saying visibility doesn’t matter. It does. But visibility that doesn’t ladder to business impact is just brand awareness wearing a lab coat.

The Same Error, on the Paid Side

The paid media discussion from Part 2 reflects the same underlying confusion: the tendency to optimize for the system rather than the signal.

The dominant take is: automation is inevitable, so layer intelligence on top. Guide the machines. Don’t fight the algorithm. I agree with the mechanics of this. The incrementality work that Frederick Vallaeys and the Optmyzr team has been doing is genuinely breakthrough-level. Brad Geddes’ documentation of the match type evolution is essential reading. There are very good, very practical ideas on how you (as a PPCer or marketing practitioner) can place controls on an increasingly resistant-to-control system.

But most of the discussion is about how to optimize within the system. There’s less exploration of whether the system itself is oriented toward the right objective.

In The Next Era of Marketing, I argued that the industry’s obsession with “more” - more impressions, more reach, more clicks, more spend - is fundamentally misguided. The platforms reward scale and optimize for volume. Every recommendation, every default setting, every “optimization” suggested by the Recommendations tab is designed to get you to spend more (in fact, I have never seen a recommendation to decrease budgets, increase tROAS, or decrease tCPA. Ever).

“Automation layering” is a smart tactical response to this reality. But it doesn’t challenge the underlying premise. Layering intelligence on top of a fundamentally flawed “more” strategy just makes you more efficiently mediocre. It doesn’t make you better.

The better question is: “Am I building something that matters deeply enough to a specific audience that the automation works in my favor because the signal is unmistakable?”

That’s the difference between guiding an algorithm and giving it something worth guiding toward.

One of the points I’ve consistently made is that all automation is like water: it finds the path of least resistance. If you allow Meta to target existing customers or warm leads or remarketing audiences in your prospecting campaign, guess what it’s going to do? The same applies to Google: if you allow PMax to include branded searches, that’s what it’s going to do — because that’s the path of least resistance.

If your brand is unremarkable - if you’re stuck in the middle - the algorithm will happily cannibalize your brand search and remarketing pool and call it “performance.”

The brands that win with automation aren’t the ones with the best layering strategy. They’re the ones with a brand strong enough that the algorithm has clear, differentiated signals to work with. The creative matters. The offer matters. The story matters. 90% of the results inside your ad account are a product of what’s going on outside your ad account.

The Real Moat: Differentiation Through Truth

This is where both sides of the conversation - organic and paid - converge to the same answer.

So much of the current discourse is focused on AI, automation, and machine systems. Every time I open LinkedIn, there’s a new post on technical SEO. Structured data. Passage-level retrieval. Vector embeddings. Agentic friction. AEO. All of those things matter, at least to some degree. But the conversation about making your target audience feel something gets a lot less airtime.

I predicted that Ogilvy-style advertising would make a comeback in 2026, simply because audiences are tired. They want something that’s real and substantive. Something that feels made with effort and intention, not just thrown together by a junior marketer and ChatGPT.

In a world of generative everything, where sameness can quite literally be automated, differentiation through truth becomes the ultimate moat. You can’t outsource belief to an AI system (at least, not yet). The brands that optimize purely for machine legibility - structured data, vector embeddings, agentic APIs - will find themselves competing in a race where the barriers to entry collapse to zero. Anyone can format schema markup. Anyone can generate 10,000 pages of “optimized” content. That’s table stakes, not a competitive advantage. When the cost of production goes to zero, the value of knowing what to create - the value of taste - goes to infinity.

The brands that will win are the ones that use those AI systems to become more human, not more optimized. The ones that tell one truth, consistently, honestly and well. The ones whose creative and perspective is so differentiated that the algorithm amplifies it not because of layering strategies, but because the signal is undeniable.

Put another way: LLMs are (inherently) giant pattern recognition systems. That makes the job of marketing to create the right pattern. Everything you create - an article, a video, a deck, a PDP, a social post - either amplifies that pattern, or detracts from it. That’s not an optimization problem, it’s a taste challenge. It’s a question of discretion, not creation.

The irony of 2026 is this: we’ve collectively published tens of thousands of words on how to optimize for AI systems. The industry’s best minds have mapped out where we are. But the brands that will define the next era will ignore much of it, in favor of exploring the territory that isn’t on the map.

The Only Problem Worth Solving

This week, I want you to do something counterintuitive:

Stop thinking about your AI visibility strategy for 48 hours. Instead, call 10 customers. Not a survey. Not a feedback form. Real conversations with real people who gave you real money. Ask them why. Not what features they like. Not how they found you. Ask them why they chose you over every other option, and what you mean to them.

Then ask yourself: if you fed those 10 answers to an AI, would it know who they were talking about? Or could it be any brand in your category?

If it could be anyone, your problem isn’t visibility. It’s that you don’t have anything worth being visible for.

That’s the only problem worth solving.

Until next week,

Cheers,

Sam

P.S. Whenever you’re ready, here’s how I can help: if your brand needs a strategic partner that blends performance marketing, analytics, and brand into one integrated team — not five siloed agencies — reach out or schedule a call.

Loving The Digital Download?
Share this Newsletter with a friend by visiting my public feed.

Follow Me on my Socials

​​​

1700 South Road, Baltimore, MD 21209 | 410-367-2700
​Unsubscribe | Manage Preferences​