Hey👋,
I'm Giacomo

Last week, I moderated a workshop at the Digital Marketplaces Association conference in Zurich.

The topic... You guessed it. AI. 😅

Actually, almost every session was about AI.

But unlike most events I’ve attended recently, the focus wasn't on efficiency or operational processes.

It was on something much more critical.

How AI chatbots are fundamentally reshaping people's behaviour online.

From search queries to questions.
From browsing to reading answers.
From clicks to mentions and direct traffic.

This has massive implications for digital marketing and online businesses in general.

No one is immune.

It was fascinating to discuss this with a group of like-minded business leaders.
We looked at the current situation, the risks, and most importantly the opportunities.

Here is the consensus:
- We are just at the beginning. But the shift is happening.
- The risks are real. But the opportunities are greater.
- There's no playbook! we'll still need to figure.

On the last point,
someone did ask me whether I could share a full guide on digital marketing in the age of AI.

Well, I wish it existed!

Exciting times ahead!

I’ve had many discussions here on LinkedIn about who is leading the AI Chatbot race.

There is a lot of noise about new features and personal opinions.

But if we look at the actual usage data, the answer is boringly simple.

ChatGPT is the undisputed king.👑

I was looking at this desktop user panel by Datos, A Semrush Company:

- ChatGPT: used by over 40% of users.
- Gemini: growing, yes. But still less than 10% market share.

Gemini has less than a quarter of GPT’s users. Despite the Google ecosystem, Android default etc.

OpenAI released a paper a couple of months ago that explains exactly what’s happening.

They analyzed how people actually use ChatGPT.

The results were unexpected.

It’s not for coding. It’s not for work.

It’s mostly used for non-work activities. 🏠
• How-to guides
• Learning new topics
• Searching for quick info

Essentially, ChatGPT has become the ultimate personal assistant for everyday life.

And that is their moat.
For now.

The other tools are fighting for the "specialist" slice: coding, enterprise workflow, etc

But for the average person, there is only one default. For now.

Brand vs. Performance.

The hottest debate in marketing is coming to an end.

In the 2010s, we lived in the golden age of performance marketing.

Google and Facebook sold us a perfect dream:

- Invest $1. Get $5 back.
- Track every single user journey, from ad click to purchase.

It worked. For a while.

Until two asteroids hit the marketing planet:

1. Privacy killed the cookie 🍪
-> GDPR and browsers like Safari blocked third-party cookies.
The bridge between ad click and purchase started to collapse.

2. AI broke the funnel
-> “Upper funnel" discovery now happens in ChatGPT or AI Overviews.
These conversations are critical for purchase decisions. But they generate zero clicks.

Think about it.
Users discover your brand by talking to ChatGPT, and then visit your site directly to complete a sale. Which ad brought the sale? Impossible to tell.

So, what now?

A landmark study by WARC found that when brand and performance activities merge, they multiply ROAS and sales, both short and long term.

Whether we like it or not, the tech giants are already forcing us there.
Look at Google Performance Max or Demand Gen.

They are AI engines designed to optimise results across the entire user journey.
With or without clicks.

AI is solving the problem that AI created 😅

So,

No more Brand vs. Performance.

Now it’s Brand x Performance x AI.

Are you in?

I don't buy individual stocks.
But the day OpenAI goes public, I’m buying. Immediately.

Here’s why:

1. It's already mainstream.
→ 10% of the world's adult population now uses ChatGPT.
→ It reached 1/3 of Instagram users, in less than three years.
→ Over 70% of all GPT conversations are non-work related.

2. Massive room for growth.
→ The 18-25 age group sends nearly half of all messages.
→ Growth in developing countries is exploding.

3. It's becoming the "Everything App".
It isn't just a chatbot anymore. It's the gateway to the web:

→ Personal Assistant: ChatGPT
→ Entertainment and social networking: Sora
→ Browser: Atlas
→ Shopping: Instant Checkout
→ App Store: GPT Apps (essentially the new Apple App Store)
→ Messaging: Group Chats (just announced!)

4. Huge potential for monetisation.
Yes, it still burns billions.
But it's already at $13B+ annual revenue.
Facebook took years to even make a single dollar.
OpenAI did it on day one.

The path is clear:
→ Ads (huge opportunity).
→ Commissions from checkouts.
→ More tiered subscriptions (free with ads, Lite?, Plus, Pro, Enterprise, etc).
→ Massive B2B potential.

Meanwhile...
The other Big Tech giants are too large and slow to keep pace.
Yes, even Google.
Gemini is growing fast but its market share is still a fraction of ChatGPT.

Many are trapped in the innovator's dilemma.
OpenAI doesn't fear any dilemma.

It managed to disrupt the most powerful companies on earth right where they were strongest: AI & machine learning.
Nobody else has ever done that.

What else should I add?

Microsoft is moving to the light side of the force.

It’s building a content marketplace where publishers and creators can sell their work directly to AI chatbots.

Auction or fixed price, still unclear.

Why this is a huge deal:

Publishers are getting crushed.
Search traffic is collapsing after “Google Zero”, or the so-called "death of clicks".

Less traffic means less ad revenue.

Meanwhile, AI chatbots keep using their content for free…
without even sending traffic back 😅.

This model is clearly not sustainable.

Sure, OpenAI has begun striking content licensing deals.
For example with the Financial Times.

But these are slow, maybe annual negotiations, while AI usage grows exponentially in the meantime.

At this point, publishers need a Big Tech giant to finally take their side.
If one moves, the rest will follow.

Problem is,
Microsoft’s marketplace would only feed CoPilot. 😐
And CoPilot is minuscule compared to the main players.

Still, it’s a start.
And Microsoft’s stake in OpenAI could make the ripple bigger.

Will it work?
Hard to tell.

But one thing is certain:
How this shift will evolve will either save the open web… or kill it.

Marketers, what can we learn from this chart?

1️⃣ Reddit alone gets 40% of all LLM citations.
That’s a massively skewed distribution toward the top.

2️⃣ The top-cited brands are huge, high-traffic websites.
LLMs clearly favour big names.
Makes sense, they’re statistical engines.
The more a brand is known, the more chances it shows up in the model’s training data.

3️⃣ Most of them are platforms.
Not publishers or blogs, not retailers, not producers.

So...

If you’re a small business,
forget about “LLM visibility.” That’s not your game.

If you’re a large business (but not in the top 20),
you can play. But keep in mind, platforms will still win.

If you’re in the top 20,
congrats, you made it.
But now hold that position tight, the challengers are more aggressive than ever.

Two obvious takeaways:

-> Be on all these platforms.
Reddit is the most cited domain, but it doesn’t have content of its own. It’s your content that will be surfaced!
Same story for most of the other top domains.

-> If you’re small, grow!
Sounds obvious, but it’s worth repeating.
How? online PR and wide-reach ads.

Stop making websites. You don’t need one.

All you need is a backend and an API.

Welcome to the future of the internet!
where users interact with your website without ever visiting it.

Sounds crazy?
Then take a look at ChatGPT Apps.

People can now:

find apartments to rent,
book hotels,
even play Spotify...

all without leaving ChatGPT.

The entire browsing and discovery process happens inside the chatbot.

Only the final action, the payment or transaction, happens outside (for now).

This changes EVERYTHING.

Traffic acquisition is dead. 💀 
You don't need traffic anymore.

All you need is to be featured in ChatGPT Apps.

Here’s what’s going to happen next:

- ChatGPT Apps will become the new App Store.
- Apps will compete for ranking, organic or paid.
- Users will choose the app they trust.
- Or GPT will choose for them based relevancy to the conversation.

A new marketing ecosystem is rising.

Think of...

Sponsored placements.
Apps PPC.
Pay-per-conversion.
Organic app ranking.

We’re witnessing the birth of a new internet,
and the dawn of a whole new era of digital marketing.

"Hey baby, I got a bonus, let's go celebrate!"
"Well done, how much?"
"1,000,000,000,000 dollars."

It sounds so insane that actually makes sense.

To get the cash, Elon Musk must:

- Raise Tesla’s valuation to $8.5tn (twice the most valuable company today).
- Boost earnings to $400bn, about 24× what it earns now.
- Sell millions of Optimus robots.
- Sell millions of autonomous driving subscriptions.

Will he make it?

I think so. I don’t see why not.

The only real risk here, is Musk himself.

What if something happens to him before that?
He almost died at least twice.

How much is Tesla worth without Musk?
Probably just a fraction of today’s valuation.

Maybe close to nothing actually.

Remember,
Tesla is already one of the most overvalued stocks in America.
P/E ratio 298.3.

For context,
NVIDIA is at 53.5, and Meta at “just” 27.4.

So the Musk premium is already priced in.

But if he actually delivers, the world will be a much happier place.

Millions of robots doing our work.
Cheap transportation.
No road accidents.
No city pollution.

We’ll definitely get there, sooner or later.
I was in China recently and I got a glimpse of this future.

The only question is when,
and whether Tesla will be the first to get there.

Yesterday at Analytics Summit, someone asked the amazing Simo Ahava how he envisions web analytics in the next 5–10 years.

His answer was spot on:

“I don’t know what’s gonna change, but I do know what’s not going to change: team silos and broken collaboration.”

True.
But it got me thinking…

💡What if we won’t even have a “web” to analyse anymore?

What if there are no browsers, no websites.

At least not in the way we know them today?

What if the future of the internet is an omnipresent LLM chatbot, a single interface that is the internet?

In that world, brands won’t build websites or apps.

They’ll just make great products, and market them through chatbots.

Think of partnerships. Licensing deals. Sponsored results.

Imagine booking a flight:
“Hey ChatGPT, book me a flight from X to Y.”
“Sure, here’s your booking.”

Maybe the backend is Booking.com, but I’ll never see it.

At that point, “web” analytics becomes brutally simple:
“How many bookings did ChatGPT send us this month?”

Unless your brand is strong enough that people ask explicitly:

“Book me a flight with Booking.com.”

Then, once again, what truly matters is your brand. Nothing else.

This is the essence of the AI Marketing revolution unfolding before our eyes, right now.

Can’t wait to see how this unfolds!

The best quote I’ve ever read.

And it couldn’t be more relevant in the age of AI.

Yesterday I said we should treat LLMs like a real human audience.

Just like with people, advertising can influence what LLMs “think” of our brand.

But here’s the problem:
we can’t see who they “are.”

We can’t target audience segments, like we do with real people.

So performance marketing, just like a laser in a pitch-black room, won’t be effective.

Brand building then becomes the smarter play to create "mental availability" inside LLMs.

The goal is to make sure you’re known, salient, and more “popular” than competitors. So will you come to “mind” when users ask about your space.

Performance marketing has massively risen in relevance in the last 20 years.

But now it's losing ground:

-> GDPR and the cookie collapse
-> AI chat disintermediating search
-> AI "stealing" traffic from websites and display ads impressions

The foundations of data-driven performance are being questioned.

That’s why brand is back!

But it’s shouldn't be brand versus performance. It never should have in fact.

They work together, now more than ever.

Brand lights up the room.
Performance laser targets once the lights are on.

Brian Chesky couldn’t have said it better.

🔔 Stay tuned!
More AI Marketing pills next Wednesday.