There is no shortage of discussion about how AI is changing marketing.
Most of it starts with the technology: AI agents, automation, AEO, content generation, personalization, new search behaviours and new platforms.
All of those things matter. But I think starting with AI risks missing the more important change happening underneath it.
We need to start with the buyer.
B2B buying journeys have been fragmenting for years. Buyers do more research independently, move between channels, consult peers, consume content anonymously, talk to analysts, participate in communities and increasingly use AI to research problems, evaluate alternatives and synthesize information.
A meaningful portion of the buying journey can now happen without a buyer ever raising their hand – or even clicking.
That creates an obvious challenge for marketers. But I think it creates a much bigger challenge for organizations.
The centre of gravity has to be the customer.
As the buying journey becomes harder to observe, understanding the customer becomes simultaneously harder and more important.
A healthy go-to-market system needs the ability to continuously understand, surface and action the voice of the customer across every experience. That means thinking well beyond the annual research project or NPS survey. Customer intelligence is being generated constantly through sales conversations, customer service interactions, search behaviour, website activity, account intelligence, win/loss analysis, social conversations and direct research.
The problem is that much of this intelligence remains trapped wherever it was created.
Sales knows something marketing doesn’t. Customer experience sees a pattern that never reaches the brand team. Marketing generates engagement data that doesn’t influence the sales conversation. Customer research produces a presentation, gets discussed for an hour and gradually disappears into a shared drive.
We have more customer signals than we’ve ever had. I’m less convinced that most organizations have become proportionately better at listening to them.
A connected organization should be able to surface those signals, make sense of them and put that understanding back into the business. Something learned in a sales conversation might change positioning. Search behaviour might expose a disconnect between how customers describe a problem and how the company talks about it. Customer service data might reveal an experience issue that is undermining the brand promise. AI visibility might reveal that the market understands the company very differently than the company understands itself.
In a healthy system, those aren’t isolated observations. They become inputs.
The result is a continuous loop from customer signals to shared understanding, coordinated action, experience, new signals and learning. That feedback loop may ultimately be a more useful way to think about modern GTM than the funnel itself.
The funnel describes how we hope a customer will progress. A connected growth system describes how the organization learns and responds.
We built organizations around functions. Customers experience systems.
Most organizations have perfectly rational reasons for being structured the way they are. Brand has a mandate. Demand generation has another. Sales has another. Customer experience, data, technology, product marketing and operations may each have their own leaders, budgets, platforms, partners and KPIs.
Specialization has created enormous value. But somewhere along the way, we’ve become very good at optimizing the pieces, and I’m not sure we’ve been nearly as good at optimizing what happens between them.
Customers don’t experience our organizational structures. They experience one company.
They don’t care who owns the CRM, whether brand and demand report to different leaders, which agency created which experience or where responsibility shifts from marketing to sales. They encounter a collection of experiences that, taken together, form their understanding of the organization.
Increasingly, AI does something similar. It draws from information distributed across websites, content, third-party sources and other signals to form its own representation of companies, products and categories.
The integrity of the whole system therefore matters more than ever.
AI won’t fix a fragmented growth system.
One of the things that interests me most about AI is how much of its potential depends on context: good data, connected systems, accessible knowledge, consistent language, clear positioning and useful customer intelligence.
Now consider the opposite environment: customer information fragmented across systems, poor CRM hygiene, different business units describing the same company differently, product information scattered across repositories, sales intelligence that rarely makes its way back to marketing, customer insights sitting in presentations, and content disconnected from actual buying behaviour.
Then add an expanding collection of AI tools.
AI doesn’t magically resolve that fragmentation. In some cases, it may simply accelerate it.
AI can make a healthy growth system dramatically more capable. It can also make a fragmented one dramatically faster.
That’s why I suspect many organizations will discover that their biggest AI challenge isn’t selecting the right technology. It’s getting the organization underneath it ready.
That requires looking beyond individual use cases and asking harder questions about the underlying system. Is the data trustworthy? Can knowledge move across functions? Are teams operating from a common understanding of the customer? Are systems interoperable? Is the brand expressed consistently enough for both humans and machines to understand it? Can what the organization learns in one part of the customer journey influence what happens somewhere else?
Those questions aren’t really AI questions at all. They’re organizational ones.
The clickless funnel makes brand more important, not less.
There’s another important consequence of the shift happening in search and discovery.
We’re entering a world in which customers can learn considerably more about companies without visiting their websites. Search engines increasingly answer questions directly, and the behavioural impact is already becoming visible. Pew Research found that users encountering Google’s AI summaries were roughly half as likely to click a traditional search result as those who did not. Recent clickstream research suggests the shift is accelerating, with 68% of U.S. Google searches now ending without a click.
AI platforms synthesize information from multiple sources. Buyers can compare approaches, develop shortlists and form opinions before a company has any direct indication they’re interested.
We can debate exactly how quickly the “clickless funnel” will develop, but the direction seems fairly clear.
That will understandably lead organizations to invest more aggressively in AEO, structured content and AI visibility. They should. But I think there’s a danger in treating this as another optimization problem.
If buyers are making more decisions before clicking, mental availability, reputation, distinctiveness and trust become more valuable – not less. Being surfaced in an AI answer matters. Being one of the companies the buyer already knows to ask about may matter even more.
That changes the nature of the investment required. The answer isn’t brand or demand generation or AEO. It’s a more dimensional investment in brand building, discoverability, content, demand creation and customer experience, supported by the data and technology required to make them work together.
In other words, the clickless funnel doesn’t diminish the importance of brand. It raises the stakes.
When fewer interactions happen on properties we control, what customers already believe about us – and what the broader information ecosystem understands about us – becomes increasingly consequential.
Which raises an uncomfortable question: who owns it?
This is where I think organizations may have to challenge some long-standing assumptions.
If brand, demand, customer experience, AI, data, CRM, sales intelligence and technology increasingly need to operate as parts of the same growth system, are the traditional divisions of responsibility still the right ones?
Where does marketing end and sales begin? Is CRM infrastructure primarily a technology responsibility or a growth responsibility? Should customer experience operate separately from brand when the experience itself is one of the most powerful expressions of the brand? Who owns AI-driven discovery? Who owns the integrity of customer data? Who ensures that what sales learns changes what marketing says?
And perhaps most importantly, who is accountable for the health of the whole system?
I don’t think the answer is necessarily putting everything under one executive. Nor do I think specialization is the problem. Deep functional expertise remains essential.
The problem is allowing organizational separation to become operational fragmentation.
That distinction matters because reorganizing companies is expensive, disruptive and frequently doesn’t solve the underlying problem anyway. Moving boxes around an org chart doesn’t automatically create shared data, better feedback loops or a common understanding of the customer.
The more useful question is whether the organization can operate as a connected system regardless of where the reporting lines sit.
Defragmentation doesn’t mean centralization.
A modern enterprise can have highly specialized teams with different leaders, budgets and responsibilities while still operating as a connected growth system.
The objective isn’t organizational uniformity. It’s connectivity: shared customer intelligence, consistent positioning, connected data and interoperable technology, coordinated activation and measurement, and continuous feedback. Most importantly, it requires an operating model capable of turning what the organization learns into action quickly.
That’s what creates agility. Not simply moving faster, but becoming better at sensing what is happening, interpreting it and responding coherently.
This becomes particularly important in an environment where customer behaviour, technology and competitive dynamics are changing simultaneously. An organization can have enormously talented people in every function and still respond slowly if information has to fight its way across departmental boundaries before anyone can act on it.
Seen this way, defragmentation isn’t primarily an efficiency exercise. It’s about increasing the organization’s capacity to learn.
And that may become one of the most important characteristics of a healthy GTM function.
We’ve spent years building stacks. Now we need to build systems.
Over the last decade, organizations have invested enormous amounts in capability: CRM, marketing automation, CDPs, intent data, ABM platforms, analytics, content systems, sales enablement, customer experience platforms and, increasingly, AI.
Most of those investments made sense individually. But possessing a sophisticated collection of technologies isn’t the same as possessing a connected growth system.
The next question may therefore be less about what capability we’re missing and more about what we’ve already built that isn’t working together.
Can customer understanding influence brand strategy? Can brand strategy shape demand? Can demand signals improve sales conversations? Can sales intelligence improve content? Can customer experience change positioning? Can all of that learning flow through the data and technology infrastructure quickly enough to influence what the organization does next?
And can the organization measure the effect across the system rather than simply demonstrating that each individual function performed its assigned task?
That’s a much higher standard for integration.
It’s also a much more useful one.
The next transformation may be about integration, not addition.
For years, digital transformation encouraged organizations to add capabilities: more channels, more platforms, more specialists, more data and more automation.
AI will certainly add another extraordinary layer of capability. But it may also force us to confront the accumulated complexity underneath it.
The organizations that thrive in the next phase won’t necessarily be the ones that adopt AI fastest. I suspect they’ll be the ones willing to reconsider how growth actually happens – and redesign themselves around it.
That may mean challenging organizational boundaries, investment models, technology decisions, agency structures, data ownership and even some long-standing assumptions about what constitutes “marketing.”
Through all of it, the centre of gravity should remain remarkably simple: understand the customer, surface what the organization is learning about them, make that intelligence available to the people and systems that need it, act on it, and learn from what happens next.
Because increasingly, competitive advantage won’t come from any one capability.
It will come from how intelligently all of those capabilities work together.