The death of moats in marketing and martech is greatly exaggerated

SaaS, email, SEO, CDPs — supposedly dead things haven’t had this much life since zombies took over movies and TV. Hyperbole, however, remains in excellent health.

We could pull apart each of these premature obituaries in great detail, but I believe there are sources of durable differentiation that transcend all of them.

Let’s start by granting the premise: AI makes it asymptotically effortless to create almost anything digital — chiefly content and code.

It’s an inverse Field of Dreams for ideas: if they come, you can build them.

That superpower initially appeared as a gift under the banner of “do more with less.” But then we realized that everyone got the exact same superpower. And as Syndrome declared in The Incredibles, when everyone’s super, no one is.

The raw ability to build — something many marketers and product leaders prided themselves on — is now less of an edge. At least on its own.

The real art is deciding what to build.

And I don't just mean digital stuff. (Although even there, with nearly unlimited degrees of freedom, the value shifts to choosing what to make — and what not to.) I mean what you build as a brand: trust, reputation, relationships, culture, ecosystem.

These things cannot be generated from a prompt.

They cannot be “maxxed” with gobs of tokens. Not only because they don’t live fully within the grasp of AI, however clever and AGI-ish it might be. But because these things take time, in the real world, with real customers and real partners. Time that can’t be insta-accelerated with a faster model or more GPUs.

And because they require choices. There is no utopian all-things-to-all-people brand. In fact, that’s the antithesis of a brand.

Those choices — and how you follow through on them — are where differentiation is created. Compounded over time, that differentiation can become a moat.

Not an impenetrable moat — there’s no such thing. Over time, every castle can crumble. But on any reasonable horizon it can be quite effective, for a simple reason: the time and trade-offs it took to dig the moat are exactly what a competitor can’t shortcut.

Let’s look at two of the most powerful yet perennially underutilized levers for differentiation: experiences and ecosystems. The Big E’s.

(Sorry, reached for a pun there and came up short.)

Let’s start with experience — the experience you give your customers at every touchpoint they have with you (including the touchpoints they wish they had with you).

Think of all the websites, stores, restaurants, hotels, offices you’ve visited. Think of all the interactions you’ve had with salespeople, customer service reps, automated phone trees, and online chatbots. Think of all the different services you’ve used and products you’ve bought.

How many were unremarkable? How many were disappointing or frustrating?

And how many were truly great? How many did you love?

I’m going to bet a tiny fraction have landed in that last bucket. And when you did have an experience you loved, how many times did that brand continue to delight you again, and again, and again?

That is the elephant-sized opportunity in the room to differentiate and earn a reputational moat.

For all the talk of CX (customer experience) in marketing and martech, most touchpoints have not been crafted with an obsessive passion for customer delight. Because it’s hard. The span of potential interactions is vast. Making them great requires care. Making them consistent with your larger brand — a coherent position you want to own and continually reinforce in the mind of your customer — requires coordinated commitment across teams, systems, policies, and incentives. And the conviction to make trade-offs that reinforce that position.

You can’t promise simplicity while letting every product team crowd the experience with its own priorities. You can’t promise personal attention while rewarding service reps only for how quickly they end the conversation. Your brand has to survive contact with your operating model.

AI can help enormously here, making experiences possible that would previously have been impractical or unaffordable. But a competitor’s access to the same AI doesn’t automatically give them the same organizational commitment. Delivering your experience might require choices that conflict with their priorities and economics. Not because they’re stupid. Because they’ve built a different business.

That’s why copying an interface isn’t the same as copying an experience. The interface can be vibe coded. Changing how an organization works to consistently deliver the promise behind it? That’s a much deeper vibe.

And then customers have to experience it again, and again, and again. Until it’s not just what the brand says, but what customers confidently expect.

That’s a reputation you earn, not an asset you generate.

Your ecosystem is bigger than your distribution channels. It’s the network of businesses and individuals who shape how customers discover you, choose you, and get value from what you offer.

For a bicycle brand, that might include retailers, repair shops, riding clubs, coaches, and race organizers. For a hotel, restaurants, tour operators, transportation services, wedding planners, and past guests helping future ones plan their trips. These connections influence not just who buys, but how much more they can do with what they’ve bought.

Then there are creators, press, analysts, and other influential voices. The communities around your brand — including those you didn’t organize and may not even recognize. And customers themselves, through what they share, recommend, and teach one another.

Some connections are tightly woven into your offering. Others are loose associations that spark a recommendation or an unexpected opportunity. The combinatorial possibilities are astronomical. Even brands connected to many of the same participants can have very different ecosystems, depending on how those relationships work. When participation attracts more participation and increases the value for others, those relationships can generate network effects.

I saw this firsthand building HubSpot’s technology partner ecosystem. It took years. A productive ecosystem isn’t something you announce into existence. You earn participation by making it worthwhile for others to invest their time, resources, and reputation alongside yours. And then you keep earning it.

Throw as many AI agents at that as you like. They may help people find each other, coordinate work, and create new offerings. But they don’t automatically give you another business’s commitment, a community’s enthusiasm, or a customer’s willingness to recommend you. Those participants have their own goals, choices, and perceptions of you they’ve developed over time. However digitally connected, an ecosystem remains a deeply human web.

Which also makes it messy. Partners and communities have an inconvenient property of not reporting to you. Their priorities don’t necessarily align with your quarterly plan. Cultivating shared opportunity takes patience, investment, and a willingness to create value you don’t capture entirely yourself.

It’s easier to focus on your internal teams and operations. Things you can direct, measure, and hold accountable on your own timetable. Which is why ecosystem strategies so often play second fiddle.

But that difficulty — and the temptation to underinvest because of it — is precisely where the opportunity lies. While everyone races to accelerate what they can produce themselves, you can develop advantages that depend on what others choose to do with you.

A competitor can approach the same people and businesses. Earning the same place in their world? That’s a much longer game.

Look back at the two E’s and notice what they have in common. Neither is rooted solely in what a model produces. They’re forged from choices, follow-through, and time. Deciding who an experience is for and holding that line across your whole operating model. Earning a place in other people’s worlds and then keeping it.

A faster model can help with that work. It doesn’t make those advantages instantly reproducible.

That’s the flaw in the death of differentiation argument. Even if the cost of building AI-producible stuff goes to zero, not everything else goes to zero with it.

Which puts me in the odd position of finding a silver lining in Martec’s Law, my cheeky observation that technology changes exponentially while organizations change logarithmically. I’ve spent a career framing that as a problem. It is one. But it also suggests a corollary.

Martec’s Corollary: The same forces that make organizations slow to change can make their advantages slow to copy.

Not because being slow is an advantage. But because reproducing your experience or ecosystem requires a competitor to change how their own organization works. Different priorities, incentives, relationships, commitments. They may acquire the same technology and mimic you with prompts, but the logarithmic curve of organizational change can only bend so far so fast.

So the question is: what are you building that your competitors can’t simply prompt into existence?

Scott

P.S. Coming to Dreamforce next week? Please join me for a fireside chat session I'm having with Doug Tallmadge, CEO of Gradial, about “life after martech” — or at least life after martech-as-we’ve-known-it. We’ll be speaking Tuesday, September 15, 1pm at Moscone North, LL, Campground, Theater 2. (You need to reserve a seat, as space is limited.)

After the session, we’ll be meeting with folks for a more intimate discussion at Gradial’s booth #205, in that same campground, between 2pm-3:30pm. To snag a spot for that, sign up here. Hope to see you there!