What VCs Get Wrong About AI Moats

Venture capitalists have been asking founders the same question for decades: What's your moat?

In the age of AI, that question has become urgent. And VCs have responded with competing frameworks for defensibility. Some push portfolio companies toward technical moats that compound over time. Others have concluded that traditional moats no longer apply and velocity is all that matters.

Both perspectives contain truth. Both are incomplete.

Let's give credit where it's due. The best investors have converged on defensibility theses that are operationally useful.

They define a moat as the set of structural advantages that prevent a well-funded competitor from taking your customers, margins, or position within a few years. They care less about uniqueness at launch and more about how defensibility widens as the company scales.

The moat types they push are familiar: network effects, proprietary data, switching costs, brand and trust, cost or distribution advantage. Each of these can compound. Each can be measured. Each can be articulated to the next investor in the funding chain.

The operational guidance is equally clear. Design roadmaps that deliberately deepen one or two moat types. Show moat proof through metrics: organic versus paid acquisition, retention and expansion, CAC payback. Build capabilities first that later crystallise into moats.

This is good advice. It creates discipline. It forces founders to think beyond features.

But it's missing something.

Over the next year, Roku predicts that 100% of the streaming audience will see ads. For growth marketers in 2026, CTV will remain an important “safe space” as AI creates widespread disruption in the search and social channels. Plus, easier access to self-serve CTV ad buying tools and targeting options will lead to a surge in locally-targeted streaming campaigns.

Read our guide to find out why growth marketers should make sure CTV is part of their 2026 media mix.

Learn more.

A recent piece from Insight Partners, one of the most sophisticated growth-stage investors, crystallises one strain of current VC thinking on AI-era defensibility. They identify two categories where vertical AI companies can build real moats: workflow depth and proprietary data.

Workflow depth means owning the system of record. Platforms users live in all day, where work is created, shared, and stored. The argument is that a chat window fine-tuned by prompt engineers won't create defensibility. The companies that endure are the ones that embed AI into core workflows.

Proprietary data means access to specific, hard-to-replicate information: custom integrations, customer data that feeds training loops, hardware and sensor capture. Foundation models can do powerful things with public data. They can't access your private corpus.

These are real moats. Filevine owns the legal workflow. Motorq has direct integrations with auto manufacturers that took years to build. ExaCare's AI recommendations create reinforcement loops that improve with every human decision.

The thesis is credible: own the workflow, own the data, and you own something AI can't easily replicate.

But there's another school of thought that's given up on traditional moats entirely.

Andreessen Horowitz takes a more radical position. In their view, there is no moat in consumer AI right now. The landscape is simply evolving too quickly. When foundational models shift every month, it's become nearly impossible to build methodically.

Their conclusion: what matters is velocity. How fast you can launch, gain traction, and seize mindshare. Early distribution is everything.

They use a vivid metaphor: launching an AI company today is like ejecting a pigeon into the sky and willing it to take flight. There's a whole flock of AI startups flapping around together, trying to gain enough altitude so they don't burn out and plummet. The rarefied few that blast straight up, pierce through the cloud cover, and continue to climb are the ones that make it into mainstream consciousness. They reach what a16z calls "mindshare air."

The tactical advice follows from this framing: hackathons as public performance, social experiments with big prizes, AI starter packs and alliances that cross-pollinate audiences, insider influencers who have credibility in niche communities, building in public to make users feel like participants.

The core belief is that whoever builds first, iterates fastest, and distributes best wins.

But here's the most interesting part of their argument. When describing the companies that are pulling away from the pack, a16z notes: "By establishing a 'cult of diehards,' leading AI companies are dredging moats rooted in community and hands-on experimentation."

They're circling the truth without fully grasping it.

Insight Partners' framework is fundamentally technical. Systems. Data. Integrations. The moats they describe exist in infrastructure, not in human behaviour.

They mention "customer relationships" and "trust" in passing. They note that the winning companies will be "the most human," the ones that truly understand how their customers work. But they don't explain how to build that. It's treated as an outcome, not a strategy.

a16z gets closer. They've recognised that traditional moats don't hold in a fast-moving market. They've identified that community and "cults of diehards" matter. But they treat community as a distribution tactic rather than an architecture. Their advice is about creating viral moments and passionate early adopters without explaining how to convert momentum into lasting defensibility.

Both frameworks share the same blind spot: they're advising companies on how to own the workflow, the data, or the moment, but not how to own the relationship.

And in an AI-commoditised market, the relationship is the only layer that's truly defensible.

Consider the moat types VCs push:

Network effects. Each new user makes the product more valuable to others. But network effects can be replicated. LinkedIn has network effects. So does every professional network that's tried to unseat it. Network effects create stickiness, but they don't create belonging.

Data moats. Proprietary data improves models over time. But data advantages erode. Foundation models get better. Synthetic data gets cheaper. The corpus you spent five years building might be matched in eighteen months by a competitor with better architecture.

Switching costs. Workflows, integrations, and learning curves make churn painful. But switching costs are friction, not loyalty. Users tolerate friction until something better comes along. Then they leave anyway.

Brand and trust. Perceived reliability matters, especially in AI and fintech. But brand is expensive to build and easy to lose. Trust is fragile. Neither creates the kind of stickiness that survives a superior competitor.

Each of these moats can be measured. Each can be articulated in a pitch deck. And each can be overcome by a well-funded competitor with enough time and capital.

a16z is right that velocity matters. In a market where everything is shifting, the ability to ship fast and capture attention creates real advantage.

But momentum without structure dissipates. The pigeon eventually gets tired.

What happens when you stop shipping weekly? What happens when a competitor with more capital outspends you on distribution? What happens when the next viral moment belongs to someone else?

The "cult of diehards" a16z describes is real. But cults need architecture. They need rituals, status systems, identity markers, and relationship infrastructure. Without these, diehards drift away when the next exciting thing launches.

Momentum gets you attention. Community converts attention into belonging. And belonging is the only defensibility that survives commoditisation.

Community is not a marketing channel. It's not a Slack workspace. It's not engagement metrics or NPS scores. It's not even a "cult of diehards," though that's closer.

Community is the structure that makes behaviour change, learning, and loyalty spread faster inside and around a product than any rival can match.

When you build community correctly, you create:

Switching costs that don't travel. Shared language, relationships, rituals, and status exist only inside your ecosystem. A user who leaves doesn't just lose access to features. They lose their standing. They lose their people. They lose the identity they've constructed.

A compounding intelligence loop. Members co-create playbooks, templates, frameworks, and stories that you own, not the market. Every contribution makes the community more valuable. Every valuable community attracts more contributors.

Non-transferable status. Someone who has spent two years building reputation in your ecosystem loses that standing completely if they switch. Stack Overflow's power isn't the Q&A format. It's that developers have invested years in reputation they cannot export.

Tribal identity. When users define themselves through membership in your community, leaving means abandoning that declared self. AI can replicate your features tomorrow. It cannot replicate the identity someone spent months constructing.

This is the layer VCs are missing. Technical moats create friction. Momentum moats create attention. Community moats create belonging. And belonging is the only defensibility that survives commoditisation.

To be clear: this isn't an argument against technical moats or momentum. Workflow depth and proprietary data matter. Speed and distribution matter. The companies that win will own systems of record, accumulate data advantages, and ship faster than competitors.

But technical moats without community moats are incomplete. They create switching costs without creating loyalty. They capture data without capturing hearts.

And momentum without community is unsustainable. You can't keep flapping forever. Eventually you need to build something that holds altitude on its own.

The complete defensibility stack looks like this:

Most companies focus on layers one through three. They ignore layer four entirely, or treat it as a marketing function. That's a strategic error.

The businesses that survive AI commoditisation won't just have the best workflows, the richest data, or the fastest shipping cadence. They'll have something competitors cannot replicate no matter how much capital they deploy: a community of users who belong.

VCs tell founders to make the moat visible in behaviour and metrics. The same applies to community moats, but the metrics are different.

Instead of just tracking retention, expansion, and velocity, track:

These metrics don't show up in standard SaaS dashboards. But they're the leading indicators of defensibility that survives feature parity and outlasts momentum cycles.

AI is flattening the competitive landscape faster than founders realise. Features are becoming commodities. But community moats, identity, contribution, relationships, status - compound.

If you want to build a business AI can’t replicate, now is the moment to start.

Here’s what you can do next:

👉 Join the waitlist for the BuildMoat Defensibility Workshop (January)
A half-day working session where we diagnose your moat and design your community architecture.

👉 Follow me here on LinkedIn
I share weekly insights on building AI-proof defensibility, community strategy, and the future of product.

👉 Learn more about the MOAT Framework. The Framework is a proven, adaptable system for building sustainable, defensible community infrastructure that lowers acquisition costs and creates long-term competitive advantage, successfully demonstrated through Broxtown and Hertstown.

👉 Subscribe to The BuildMoat newsletter
For community-based moat insights, deep dives, frameworks, and real-world examples you can subscribe immediately.
🔗 the-moat.beehiiv.com

The companies that build now will own the next decade.
The ones that don’t… won’t.

We appreciate you taking the time to read this. Building community isn't just theory for us, it's a framework we've developed and refined with companies at every stage.

Have questions about adopting The Moat framework for your business? We'd love to help. Book a discovery call to speak directly with one of our consultants. We'll audit your current community engagement process and show you exactly where the opportunities are.

Did someone forward this to you? Subscribe here to get strategic insights on building community as your competitive advantage, delivered directly to your inbox.

Keep building your community until next time.

Build Moat
Where Community Becomes Your Competitive Edge