By SC Moatti

If your pitch still leads with “we use AI,” you’re describing infrastructure, not a business. Consider that 97% of the products nominated for this year’s Products That Count Product Awards are deeply integrated with AI, highlighting the extent to which the days of AI as a differentiator are officially gone.

The real moat investors are seeking today is something many founders cannot identify: What will survive a well-funded competitor who could launch tomorrow with a better model.

We analyzed Crunchbase data on 576 venture-backed, AI B2B companies that raised $50 million-plus rounds since the start of 2025 using Hamilton Helmer’s 7 Powers framework and layered in insights from Products That Count’s 600,000-plus product leader community.

The data gives a clear answer: When building is nearly free, the only moats that hold are those that a model cannot generate.

Two of them work without requiring you to out-raise OpenAI. Two of them are traps. And one isn’t a game most founders are playing.

Counter-positioning: The moat that costs nothing to defend

Counter-positioning is what happens when a newcomer builds a business model so structurally different that the incumbent can’t copy it without destroying their own economics. Netflix versus Blockbuster is the canonical example. Blockbuster could have matched the subscription model, but doing so would have gutted late-fee revenue, which kept their stores alive. So they didn’t, until it was too late.

In the AI era, this power is rare and underutilized. Only 5% of companies in our dataset leverage counter-positioning. Investors price that scarcity at a median enterprise value of 5.3x per dollar raised — the highest multiple of any power in the analysis.

The pattern shows up in vertically integrated AI insurers that sell directly to employers, a model traditional brokers can’t replicate without cannibalizing their own relationships and underwriting margins. It shows up in AI-native revenue management systems that would gut the high-margin consulting revenue legacy vendors depend on if those vendors tried to match them.

The incumbent sees the threat. They choose not to respond. That rational inaction is the moat.

For founders, the diagnostic question is this: Could a well-resourced incumbent copy your model if they wanted to? The right signal is answering “technically yes, but it would cost them more than it would cost us to build.”

Network economies: The moat that builds itself

Network economies arise when a product becomes more valuable to each user as more users join. At scale, this tends toward winner-take-all outcomes within the network’s boundaries, whether in geography, professional context or industry vertical. LinkedIn is the textbook case: More recruiters attract more candidates, which in turn attract more professionals, which in turn attract more recruiters.

In our dataset, network economies appear in only 5% of companies, but command a 4.2x multiple. That makes it the most capital-efficient path to a strong valuation premium available to founders today.

The B2B variant here is particularly underappreciated. Rather than individual users as nodes, the network connects companies: brands to factories, advertisers to audiences, platforms to partners. Every new participant makes the network more valuable for every existing participant.

The data that accumulates across those interactions — costing, production cycles, audience behavior — compounds in ways that become progressively harder to replicate.

Getting both sides of a two-sided market to commit simultaneously is hard. But founders who solve the cold-start problem own something that a well-funded competitor with a better model still cannot buy.

What looks like a moat but isn’t?

At 44%, cornered resources like proprietary data, unique IP and exclusive access command the second-highest prevalence in our dataset. But they also have the worst multiple: 2.6x. Because investors have watched too many proprietary datasets get eroded by foundation models and synthetic data, if a data advantage doesn’t compound in ways that get harder to replicate over time, it can’t be considered a moat.

Switching costs are the most crowded power at 37%, and look like a moat because customers really don’t leave.

But the cost of building them creates a problem, because the requirement for deep enterprise entanglement — baked-in instrumentation, institutional memory, rearchitecture risk — means expensive sales cycles before the stickiness kicks in. The multiple is comparable to network economies (4x), but the capital required to reach it is roughly 10x higher. It’s a viable path for founders who build a product-led growth motion to reduce that cost, or who engineer a reason for users to collaborate on the platform, converting switching costs into network economies over time.

Scale economies are not the game most founders are playing. The median multiple, excluding OpenAI and Anthropic, collapses from 6.1x to 3.2x, and 88% of the category’s capital belongs to those two companies. Believing your unit economics improve with growth does not equate to building a scale moat. The gap between the two is measured in billions of dollars that most startups will never raise.

The only question that matters

Every company in this dataset has AI in its product. The ones commanding premium multiples have built something underneath the AI that a model cannot generate on its own.

That something is structural. It lies in the business model design or the network architecture, rather than model quality, feature set or data volume. It answers the question every founder should be able to state in one sentence: What about my business would survive a competitor who starts today with more capital and a better model?

If you can’t answer that question, you’re building a product. The founders commanding 4x to 5x multiples are building a power.

---

 SC Moatti is the founding managing partner of Mighty Capital and board chair at Products That Count. As a venture capitalist honored on the Kauffman Top 30 Index and Power100, she invested in pioneering companies including Amplitude, Netskope and Groq. She earned a reputation for developing products that people love during the cloud and mobile era, when she built products that billions of people use at Meta and Siebel Systems, won industry awards and nominations from The Wall Street Journal and the Emmy’s Foundation, and wrote an award-winning bestseller on what makes a great product. She holds a master’s in electrical engineering and a Stanford MBA, and is a Kauffman Fellow and member of Young Presidents Organization.

Illustration: Dom Guzman