The AI Startup Moats Investors Still Reward: Counter-Positioning and Network Economies

SC Moatti of Mighty Capital argues that leading with “we use AI” now describes infrastructure rather than a durable company advantage. The article points to widespread AI integration across products nominated for Products That Count Product Awards as evidence that AI itself has become common. For founders, the practical takeaway is direct: investors are looking for what survives when a better-funded competitor can launch quickly with a better model.

Counter-positioning works when a startup’s business model is so structurally different that an incumbent cannot copy it without damaging its own economics. The article uses Netflix versus Blockbuster as the familiar example, then applies the idea to AI-native companies whose models would cannibalize legacy revenue streams. Moatti’s key test for founders: could an incumbent technically copy you, but only at a cost that hurts them more than it helps?

Network economies arise when a product becomes more valuable as more users, companies or partners join. In B2B AI, the article highlights networks that connect companies, brands, factories, advertisers, audiences, platforms or partners. The value comes not just from participation, but from interaction data that compounds over time and becomes harder for rivals to reproduce.

The article warns that proprietary data, unique IP, exclusive access and switching costs can appear defensible but may not hold up on their own. Data advantages can erode if they do not compound in ways that become harder to replicate, while switching costs can require expensive enterprise sales cycles before stickiness appears. Scale economies are also framed as a difficult path for most startups, especially when the capital requirements are beyond what most companies can raise.

The central question is simple: what about the business would survive a competitor starting today with more capital and a better model? According to Moatti, premium AI companies are not relying on model quality, feature sets or data volume alone. They are building structural power through business model design or network architecture.