Why App Layer Startups Should Stop Lying About Their Moat

The article confronts a tension that every application-layer founder faces: investors demand a moat on slide three of the deck, but for most app companies the defensible advantage cannot exist yet. Infrastructure companies can point to novel architectures or proprietary datasets at founding, what the author calls leading moats. Application companies, by contrast, rarely possess such technical differentiation out of the gate. Their moats must be earned through years of execution, and that reality clashes with the standard venture framing of immediate competitive advantage. The problem is not that the company lacks defensibility; it is that the timeline for building it is longer than a pitch deck allows.

To make the distinction operational, the article draws on Hamilton Helmer’s 7 Powers framework. Scale economies, brand, and switching costs are lagging moats: they require volume, time, and embedded customers. Counter-positioning, cornered resources, and process power can be leading if present at founding, but most app-layer startups do not have them. Salesforce is the canonical example. Siebel had better technology in 1999, but Salesforce won on sales muscle, brand, and a decade-long head start in the cloud-CRM category. Snowflake took the other path: separating storage from compute was a true leading moat that bought the runway to build lagging moats through marketplace distribution and pipeline switching costs. Network economies, like those of Slack, Figma, and GitHub, are earned only after scale arrives.

The takeaway for builders is practical and honest. If you are building at the application layer, the correct answer to the moat question is not a fabricated technical differentiator. It is: we are building one. A lagging moat requires focus, execution, and a market moving faster than incumbents can defend. Founders should not feel pressure to claim a technical lead that will dissolve within a year. Instead, they should design the business from day one to eventually earn scale economies, brand, embedded workflow switching costs, or network effects. The moat shows up later, but it is no less real once it arrives.