App generation is one of the fastest growing AI spaces. Cos are scaling to $100M+ ARR at record pace, seemingly in a race to dominate the market. But we think it won't be "winner takes all." The market will segment, with different winners in different areas. Our thesis 👇

Why won't this space be zero sum? AI markets haven't played out this way. Think about LLMs - OpenAI, Anthropic, xAI, Mistral, Gemini all have superfans. Same with image + video - Midjourney, Krea, BFL, Ideogram, Veo, etc. - with further specialization via workflows.

There can be multiple winners because these markets are HUGE and use cases vary. For example - many developers using an LLM in an IDE prefer Claude, while consumers looking for a general assistant turn to ChatGPT. We think the same segmentation will occur in AI app creation.

What does this look like? We've seen the following use cases, with platforms "spiking" on each 1) Prototyping - quick & dirty implementation of functionality/design 2) Personal software - bespoke, single user apps 3) Production apps - can be scaled by teams + used by the public

Each of these categories can be further segmented by user type. Ex. A consumer with zero technical experience probably needs a different interface and workflow than a developer - even if they're both prototyping the same product! There's no "one size fits all" here.

This crossover in app gen tools shows in web traffic overlap data - what % of site X's users visited site Y? Two user types emerge: (1) Loyalists who prefer the UI, features, etc. of one platform (e.g. 82% of Replit users are only on that platform) (2) Power users active across platforms in a complementary way (e.g. the 20% of Bolt users on Lovable)

The bottom line is people want the BEST product for their use case. This means being great at generating one particular type of app > being okay at all of them. Multiple specialized players will win, focused on different integrations, constraints, and workflows. Examples: