Get your free personalized podcast brief

We scan new podcasts and send you the top 5 insights daily.

Kyle says every company and designer had a different flow, so chasing features for billion-dollar customers did not transfer. They learned to get hyper-narrow, but Ridd notes everything is vibe-codable and too small a niche may not sustain the company.

Related Insights

While you gain deep empathy for one user (yourself), you risk creating a product so tailored to your expert needs that it alienates the broader market. This "market of one" paradox can lead to building powerful but commercially unviable tools for a niche group of power users.

While foundational AI models threaten broad applications like writing aids, startups can thrive by focusing on vertical-specific needs. Building for niche workflows, compliance, and deep integrations creates a moat that large, generalist AI companies are unlikely to cross.

To avoid building a niche product, Sandstone actively stopped working with early design partners. If a request didn't align with the problems of at least five other potential clients, they recognized it as a distraction and parted ways to maintain focus on scalable solutions.

Instead of serving everyone, AirOps focused on marketers who used their tool in unexpectedly complex ways. Halliday's advice is to build for the user with "high taste" who will keep rejecting your product until it meets their high standards, forcing you to achieve excellence quickly.

While competitors built all-in-one no-code platforms, UX Pilot concentrated only on AI-powered design generation. This narrow focus resulted in a faster, cheaper, and higher-quality product that resonated with its target audience and created a clear competitive advantage in a crowded market.

In the AI era, a narrow, deep product is easily replicated. Choi argues for building breadth across an entire workflow. While a single feature can be "vibe-coded" by an LLM, replicating an interconnected system with multiple integrations and steps creates a much stronger competitive moat.

Kyle says around February they could build almost anything quickly, which removed the earlier clarity of 'build for Figma.' They had to keep reassessing where opportunity was as tools like Conductor changed their own workflow every few days. Ridd says they leaned on Kyle's early trend sense.

In an era of powerful general AI models, smaller software companies' advantage is deep vertical expertise. They win by creating a product so tailored to a specific niche that it feels like a custom, in-house solution. This 'for me' experience is something large, horizontal models cannot replicate.

In a space like AI where everyone uses the same models and tech moats are rare, competing on technology is futile. The winning strategy is to ignore the competition, focus intensely on a narrow ideal customer, and build an amazing product vision tailored specifically to their needs.

Kernel's product strategy is to go deeper into company data challenges (e.g., complex APAC or government hierarchies) before going broader. This 'earn the right' approach builds customer trust by solving the core problem exceptionally well, creating pull for future product expansions rather than pushing a bloated, mediocre feature set.