Product StrategyEmerging Pattern

Being newer can be a technical advantage—skip legacy infrastructure and build for the future, not the past

Zapier and N8N are stuck with drag-and-drop workflow builders because that's what they built and what their customers expect. Cotera, being newer, went fully prompt-based from day one—users write what they want (like emailing a new hire) and AI figures out the workflow. This agentic approach handles ambiguous, judgment-based tasks that rigid workflows cannot. Example: 'Send Pagerduty alert if Google review mentions someone getting sick'—keyword approach either misses cases or pings constantly; AI understands intent. Being new means no technical debt, no legacy customers expecting the old interface, no sunken infrastructure costs preventing better approaches. Sometimes the best technical decision is being late enough to skip the interim solution everyone else is stuck with.

When to use

Use this when entering a market where incumbents are constrained by their own past decisions. Look for opportunities where new technology enables a fundamentally better approach that incumbents can't adopt because of technical debt or customer expectations. Don't try to copy their stack—build for where the puck is going.

Don't do this

Copying incumbent approaches because 'that's how it's done' or because investors/customers say 'we want something like Zapier but for X.' If you're copying their approach, you're competing on features/execution where they have advantages. Better to use newer technology to solve the problem differently and better.

1 Founder Who Did This

1
Coteraby Ibby Syed

Zapier/N8N stuck with drag-and-drop workflows (legacy infrastructure, customer expectations). Cotera went fully prompt-based—users write what they want, AI builds workflow. Handles ambiguous tasks drag-and-drop cannot. Example: 'alert if review mentions sickness' needs judgment, not keywords.

Result:$1M ARR competing against $5B+ (Zapier) and well-funded incumbents by using prompts/AI to solve workflow problem better than drag-and-drop can
Read full story →