The aha moment is no longer enough for AI products
The first output can impress. The real product work starts when a user decides whether they can trust it enough to act.
Writing
Writing
Notes on AI design, growth systems, and building product experiences that feel sharper, calmer, and more intentional in use.
Your users may not need another product tour
Product tours explain the map. Activation needs a route to the first meaningful outcome.
Time-to-value is a design problem
The first experience of value is shaped by product design decisions, not just growth metrics.
Product-led growth does not mean sales-less growth
Self-serve should reduce unnecessary dependency on sales, not pretend complex buying journeys do not exist.
A design system should standardise decisions, not just components
Components standardise output. Systems become useful when they standardise behaviour and decisions.
AI will expose weak design systems
Generation tools won’t replace design systems. They’ll expose the gaps teams never documented.
Activation is not the same as onboarding completion
Completion measures the ceremony you designed. Activation measures the outcome the user came for.
Design around progress, not features
Users do not arrive asking for features. They arrive asking whether they are getting closer.
What building Omi is teaching me about designing for uncertainty
Building Omi keeps reinforcing that clarity in high-stakes products is really a question of orientation.
Branding cannot rescue an unclear product
Strong branding sets expectation. The product decides whether that expectation survives contact.