AI app defensibility: useful habits, switching costs and value density
As software becomes easier to create, Phil Carter argues that lasting value depends on why people return. His framework focuses on useful habits, a clear product experience, trusted relationships and context that becomes more valuable over time.
By NewForm · Updated
Key takeaways
- Retaining users across years requires defensible moats like emotional trust, reputable branding, and network effects rather than quick feature additions.
- Users evaluate software through value density rather than total feature count, making single-action habit loops more effective than crowded interfaces.
- Maintaining product clarity requires actively deprecating features that fail to pull their weight as new capabilities are introduced.
- Switching costs in AI environments depend on whether value lives in accumulated personal context or routine operational data that agents can easily port.
The retention challenge in a generative software market
In our view, creating sustained product retention is far more demanding than launching effective marketing campaigns. Generative tools make it relatively straightforward to deploy ad creative and spin up functional software prototypes within hours, but rapid deployment does not guarantee that users will return over years.
Phil Carter points out that as the barrier to producing software falls, enduring market defensibility relies on elements that competitors cannot replicate overnight. Products build staying power when they establish emotional connection, develop a reputable brand that users trust, and cultivate network effects that reinforce continued usage.
Why value density outperforms feature volume
The assumption that adding more features automatically creates more value often backfires. Carter frames this through user psychology: because people possess limited energy and attention spans, they evaluate software based on value density—the speed and ease with which they realize utility relative to cognitive effort.
When teams continually pile on capabilities without restraint, feature bloat obscures core navigation and leaves users struggling to locate value. Carter observes that many enduring consumer applications initially succeeded by narrowing their scope to a single primary screen and a compounding atomic habit.
To protect value density over time, product development must include regular pruning alongside new releases. If a team introduces new capabilities, it should systematically evaluate existing workflows and deprecate those that no longer justify the interface real estate.
Switching costs in an agentic landscape
Whether artificial intelligence increases or lowers switching costs depends heavily on the category and specific use case. In products where a user builds extensive personal session history and an emotional connection over weeks or months, moving to an alternative product becomes increasingly difficult because that accumulated context cannot easily be duplicated.
Carter also considers whether agents could reduce the work of migrating operational data and configurations between products. That is a conditional scenario, not a claim that such migrations are currently reliable in every setting. His framework asks what customers would lose by switching: routine setup, accumulated context, a trusted relationship or something else.
Adapted from NewForm’s original videos on creative strategy and paid social.
Explore NewForm Framework →