Patterns
Use mined product patterns to find hypotheses and bottlenecks.
Patterns surface repeated behaviors, bottlenecks, and testable hypotheses mined from connected product behavior.
Pattern Types
Hypotheses
Candidate product changes suggested by repeated behavior or friction.
Bottlenecks
Points where users slow down, loop, drop off, or fail to reach value.
Signals
Events, paths, and cohorts that explain why a hypothesis exists.
Inspect Patterns From The CLI
oriyn patternsUse patterns as a starting point for research questions. A mined pattern is not the final answer; it is a signal that a specific product change is worth testing.
From Pattern To Experiment
Pick a high-signal pattern
Start with a pattern tied to activation, conversion, retention, or a known roadmap decision.
Write one change
Convert the pattern into a single user-visible product change.
Run research
Use participant summaries to decide whether to ship, revise, or gather more data.