User Flows & Navigation Graphs

Construct directed acyclic visitor graphs to evaluate route transitions, drop-off rates, and multi-step conversion funnel performance.

Directed Acyclic Graph (DAG)Markov Transition ProbabilitiesCohort Sliding Windows

Directed Navigation Graphs

Tracium models user sessions as directed transition graphs where vertices represent distinct URL endpoints and edges represent sequential route changes weighted by volume and latency.

The visual Sankey topology renders first-order Markov transition probabilities across your site architecture:

[ Ingress Point: / ] (100% Volume)
         │
         ├─── (62% Transition) ──► [ /docs ] ── (38%) ──► [ /docs/quickstart ]
         │
         ├─── (24% Transition) ──► [ /pricing ] ── (71%) ──► [ /register ]
         │
         └─── (14% Terminal)   ──► [ Session Exit Boundary ]

Pipeline Configuration & Step Ordering

Conversion funnels enforce an ordered state pipeline across user journeys:

Step Definition: Bind target route patterns (e.g. /pricing) or custom telemetry event IDs (e.g. checkout_initiated).
Sequence Enforcement: Evaluate whether users followed strict consecutive routing or non-linear multi-session traversal.
Window Attribution: Bound step progression to a 24-hour or 7-day conversion window.

Attrition Calculation & Conversion Rates

Step-to-step attrition is computed as a normalized completion ratio:

Conversion_Rate(Step_N) = ( Unique_Sessions(Step_N) / Unique_Sessions(Step_N-1) ) * 100

Sudden drop-off deviations exceeding 2.5 standard deviations trigger automated anomaly notifications within Tracium Intelligence.