Using graph databases to perform pathing analysis - initial experiments with Neo4J

The blog post explores the use of Neo4J, a graph database, to analyze user journey data from a website in ways that are challenging with traditional SQL databases. By loading Snowplow page view event data into Neo4J, the author investigates various user pathing queries, such as the number of homepage views, bounce rates, and the most common paths leading to specific pages. The analysis demonstrates Neo4J's ability to efficiently handle complex queries that involve tracing user paths through a website, identifying common journeys, and measuring time between page visits. These analyses reveal user behavior patterns that would be difficult to discern with SQL, highlighting Neo4J's potential for facilitating open-ended pathing analysis without preconceived funnel constraints. The author notes the promising results of these experiments and plans further exploration, citing a contribution by Nicole White on using the UNWIND function to exclude looped paths, enhancing the ability to track longer user journeys accurately.