A data network effect is a network effect where a user's engagement generates data that improves recommendations/value not just for that user but for similar users — distinct from a friend/social-graph network effect, where value comes from direct connections between specific people.
Pinterest is Casey Winters' central example: it runs on a data network effect built from an interest/topic graph rather than a friend graph. Pinterest's early, relatively homogeneous user base was well served by friend-based recommendations, but as its audience diversified across geographies and demographics, interest heterogeneity among connected users grew, and friend-based recommendations broke down — Pinterest pivoted to an interest-based graph, matching users to content by topic affinity rather than by who they know.
Apply: when your product's value comes from matching users to interests/content rather than to specific people, expect a data/interest-graph model to outperform a friend-graph model as your audience scales and diversifies — and treat growing audience heterogeneity as the signal to make that pivot, not just an assumption you can defer indefinitely.
Relates to Content Loop: PLG Acquisition via Shared Artifacts as another alternative acquisition/retention engine that doesn't depend on a social graph.