Bookwise Social Activation
Page 3 of 7 — Does social engagement predict reading activity?
Social Engagement Buckets vs Reading Hours
Average and median lifetime reading hours per user, grouped by social engagement level. User count shown above each bar.
Across every social metric, more social engagement correlates with dramatically more reading. Users in the highest bracket read 10-50x more than those with zero social activity.
Individual User Scatter: Social Activity vs Reading
Each dot is a user with >0 reading hours. Both axes use logarithmic scale. Semi-transparent dots reveal density.
Following vs Reading Hours
Followers vs Reading Hours
Each dot is one user. The upward trend confirms the relationship isn't driven by a few outliers — it's consistent across the entire user base. The log scale reveals that even small amounts of social activity (1-5 connections) correspond to meaningfully more reading than zero.
First-N-Days: Social State at Day 7 vs Subsequent Reading
Addressing causality: Cross-sectional correlations can't distinguish cause from effect — heavy readers might attract followers, not the other way around. By measuring each user's social state at day 7 and their reading in the subsequent 30 days (days 7–37), we can better isolate whether early social engagement predicts future reading behavior.
Following at Day 7 → Reading (Days 7–37)
Followers at Day 7 → Reading (Days 7–37)
This is the key causal test. Social state at day 7 — before the reading outcome window begins — predicts reading in the subsequent 30 days. Users who follow 4+ people in their first week read substantially more afterward.
Hypotheses for Page 4
The correlations above motivate seven testable hypotheses. Page 4 applies statistical tests to each.
H1
Following Threshold
Users who follow ≥N people within their first 7 days read significantly more in the subsequent 30 days. Tests whether actively seeking content from others drives reading engagement.
H2
Follower Threshold
Users who gain ≥N followers within their first 7 days read significantly more. Tests whether social accountability (being watched) increases reading motivation.
H3a
Kudos Received Threshold
Users who receive ≥N kudos within their first 7 days read significantly more. Tests whether positive reinforcement from the community drives continued engagement.
H3b
Kudos Given Threshold
Users who give ≥N kudos within their first 7 days read significantly more. Tests whether actively engaging with others' reading content predicts one's own reading behavior.
H4
Combined Social Score
A composite metric (following + followers + kudos) predicts subsequent reading better than any single metric alone. Tests whether social breadth matters more than any one dimension.
H5
Bidirectional Engagement
Users who both follow others AND receive follows (bidirectional engagement) retain better than those with only one-directional social activity. Tests whether reciprocity is the key ingredient.
H6
Reading Momentum
Do reading sessions in the first 7 days predict reading in the subsequent 30 days? This serves as a control: if social metrics predict future reading above and beyond past reading alone, social engagement has independent value.