1) The study presents a novel recommendation delivery paradigm called Social Referral that leverages a user's social network to deliver relevant content. 2) An implementation of Social Referral was deployed and evaluated in a large online professional network. 3) The study examined the interaction between the recommender system, trusted social connections, and the end user receiving recommendations. The study found that recommendations delivered via social referrals had a significantly higher acceptance rate than direct recommendations.

1. Social Referral: Leveraging network connections
to deliver recommendations
Mohammad Amin, Baoshi Yan, Sripad Sriram, Anmol Bhasin, Christian Posse
Abstract Results
Our contributions with this study are :
1) A novel recommendation delivery paradigm
called Social Referral, which utilizes a user's social
network for the delivery of relevant content.
2) An implementation of the paradigm is deployed and
evaluated in a real industrial production setting of a
large online professional network.
3) A study of the interaction between the trifecta of the
recommender system, the trusted connections and the
end consumer of the recommendation.
Example: Social Referral for
LinkedIn Group Recommendation
Potential Members
User is nudged to
invite
1
2
Experiments
 Emails sent to referrer: 85K
 Direct Recomm Email (control grp) :16K
User’s network  Percentage of Referral Sent by
referrers: 7.8%
 Join Rate - referral: 22.5%
 Join Rate – direct recomm: 9.6%
Identifying Groups to Refer • Stronger Connection Strength
=> Higher Acceptance
 Group Quality:
 Better Recommendation
 Strongest Connection: ReleVance
=> Higher Acceptance
 Combination: Summary
 Stronger Connection Strength
=> Higher CTR
We present Social Referral, a novel
Email Experiment:  No significant change in recommendation delivery mechanism which
Social Referral recommendation relevance for leverages a target user's social network for the
vs referrees after filtering by effective delivery of relevant recommendations
referrers! to the user. To our knowledge, most previous
Direct Recommendations work has been focused on incorporating social
network to improve recommendation
Discussion: relevance or explain recommendations. Little
Social Referral has been explored on leveraging social
vs network for the delivery of recommendations.
Direct Recommendation
We carry out a large scale user study with real
 2 step loss for social referral vs one- world data that demonstrated the effectiveness
step loss for direct recommendation. of social referral. Our study showed that
referrers are receptive to passing along the
VS.  Different user experience:
Warm recommendation (social
recommendations to their connections, and

recommendations delivered via social referrals
referral) vs cold recommendation are more than twice as likely to be accepted
(direct recommendation) by target users than those directly delivered to
 For referrs, we could piggyback the them.
social referral process at various
engaging points for free