Science of Engagement (KEEN 2013)

Answers the age old question of what Miley Cyrus has to do with a Presbyterian minister from the 1700s and how that impacts your social media engagement.

Transcript

1. KEEN: The Science of Engagement @hayesdavis The Science of Engagement Hayes Davis, CEO • Union Metrics
2. @hayesdavis KEEN: The Science of Engagement Why me?
3. KEEN: The Science of Engagement @hayesdavis
4. KEEN: The Science of Engagement @hayesdavis
5. KEEN: The Science of Engagement @hayesdavis
6. @hayesdavis KEEN: The Science of Engagement The 5 verbs of engagement
7. @hayesdavis KEEN: The Science of Engagement Ignore (v): To take no action as a result of seeing a message. See also: 99.9% of everything
8. @hayesdavis KEEN: The Science of Engagement Amplify (v): To forward a message to others on a network. See also: retweet, reblog, repin
9. @hayesdavis KEEN: The Science of Engagement Acknowledge (v): To perform a gesture indicating a message has value. See also: like, favorite
10. @hayesdavis KEEN: The Science of Engagement Respond (v): To send a message back to the author of another message. See also: comment, @-reply
11. @hayesdavis KEEN: The Science of Engagement Leap (v): To further interact with the content or author of a message. See also: click through, follow
12. @hayesdavis KEEN: The Science of Engagement Reaching people
13. @hayesdavis KEEN: The Science of Engagement Delivery (n): When a message is placed in a list of messages for a user to access
14. @hayesdavis KEEN: The Science of Engagement View (n): When a user actually sees a post in a browser or app
15. @hayesdavis KEEN: The Science of Engagement Exposure (n): The total times posts were delivered or viewed
16. @hayesdavis KEEN: The Science of Engagement Reach (n): The number of unique users that receive deliveries or view posts
17. @hayesdavis KEEN: The Science of Engagement Exposure >= Reach
18. KEEN: The Science of Engagement @hayesdavis
19. KEEN: The Science of Engagement @hayesdavis High Exposure (N x Reach) Low Exposure (1 x Reach) Low Reach High Reach Saturation! People get annoyed No Interest Nobody's talking Intense Interest Repeated messages Passing Interest One time messages
20. @hayesdavis KEEN: The Science of Engagement Information cascades
21. KEEN: The Science of Engagement @hayesdavis P(E|H) • P(H) P(E) P(H|E) =
22. @hayesdavis KEEN: The Science of Engagement Bayesian inference (n): A really complicated way to say that you change your mind as you learn new stuff.
23. @hayesdavis KEEN: The Science of Engagement Bayesian with Bangerz 1. You: Miley Cyrus' Bangerz is terrible 2. Friend 1 tweets that it's great 3. You: Meh, still bad 4. Friend 2 tweets that it's really great 5. You: Ok, I'll listen 6. You tweet that Bangerz is dope
24. KEEN: The Science of Engagement @hayesdavis Trending topics
25. KEEN: The Science of Engagement @hayesdavis Memes
26. @hayesdavis KEEN: The Science of Engagement Inaccuracies
27. @hayesdavis KEEN: The Science of Engagement Network Structures
28. @hayesdavis KEEN: The Science of Engagement Networks are graphs Node 1 Node 2 Node 3 Edge
29. @hayesdavis KEEN: The Science of Engagement Clique (n): A group of nodes with edges connecting each node to each other node
30. @hayesdavis KEEN: The Science of Engagement Node 1 Node 2 Node 3 A clique!
31. @hayesdavis KEEN: The Science of Engagement Connections matter You Node 2 Node 1 Node 4 Node 3 Node 5
32. @hayesdavis KEEN: The Science of Engagement Centrality (n): A measure of the importance of a node in a graph based on where it fits in the structure of that graph.
33. @hayesdavis KEEN: The Science of Engagement 4 types of centrality: • Degree: How many edges does a node have? (See also: followers) • Closeness: How close is a node to all the other nodes? • Betweenness: How often does a node bridge other nodes? • Eigenvector: How important is this node to other important nodes? (See also: PageRank)
34. @hayesdavis KEEN: The Science of Engagement Putting it together
35. @hayesdavis KEEN: The Science of Engagement Optimize • Which type of engagement do you want? • Are you saturating your audience? Not giving them enough? • Are you updating people's bayesian priors? • Are you stuck in a clique? Or are you targeting connectors?