Engagement Maximization
Fuente:
arXiv
Enregistré dans:
| Auteurs principaux: | , |
|---|---|
| Format: | Preprint |
| Publié: |
2022
|
| Sujets: | |
| Accès en ligne: | |
| Tags: |
Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
|
| _version_ | 1866917137523671040 |
|---|---|
| author | Hébert, Benjamin Zhong, Weijie |
| author_facet | Hébert, Benjamin Zhong, Weijie |
| contents | We investigate the management of information provision to maximize user engagement. A principal sequentially reveals signals to an agent who has a limited amount of information processing capacity and can choose to exit at any time. We identify a ``dilution'' strategy -- sending rare but highly informative signals -- that maximizes user engagement. The platform's engagement metric shapes the direction and magnitude of biases in provided information relative to a user-optimal benchmark. Even without intertemporal commitment, the platform replicates full-commitment revenue by inducing the user's belief to remain ``as uncertain as'' the prior until the rare, decisive signal arrives and induces stopping. We apply our results to two contexts: an ad-supported internet media platform and a teacher attempting to engage test-motivated students. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2207_00685 |
| institution | arXiv |
| publishDate | 2022 |
| record_format | arxiv |
| spellingShingle | Engagement Maximization Hébert, Benjamin Zhong, Weijie Theoretical Economics We investigate the management of information provision to maximize user engagement. A principal sequentially reveals signals to an agent who has a limited amount of information processing capacity and can choose to exit at any time. We identify a ``dilution'' strategy -- sending rare but highly informative signals -- that maximizes user engagement. The platform's engagement metric shapes the direction and magnitude of biases in provided information relative to a user-optimal benchmark. Even without intertemporal commitment, the platform replicates full-commitment revenue by inducing the user's belief to remain ``as uncertain as'' the prior until the rare, decisive signal arrives and induces stopping. We apply our results to two contexts: an ad-supported internet media platform and a teacher attempting to engage test-motivated students. |
| title | Engagement Maximization |
| topic | Theoretical Economics |
| url | https://arxiv.org/abs/2207.00685 |