skip to main content
10.1145/2740908.2742002acmotherconferencesArticle/Chapter ViewAbstractPublication PagesthewebconfConference Proceedingsconference-collections
research-article

Targeted Content for a Real-Time Activity Feed: For First Time Visitors to Power Users

Published: 18 May 2015 Publication History

Abstract

The Activity Feed is Etsy's take on the ubiquitous "web feed" - a continuous stream of aggregated content, personalized for each user. These streams have become the de facto means of serving advertisements in the context of social media. Any visitor to Facebook or Twitter has seen advertisements placed on their web feed. For Etsy, an online marketplace for handmade and vintage goods with over 29 million unique items, the AF makes the marketplace feel a bit smaller for users. It enables discovery of relevant content, including activities from their social graph, recommended shops and items, and new listings from favorite shops. At the same time, Etsy's AF provides a platform for presenting users with targeted content, as well as advertisements, served alongside relevant and timely content.
One of the biggest challenges for building such a feed is providing an engaging experience for all users across Etsy. Some users are first-time visitors who may find Etsy to be overwhelming. Others are long-time power users who already know what they're looking for and how to find it. In this work, we describe solutions to the challenges encountered while delivering targeted content to our tens of million of users. We also cover our means of adapting to each user's actions, evolving our targeted content offerings as the user's familiarity with Etsy grows. Finally, we discuss the impact of our system through extensive experimentation on live traffic, and show how these improvements have led to increased user engagement.

References

[1]
D. Agarwal, B.-C. Chen, R. Gupta, J. Hartman, Q. He, A. Iyer, S. Kolar, Y. Ma, P. Shivaswamy, A. Singh, and L. Zhang. Activity ranking in linkedin feed. KDD '14, 2014.
[2]
L. Backstrom and J. Leskovec. Supervised random walks: Predicting and recommending links in social networks. WSDM '11, 2011.
[3]
D. M. Blei, A. Y. Ng, M. I. Jordan, and J. Lafferty. Latent dirichlet allocation. JMLR, 2003.
[4]
M. Burke, C. Marlow, and T. Lento. Feed me: Motivating newcomer contribution in social network sites. CHI '09, 2009.
[5]
P. Gupta, A. Goel, J. Lin, A. Sharma, D. Wang, and R. Zadeh. Wtf: The who to follow service at twitter. WWW '13, 2013.
[6]
N. Halko, P. Martinsson, and J. Tropp. Finding structure with randomness: Probabilistic algorithms for constructing approximate matrix decompositions. SIAM Review, 53(2), 2011.
[7]
D. Hu, R. Hall, and J. Attenberg. Style in the long tail: Discovering unique interests with latent variable models in large scale social e-commerce. KDD, 2014.
[8]
J. Kleinberg. Hubs, authorities, and communities. In ACM Computer Survey, 1999.
[9]
Y. Koren, R. Bell, and C. Volinsky. Matrix factorization techniques for recommender systems. Computer, 2009.
[10]
M. Zuckerberg, R. Sanghvi, A. Bosworth, C. Cox, A. Sittig, C. Hughes, K. Geminder, and D. Corson. Dynamically providing a news feed about a user of a social network, Feb. 23 2010. US Patent 7,669,123.

Cited By

View all
  • (2016)A Joint Model for Who-to-Follow and What-to-View Recommendations on BehanceProceedings of the 25th International Conference Companion on World Wide Web10.1145/2872518.2890083(581-584)Online publication date: 11-Apr-2016

Index Terms

  1. Targeted Content for a Real-Time Activity Feed: For First Time Visitors to Power Users

    Comments

    Information & Contributors

    Information

    Published In

    cover image ACM Other conferences
    WWW '15 Companion: Proceedings of the 24th International Conference on World Wide Web
    May 2015
    1602 pages
    ISBN:9781450334730
    DOI:10.1145/2740908

    Sponsors

    • IW3C2: International World Wide Web Conference Committee

    In-Cooperation

    Publisher

    Association for Computing Machinery

    New York, NY, United States

    Publication History

    Published: 18 May 2015

    Permissions

    Request permissions for this article.

    Check for updates

    Author Tags

    1. hadoop
    2. large-scale systems
    3. recommender systems

    Qualifiers

    • Research-article

    Conference

    WWW '15
    Sponsor:
    • IW3C2

    Acceptance Rates

    Overall Acceptance Rate 1,899 of 8,196 submissions, 23%

    Contributors

    Other Metrics

    Bibliometrics & Citations

    Bibliometrics

    Article Metrics

    • Downloads (Last 12 months)5
    • Downloads (Last 6 weeks)0
    Reflects downloads up to 05 Mar 2025

    Other Metrics

    Citations

    Cited By

    View all
    • (2016)A Joint Model for Who-to-Follow and What-to-View Recommendations on BehanceProceedings of the 25th International Conference Companion on World Wide Web10.1145/2872518.2890083(581-584)Online publication date: 11-Apr-2016

    View Options

    Login options

    View options

    PDF

    View or Download as a PDF file.

    PDF

    eReader

    View online with eReader.

    eReader

    Figures

    Tables

    Media

    Share

    Share

    Share this Publication link

    Share on social media