Activity Feed Personalization 101: Top Feed Features to Improve User Engagement

Understand how Stream’s data science team helps implement personalized activity feeds, plus top feed features to boost your platform’s user metrics.

Jenna B.
Jenna B.
Published March 19, 2021 Updated March 23, 2021

Personalization comes in many flavors, and the data science team at Stream can help you build your own feeds personalization engine based on your specific needs. In conjunction with our analytics client we recommend tracking every event for every user, such as clicking on a link) we use both engagement and feed data to power and improve your app’s user experience by leveraging cutting-edge machine learning methods.

Here, you’ll learn how Stream’s data scientists approach creating a personalized feed to solve your unique challenges, and some of the powerful feeds features that we implement.

Activity Feed Personalization Process

After you’ve decided that an enhanced activity feed experience is the right option for your product, the first step to personalization is to have a kickoff meeting with your Stream account manager and our entire personalization data science team so we can understand your vision, document which feeds you want personalized, and discuss the data requirements we’ll need in order to create a customized algorithm.

For machine learning to deliver the right content, to the right person, at the right time, feed personalization requires user data. Most customers have best results if they use Stream’s activity feeds for at least one to three months prior to personalization to adequately collect user data. Then, our team can better create a model that boosts your user activity and brand engagement.

While tens of thousands of user engagement data points are ideal to develop personalized feeds, Stream data scientists can launch an accurate activity feed algorithm with just several thousand feed users.

Ongoing Personalized Feed Support

After your personalized activity feed is up and running, you’ll meet with Stream’s data science team bi-weekly to ensure the algorithm is functioning how you envisioned. Here, you can also request new features and tweaks to the feed.

These follow-up meetings will also enable your team and Stream to collaboratively improve the model. “Generally, building a personalized feed algorithm is an iterative process. It needs to be collaborative by design because our clients have the domain knowledge,” says Neha Rao, data scientist at Stream. “V1 isn’t always the best. During our bi-weekly meetings, we talk about new data to refine algorithms, and we run two models side by side to validate which one best achieves a client’s goals.”

Long term, Stream’s personalization team is available as needed to ensure the feeds personalization model continues to run smoothly.

Top Personalized Feed Features

An overview of the best-in-class feed features Stream offers to help you grow without worrying about scalability.

Discovery & Interest Feeds

Discovery feeds usually contain activity posts that are not directly available through a user’s network but that might be of interest to them. A prominent example of a discovery feed is Instagram’s Explore section, which enables users to learn about new creators and accounts they do not follow. Content on a discovery feed is typically based on:

  • Trending content
  • Activities that match a user’s interests
  • Content that is popular amongst a user’s extended (friends of friends) network.

Interest profiles are based on the types of activities users engage with — if you click many pictures of snowboarding on Instagram it will, over time, show you more content related to snowboarding. At Stream, we use our analytics package combined with the personalization offering to build a similar experience.

AI Feed Ranking

Stream’s most advanced feature, think of personalized feed ranking as a more powerful implementation of our Ranked Feeds. The ranking for personalized feeds is based on each user’s individual interest profile. It allows you to build complex ranking rules that show the most relevant content at the top of users’ feeds, such as “while you were away” content, content from users’ inner circle of friends, fresh activities, or recommended items.

For example, by using Stream’s personalized feed ranking, Healthline’s peer-supported social network apps focusing on chronic health conditions (think: breast cancer, type 2 diabetes, rheumatoid arthritis, and more) can fill members’ feeds with highly individualized content. If a user likes a post about arthritis treatments, their feed will populate with content relating to medications, alternative therapies, and posts from other users who are also interested in managing arthritis symptoms.

Follow Suggestions

If your app implements an activity feed, it likely contains a social component. Most successful social apps have a highly connected network. Stream’s Personalization Engine can help the organic growth of your app by suggesting other feeds to follow (such as users or topics) based on authority, common interests, and friend proximity.

Item Recommendations

Item recommendations are an essential feature if you need to provide users with highly relevant and valuable content. There are thousands of possible applications for item recommendations, such as recommending songs to users, matching job seekers with open positions, and recommending fashion products that are similar to items previously purchased.

Email Personalization & E-Commerce

Stream’s personalization algorithms learn user interests by understanding likes, follows, and reposts. Using this detailed interest profile, you’ll be able to email the most relevant and “sticky” content to your user base to encourage them to return to your app. For example, if a user continuously engages with content that contains makeup tutorials, Stream’s email optimization feature can enable you to send them content from accounts that sell beauty and skincare products. This feature is particularly effective for e-commerce sites with a community aspect.

Data-Driven Functionality

The above personalized activity feed features aren’t just bells and whistles. They represent data-driven functionality that can help your community develop deeper relationships with your app and your brand. By anticipating the kind of activity feed content users will resonate with, a personalized activity feed will boost key app metrics such as weekly active users and retention rate.

More good news: Stream’s feed features are continuously expanding to keep your app growing, and your community engaged.

Learn what Stream’s personalized activity feeds can do for your community! Check out our activity feed SDKs, and contact a Stream activity feed expert today.