Sorry. This page is not yet translated.

Engagement Part II: Content Production

Data Science Team

In the introductory post for this series, we explained why engagement matters and how production-consumption systems work in the context of news feeds. In this post, we explain how to understand, track and increase content production.

Content production is the single most important factor that influences engagement. If no content is produced, there will be nothing for your users to consume and give feedback on. Visitation will decrease from multiple times per day to daily, weekly, monthly, etc., and your product will ultimately die (see Figure 1).

15 1

Content production can be broken down into three phases: creation, capture and sharing (see Figure 2). Creation happens every moment, around the world?—?a simple conversation between two people can be considered content creation. Of course, most of this content is not captured, even today. But since the introduction of mobile phones with cameras, the amount of content?—?and in particular, visual content?—?that is captured has increased dramatically.

Many social products, such as Instagram and Snapchat, further emphasize the capture of visuals over text.

Even when content is captured, it may or may not be shared. Many products encourage sharing by adopting a camera-first strategy.

17 2

CONTENT CREATION AND CAPTURE

Content creation and capture can be understood across multiple dimensions, such as creators (e.g., UGC, professionals), types (social, entertainment, informational), frequency (times per day, per week), medium (VR, 360-degree video, text).

Content creators

In the context of news feeds offered by social apps, there are four primary categories of creators: friends, pages, groups and news. Content generated by friends is often personal in nature, but may also include links and re-shares of content originally generated by others. Content generated by a page might come from a business, a celebrity, or even a profile of a pet. Groups generally reflect shared interests, such as members of a soccer team or professional organization. News usually comes from a media organization, such as The New York Times. In each category, a creator is eligible to appear in a user’s news feed as long as the user has subscribed to their content.

Social networks today increasingly feature user-generated content (UGC), and the landscape of such content is changing radically due to advancements in technology?—?including faster and cheaper smartphones, new social apps, and developments in virtual reality (VR) and augmented reality (AR). In particular, improvements in video and photo quality have democratized production and led to an explosion in UGC, and sharing habits are trending away from text toward pictures and video (360-degree and otherwise) as creating visual content becomes easier and more fun. Because it is authentic and personal, UGC also tends to generate relatively high levels of engagement; people are inherently social and are curious to know what their friends and family are doing.

CONTENT SHARING

To increase engagement with your product, content must be not only created and captured, but shared. Successfully encouraging users to share can be difficult, especially when you are competing with many other products for a share of the market. The success of a social product ultimately depends on the volume of content its users share, and many such products have failed because their content production dropped off over time.

Like content creation and capture, content sharing can be understood across multiple dimensions, including audience, social acceptability of the content, feedback, perception, content type, format type, product simplicity and permanence of the content. Below, we explore each of these dimensions in detail.

Audience

We are all inherently insecure social beings. Furthermore, we care about our privacy and are generally not willing to broadcast our lives to the public about anything and everything about our lives. As a result, the number of people that are eligible to see our posts determines what we are willing to post. Of course, how close we are to the people that receive our posts also influences what we are willing to post.

Therefore, from a product perspective, the size of a user’s audience has significant implications on content sharing. As a product grows, the number of each user’s friends and subscribers generally grows, as well. Some users are therefore more willing to share content, and with a greater percentage of their followers, during a product’s early stages. As their audience grows, they may even stop sharing entirely, perhaps shifting their use to alternative products such as one-on-one and group messaging. However, an increase in audience may motivate other users (such as celebrities) to share more.

Type and quality of content

Similarly, the type of content a user is willing to share may depend on the size of their audience. As their audience grows, the user often remains willing to reshare content such as a news article or a funny video, but may stop posting deeply personal content, such as family photos (see Figure 3). They may still post updates about major life events such as a wedding or the birth of a child, but are far less willing to share the “everyday” moments.

Users are also generally less willing to share low-quality content with a larger audience.

17 3

Mimicry

Human beings often mimic or imitate others either consciously or unconsciously. Mimicry has social benefits and helps build a bond between people or groups. Copying others can range from unconscious processes whereby perception (seeing others engage in a certain behavior) becomes directly linked to our own behavior, to conscious strategies, where we choose to imitate either because we’re uncertain about the best course of action or because we want to fit in.

At a restaurant, we may order the same meal that we just saw another person eat, while being unaware of the copying process. Or we may consciously copy others by asking the waiter for the most popular item on the menu. This is an important phenomena in social media, where it manifests as using popular hashtags, uploading photos from popular locations, liking a popular artist’s page, sharing popular news articles, etc. Mimicry also drives the types of media people capture and share; if they see more videos on their news feed, they feel it is socially acceptable to share videos and are more likely to do so. Therefore, from a product perspective, generating strong mimicry will likely lead to greater content production.

Both the declared and perceived intent of your product will strongly influence whether people will use it, and in what use cases. For example, consider the success of Yik Yak, an anonymous gossip product that targeted people within a five-mile radius, versus that of LinkedIn, which is primarily used for professional networking. While mimicry often drives the social acceptance of a new product or behavior, so, too, do previously established social norms set by the product.

Perception

In a high-inventory news feed environment such as Facebook, where users consume only a small percentage of the content available to them, the posts at the top of their feed are generally ranked more highly by the product’s algorithms?—?that is, those posts have a relatively high number of likes and comments and are likely to be of high quality. This often leads users to mistakenly believe most posts are more engaging and of higher quality than their own, which can lead to various changes in behavior. A user might increase the quality of their posts, but might also reduce the number of times they share?—?or even stop sharing entirely?—?out of a misplaced fear that engagement for their posts is lower than others.

Impermanence

Content that self-destructs seconds after it is opened has strong appeal with younger demographics. This format (popularized by Snapchat) is an antidote to a world in which nearly every feeling, celebration and life moment is captured to be shared, logged, liked, commented on, stored, searched and sold. For people who don’t want to worry about unflattering pictures or embarrassing status updates coming back to haunt them, the format’s appeal is obvious. A goofy, spontaneous moment is often much more likely to be shared when the user knows it won’t last forever.

Moreover, younger users are becoming acutely aware of the permanence of the content shared through the web?—?and its repercussions later in life. Therefore, the impermanence of a social post on a platform such as Snapchat (and stories as a format) can also strongly influence the type and quality of content shared.

KEY METRICS TO TRACK

  • Frequency of posts per user?—?The higher the frequency of posts, the greater the inventory and higher opportunities for relevance.
  • Number of unique users who post?—?The greater the percentage of people that post, there is a higher probability of people visiting the product.
  • Type of content shared?—?The greater the variety of content that is unique and authentic, the higher the chance of relevance. People will return to the product to consume more often.
  • Quality of content shared?—?Understanding quality at the user level and how that has changed over time is valuable to build the right type of products.
  • Waterfall analysis of content production, e.g., content composed but not shared. People may compose a post but not share because of any number of reasons including quality, perception and audience.

TAKEAWAYS

  • Content production is the single most important factor that influences engagement. If no content is produced, there will be nothing for your users to consume and give feedback on.
  • Content sharing is strongly influenced by audience, social acceptability of the content, feedback, perception, content type, format type, product simplicity and permanence of the content.

Follow us on Medium for weekly updates.


This work is a product of Sequoia Capital’s Data Science team. Chandra Narayanan and Hem Wadhar wrote this post. We would like to thank Jamie Cuffe and Jenny Wang for their contributions to this post. Please email data-science@sequoiacap.com with questions, comments and other feedback.

Related Article Data-informed product building playbook.
国产人妻熟女偷在线视频_人妻熟女中文字幕在线播放_人妻熟女中文字幕在线播放