Content Discovery: Purpose of Page Visits

The Context Graph project is driven in large part by our desire to ensure that internet users can effectively explore the web and find relevant content — and that the web stays an open public resource. To that end, we would like to better understand how people currently access the web, and why.

After a review of academic literature, prior work done by the Firefox User Research Team, and conversations with folks at both the Mozilla Corporation and the Mozilla Foundation, we compiled a list of open questions about user intents and behaviors on the web that we do not have answers for — and could not locate publicly available information.

In this post, we will introduce a series of surveys that we conducted to help educate ourselves, inform future analysis efforts, and shed light on the state of internet use today. We will highlight some findings from the first part of the study, and follow up with more results shortly.

Goals and Design

We designed and deployed a sequence of three surveys from March to June 2017 to examine five topics on why and how people access the web.

• Purpose of Page Visit: Why is a user visiting a webpage today?
• Prior Visits: Has the user previously visited the page? Does the user regularly revisit the page?
• Page Interactions: What tools, interactions, or content does the user find the most and least helpful for the page visit?
• Discovery and Navigation: How did the user first learn about the page, first navigate to the page, and return to the page over time?
• Workflow and Follow-up Actions: How does the page visit fit into the user’s workflow? By what criteria does the user assess whether the page is relevant? What follow-up actions does the user take to accomplish their task?

Deployment

The surveys were 4–5 pages in length and conducted online via Heartbeat in Firefox. Depending on their assigned track and conditions, most users were asked 16–21 questions, which included both multiple choice questions with optional write-ins as well as free-form text responses. Write-ins and free form text responses were manually coded for analysis. The three surveys reached over 600 users each, or over 2,000 users in total. Response rates ranged between 53% and 90% depending on the question.

Limitations

Below, we report on the relative magnitude of various responses, and highlight interesting or prominent patterns. However, we caution against reading too deeply into the exact numbers. While sizable for qualitative studies, the surveys are designed to be exploratory in nature. All responses are self reported and not yet cross referenced with other data sources. Counts between default values and write-in values are directly not comparable. Coding is done with only two passes and by a single researcher. Participants are self selected and do not represent the general web user population.

With these caveats in mind, we do see a lot of fascinating patterns and have many observations to share. Let’s get to work!

Purpose of Page Visit

The surveys were randomly sent out to select Firefox users. If a user accepts the survey, we ask them the following questions about the webpage they are currently looking at.

The following tables show the most frequent answers. Values in parentheses denote common write-ins. The total percentages may not add up to 100% due to rare responses not shown here, or because some users opted to not answer the question.

Top-Level Purpose

Why are our users visiting the webpage they’re currently on?

In 2004, Pew Internet and American Life project found that the top four reasons American went online were to get information, communicate with others, conduct transactions, and entertain themselves. In our surveys, we find that these four purposes continue to dominate at least desktop web usage.

Aside from the default values, the most frequent write-in is “work.” We suspect work-related tasks might be sufficiently different from non-work-related tasks, that some users perceive work-personal separation to be more salient than categorizing the purpose of a webpage visit by the four functions above. Another frequent write-in is “email.” We suspect that some users may use email for a wide range of activities, beyond just communication.

Breakdown of Page Visit Purposes

Information Purposes

Among the users seeking information on the web, what are they looking for?

This category has a long tail of responses. The top answers, looking for news and contact information, are consistent with the earlier Pew Research findings. Two other common answers in 2004, driving directions and weather, however, receive notably low responses now. We suspect these tasks are still important but might have migrated to mobile in recent years.

Communication Purposes

Among the users communicating with others on the web, whom are they contacting?

The answers are dominated by communication with family and friend. A large number of users take the time to specify that their communication is for work and/or with a business entity. Some users describe their communication as “connecting with like-minded people” via discussion forums or social media without calling the recipients “friends.”

Transaction Purposes

Among the users conducting a transaction online, what are they doing?

The responses are dominated by payments, banking, and shopping — consistent with earlier results. However, a large number of users state that they are currently completing a work-related task.

Entertainment Purposes

Among the users looking for entertainment online, what are they doing?

Nearly half of the users state that they are looking for a video, which has overtaken all other forms of entertainment on the desktop including gaming, personal hobbies, music, and reading for pleasure (all of which ranked higher than video in 2004).

Prior Visits

Have our users previously visited the webpage that they are currently on?

Summary and Next Steps

These surveys provide us with a rough sense of what our users are trying to accomplish on the web. Compared to earlier Pew Research results, we find that a few tasks (e.g., looking for maps and driving directions, checking current weather, dating) might have moved from desktop to mobile. Work-related activities now make up a larger part of our study participants’ use of the internet. Video is becoming the most prominent form of entertainment on web. Some users consider email as serving a distinct function on the web than their other tasks.

What do you think of the responses? Please let us know if you have alternative interpretations of the survey results, or would like to share with us any personal experiences that resonate or contradict the above findings.

In future posts, we will present more results on how our users interact with a webpage, discover content, navigate the web, and how a page visit fits into the user’s workflow. Stay tuned!