Instant Census Blog

Back to all blog posts

Pew Study Finds Twitter Users Have Multiple Personalities, LinkedIn Is A Ghost Town, And Facebook Is Full of Teenagers

By Josh Zagorsky on September 3, 2015


Twitter, Facebook, and LinkedIn

The Pew Research Center recently released a study on social media usage, which we compared with published user counts from Facebook, Twitter, and LinkedIn. Here are some things we discovered by comparing Pew’s findings with the companies’ own numbers:

  • A large fraction of Twitter users have multiple accounts

  • Most LinkedIn users log in very rarely

  • Facebook has a lot of users who are under 18 years old

Twitter Logo

Twitter

Twitter data: 66 million monthly active US users

Pew data: 53 million Americans say they use Twitter

Based on Pew’s data, the number of people using Twitter is actually much lower than Twitter’s reported number of monthly active users, probably because many people who use Twitter have multiple accounts (work accounts, personal accounts, joke accounts, and spam accounts).1

LinkedIn Logo

LinkedIn

LinkedIn data: 31 million monthly active US users; 116 million total US user accounts

Pew data: 54 million Americans say they use LinkedIn

Only about one quarter of US LinkedIn accounts are active during a given month. Based on the Pew data, almost half of all people who say they use LinkedIn haven’t actually logged on to LinkedIn in the last month. Half of all US LinkedIn accounts are duplicate, spam, or forgotten (they belong to people who do not report using LinkedIn).

Facebook Logo

Facebook

Facebook data: 145 million daily active US users; 189 million monthly active US users

Pew data: 116-124 million Americans say they use Facebook daily; 166-169 million total say they use Facebook

Facebook’s numbers are slightly higher than Pew’s. We have to estimate how many minors are on Facebook, because Pew only included survey results from people 18 and older, but Facebook’s minimum age is 13. If we assume that 80% of 13-17 year-olds are monthly active users, we get 169 million Americans on Facebook, 11% short of Facebook’s reported count. Facebook says in its quarterly earnings that it estimates duplicate user accounts and spam accounts to be fewer than 7% of monthly active users. If we assume duplicate and spam accounts are close to 7%, we get an estimate only a few percentage points higher than Pew’s 2.6% margin of error.

Pew’s data and Facebook’s data have a wider disparity for daily active users: Facebook reports 16% more daily active users than Pew finds (assuming 80% of 13-17 year-olds are on Facebook daily). This could stem from a difference in definitions: it’s possible that people who told Pew they use Facebook “weekly” actually use it several times a week, and on any given day, Facebook counts a large proportion of them in its daily active user tally.

A Longitudinal Study Could Provide Even More Insightful Data

The Pew study is a snapshot of social media usage at one point in time. A longitudinal study might provide even greater insight by asking the same participants the same questions regularly over a period of time. For example:

  1. A longitudinal study could explore whether teens’ use of social media changes between vacation periods and the school year, and whether adults’ social media use changes between weekdays and weekends.

  2. Pew’s study showed LinkedIn usage dropping 3% from 2014 to 2015, but this is within the study’s margin of error. It would be interesting to see whether any respondents report using LinkedIn one year but then stop using it the next year.

  3. A longitudinal study could examine whether individuals’ use of social media is relatively steady over time, or whether people alternate bingeing on social media sites with periods of lower use.

Longitudinal studies have very effectively leveraged Instant Census text message surveys to regularly and unobtrusively ask people the same set of questions every few days. Text messaging is uniquely suited to longitudinal and diary studies, because respondents are much more willing to quickly answer short questionnaires than to regularly take a long phone call or periodically fill out a web survey.

Survey Methodology

Methodology

The Pew Research Center’s study was based on a landline and cell phone survey of a sample of 1,907 United States adults age 18+ (see Pew’s methodology page). The Pew survey included questions on usage of Facebook, LinkedIn, Twitter, Pinterest, and Instagram. We compared Pew’s results with the user counts that Facebook, LinkedIn, and Twitter announce quarterly. We didn’t examine Pinterest or Instagram because neither network releases regular, precise user counts (Pinterest is privately held and Instagram is owned by Facebook).

You can download our calculations here. We made several assumptions, especially about the proportion of minors using each social network. We assumed that virtually nobody under 18 uses LinkedIn, that 13-17 year-olds use Twitter in the same proportion as adults, and that 13-17 year-olds use Facebook much more than adults.

Looking to set up a one-time or longitudinal study using text messaging as one your interview methods? Contact one of our friendly software engineers for a demo and we’ll walk you through how to use Instant Census’s survey builder and explain how text message surveys can achieve higher response rates and faster response times that email or phone surveys.


  1. Facebook prohibits multiple accounts: “You will not create more than one personal account.” (https://www.facebook.com/legal/terms) Twitter encourages it on an FAQ page: “I want more than one Twitter account. Is that possible? Yes! You can have more than one account on Twitter.” (https://support.twitter.com/articles/20169956) 

Want to learn more about our capabilities and what we do? Get in touch!

Whatever your target audience, discover how easy it is to start connecting and get results.