Research · 2017 social media trends
What does 2017 social media trends mean in practice?
The strongest 2017 evidence shows social media’s role in U.S. news use widening, especially among older adults, while news audiences overlapped across platforms. Twitter also expanded its text limit after a measured test, and Facebook and Twitter reported different account-growth metrics. These observations are dated and method-specific; they do not form one global adoption total.
Define 2017 through measurements, not memories
A useful 2017 retrospective begins with contemporary evidence. Pew surveyed U.S. adults in August about getting news on social media. Twitter documented a character-limit test and rollout in September and November. Facebook and Twitter later filed year-end account metrics. These sources support distinct claims about news pathways, audience overlap, a product-format change, and platform-defined activity. They do not establish every important event of the year or a representative worldwide platform ranking. Keep survey field dates, rollout dates, filing periods, populations, and definitions visible. A later filing can describe a December or fourth-quarter measure without becoming an annual average, and an official product post can document a change without proving its effect on society or marketing performance.
Social news use increased in a defined U.S. survey
Pew Research Center reported that 67% of U.S. adults got at least some news on social media in August 2017 and 20% did so often. The overall estimate was a modest rise from 62% in early 2016. For the first time in Pew’s series, 55% of adults ages 50 and older reported getting news this way, ten percentage points above 2016, while 78% of adults under 50 did so. These are survey estimates about a particular purpose—news—not total platform adoption, logged activity, or verified exposure to accurate reporting. The age pattern is an association and does not show why use changed. Sampling, weighting, wording, and panel participation remain part of the claim.
Platform-user rates and population reach differed
Pew found that 74% of Twitter users got news on Twitter, 32% of YouTube users got news on YouTube, and 29% of Snapchat users got news on Snapchat. Yet platform audience sizes differed: the report estimated that 45% of all U.S. adults got news on Facebook, 18% on YouTube, and 11% on Twitter. A high news rate within a platform’s users can coexist with lower population reach. Do not compare the 74% conditional Twitter figure directly with Facebook’s 45% all-adult figure. The report also found 26% of all adults getting news from two or more studied social sites, up from 18% in 2016. News audiences therefore overlap, and platform percentages cannot be added to estimate unique people.
A format change had its own evidence boundary
Twitter’s official product posts documented a September 2017 test of a 280-character limit and a November rollout to languages affected by cramming, while Japanese, Korean, and Chinese remained at 140 characters because the company said those languages could convey more information per character. Twitter reported from the test that only 5% of sent tweets exceeded 140 characters and 2% exceeded 190. Those are platform-reported test observations under a selected rollout, not representative survey findings or evidence that longer posts improved quality, reach, or outcomes. Preserve the language exceptions, test context, and distinction between a product rule and actual user behavior.
Filed account metrics used incompatible systems
Facebook’s 2017 Form 10-K reported 2.13 billion MAUs as of December 31 and 1.40 billion average DAUs in December, both up 14% year over year under its definitions. Twitter’s filing reported 330 million average MAUs for the fourth quarter, up 4% after a disclosed MAU adjustment, and 12% year-over-year growth in average DAU without publishing an absolute DAU count there. Facebook’s point-in-time MAU and Twitter’s quarterly average are not directly comparable. Both companies disclosed internal methodology and account-quality limitations. These figures count platform-defined activity, not unique people, news consumers, or outcomes, and they must never be summed.
Build the evidence ledger around denominators
For each material statement, record its exact wording, URL, publisher, collection or rollout date, filing date, geography, population, numerator, denominator, unit, comparison baseline, exclusions, and limitation. Label news use separately from general platform use; distinguish platform-user percentages from all-adult reach; store overlap; and classify product-test metrics apart from survey estimates. For company filings, preserve averaging periods, revisions, and whether growth is reported without an absolute value. If two measures answer different questions, do not force them into a ranking. The ledger should make it possible for another researcher to reproduce both the claim and the reason a broader claim was rejected.
Use current monitoring as a new observation
What’s Trending’s API guide documents current ranked topics and source-linked public evidence, notes that metrics vary by source, defines missing values as unavailable rather than zero, and advises opening original URLs before treating a signal as a claim. A current scan cannot reconstruct 2017, recover private or deleted activity, or establish continuous coverage. Create a fresh evidence window and compare themes only when useful. Compare numbers only when populations and constructs align. The 2017 record can improve questions about news purpose, audience overlap, format constraints, and account definitions, but it cannot guarantee that a longer post, news-oriented channel plan, or historic platform mix will work today.
Examples
- An analyst writes that Twitter became the main U.S. news platform after doubling tweet length in 2017. The ledger rejects the claim. Pew found a high news-use rate among Twitter users but much lower all-adult reach than Facebook, and its survey did not test the character limit. Twitter’s own rollout post documented test behavior, not news leadership or marketing impact. The analyst instead reports the news-use denominators, the multi-site overlap, the language-scoped product change, and each company’s separate account metrics. A current source-linked scan is kept outside the historical series.
Limits and interpretation
- Pew’s findings measure self-reported U.S. news use, not total platform activity, content accuracy, or verified exposure.
- Platform-user news rates and all-adult reach use different denominators, and news audiences overlap.
- Twitter’s character-limit observations came from its own test and do not prove content quality, reach, or outcomes.
- Facebook and Twitter filings use different account periods and definitions and retain disclosed measurement limitations.
- Current public monitoring cannot reconstruct 2017 or guarantee present performance.
How many U.S. adults got news on social media in 2017?
Pew reported 67% getting at least some news on social media and 20% doing so often in its August 2017 panel survey.
Did Twitter’s 280-character change mean most tweets became longer?
Twitter reported that in its test only 5% of sent tweets exceeded 140 characters and 2% exceeded 190. The test does not establish later universal behavior or outcomes.
Can 2017 platform account totals be compared directly?
No. Facebook reported a December point-in-time MAU while Twitter reported a quarterly average with an adjustment; both used company-specific rules.