Research · 2021 social media trends
What does 2021 social media trends mean in practice?
The 2021 record shows mostly stable reported U.S. platform adoption since 2019, notable growth for YouTube and Reddit within Pew's comparison, large age and frequency differences, and continued growth in company-defined Meta and Twitter audience measures. Each finding has a different denominator and must remain separate.
Define the survey and filing windows before comparing figures
Pew interviewed 1,502 U.S. adults by landline and cellphone from January 25 to February 8, 2021. Meta reported average December daily activity and December 31 monthly activity; Twitter reported a fourth-quarter average. The survey measured whether adults said they ever used a platform and, among users, how frequently they visited. The filings measured accounts or estimated people satisfying company rules. These sources do not form one global audience series. Preserve field dates, geography, sample or account base, question or definition, and uncertainty beside every figure. A year-end filing is not an annual average, and reported platform use is not exposure to a particular message or proof of outcomes.
Most measured platform shares were stable while two changed
Pew reported 81% of U.S. adults used YouTube and 69% Facebook. It found 40% used Instagram, roughly three-in-ten Pinterest or LinkedIn, one-quarter Snapchat, similar shares Twitter or WhatsApp, 21% TikTok, 18% Reddit, and 13% Nextdoor. Within comparable 2019 telephone-survey measures, YouTube rose from 73% to 81% and Reddit from 11% to 18%; other compared platforms were statistically unchanged. TikTok and Nextdoor did not have the same 2019 phone-survey comparison. These overlapping self-reported audiences cannot be added. A statistically unchanged national share does not imply unchanged behavior for every subgroup, campaign, or community.
Age and visit frequency changed the interpretation
Pew reported that 84% of adults ages 18–29 used any social-media site, compared with 45% of adults 65 and older. Among ages 18–29, 71% reported Instagram use, 65% Snapchat use, and 48% TikTok use; Snapchat's gap between that group and adults 65 and older was 63 percentage points. Frequency used another denominator: 70% of Facebook users visited daily, as did 59% of Snapchat users and 59% of Instagram users. Among Instagram users ages 18–29, 73% visited daily. Those percentages are conditional on users, not all adults. They measure reported visits rather than duration, content creation, public reach, satisfaction, or business effect. Age association does not prove why a platform was chosen.
Survey method and sampling uncertainty belong in the ledger
Pew's methodology says the survey used random-digit-dial landline and cellphone samples, interviews in English and Spanish, and weighting to population benchmarks. It reports sampling error and notes that question wording and practical difficulties can introduce error or bias. That method supports a nationally scoped estimate within its design, not a census. Preserve the stated margin of error for each subgroup and avoid ranking close percentages without a significance test. Mode also matters when comparing later research: a change from telephone to another survey mode can affect results. A defensible trend entry therefore stores both the reported value and the method that made it observable.
Company metrics expanded under distinct rules
Meta's 2021 Form 10-K reported Facebook average December DAUs of 1.93 billion and MAUs of 2.91 billion at December 31, up 5% and 4% year over year. It reported Family DAP of 2.82 billion and Family MAP of 3.59 billion. Twitter's filing reported 217 million average fourth-quarter mDAU, up 13%, comprising 38 million in the United States and 179 million elsewhere. Twitter also expanded the mDAU definition in the fourth quarter to include paid products such as subscriptions and said the change had no material effect then. Meta's product and Family measures and Twitter's mDAU differ in products, account eligibility, deduplication, estimation, and period; they are not directly comparable or unique-person totals across the internet.
Create one row per construct, not one popularity score
For each material claim, record publisher, URL, field or metric date, geography, population, question or company definition, unit, denominator, comparison period, uncertainty, exclusions, and limitation. Mark platform adoption, daily frequency, subgroup difference, company account activity, and metric-definition changes separately. Preserve Pew's statistically unchanged language and flag platforms without comparable prior phone data. Record Twitter's subscription-related definition update even though the company said it had no material fourth-quarter effect. Do not add overlapping platform percentages or merge Meta and Twitter measures. If a recommendation follows from the data, label it as an operating choice rather than a measured fact.
Use a current observation window for current decisions
What's Trending's API guide documents current ranked topics, public source links, source-dependent metrics, collector status, and missing values that mean unavailable rather than zero. It advises inspecting original URLs before making claims. That helps form a present evidence window, but it does not reconstruct every 2021 observation, private activity, or a continuous historical series. Collect current source-linked evidence separately and compare only compatible fields. The 2021 snapshot is useful for designing subgroup questions, distinguishing adoption from frequency, and detecting definition changes. It cannot guarantee that a historic platform mix, posting pattern, or audience measure will produce current reach, traffic, engagement, sales, or trust.
Examples
- A planning memo claims every network grew in 2021 and that younger adults' high platform use guaranteed campaign reach. Pew's comparable telephone measures show most platforms statistically unchanged since 2019, with significant growth for YouTube and Reddit; subgroup and daily-use rates use different denominators. Company filings add still different account constructs. The memo is rewritten into separate adoption, subgroup, frequency, and company-metric findings, with a fresh current collection used for the actual decision.
Limits and interpretation
- Pew's platform percentages are overlapping self-reports from U.S. adults, not global unique-person totals.
- Daily-use percentages are conditional on platform users and cannot be read as shares of all adults.
- Sampling, weighting, question wording, mode, and subgroup uncertainty limit close comparisons.
- Meta Facebook/Family metrics and Twitter mDAU use different company definitions and measurement periods.
- Current public monitoring cannot reconstruct 2021 or guarantee present outcomes.
Which measured platforms grew from 2019 to 2021 in Pew's comparison?
YouTube and Reddit showed statistically significant increases; the other platforms with comparable telephone-survey data were statistically unchanged.
Did 48% of all young adults use TikTok daily?
No. Pew's 48% figure was reported platform use among ages 18–29, not daily use; adoption and visit frequency require separate denominators.
What changed in Twitter's mDAU definition?
Twitter added paid products, including subscriptions, in the fourth quarter and said the change had no material effect on the reported measure then.