Beyond Instagram: where a second social channel adds a different audience
A platform can be popular and still bring you back to many of the same people. We used Pew Research Center’s 2025 survey to examine the difference.
US adults · Survey fielded February 5–June 18, 2025 · Analysis reviewed October 17, 2026
Abstract
YouTube and Facebook offered the largest audiences outside Instagram among the ten alternatives measured in our analysis of Pew Research Center’s 2025 US survey. Among 4,640 respondents who answered all eleven platform questions, the weighted share using YouTube but not Instagram was 38.7%; for Facebook it was 29.8%. These are additional percentage points beyond the same cohort’s 52.3% Instagram share. YouTube and Facebook remain the leading two platforms in our checks on missing answers.[1][2]
The practical implication is to give a second channel a clear job. If the goal is to find people outside Instagram, YouTube and Facebook deserve consideration. If the goal is to deepen relationships with people already using Instagram, overlap may be useful. The survey measures reported ever-use, including occasional use. It does not predict who will see a post, become a customer or respond to a particular brand.
Imagine a small team that can sustain one more social channel. Instagram is already part of its routine. The team can make useful posts, but it cannot create a different campaign for every platform. The next decision sounds simple: where should it go?
A list of the biggest platforms does not quite answer that question. The team needs to know how much of the second platform’s audience is already on Instagram, and whether the people outside Instagram fit the business. Those are different questions from which app is growing or which format is fashionable.
This study addresses the first part. We compared answers from the same respondents, so a person who uses both platforms is counted once in the combined audience. The result is a way to narrow a shortlist, followed by a test of the content and customers that matter to the business.
Two platforms, three audience groups
Instagram only
Both platforms
Second platform only
The useful number is the audience outside Instagram
Consider TikTok. Among the people who answered all eleven platform questions, 38.8% reported using it. But 74.2% of those TikTok users also reported using Instagram. TikTok’s audience outside Instagram therefore represented 10.0% of that comparison group. Its total audience and its additional audience answer different planning questions.
YouTube tells a different story. Its total share was 87.8%, and 55.9% of its users also used Instagram. The remaining YouTube users represented 38.7% of the comparison group. Adding that group to Instagram’s 52.3% produces a combined share of 90.9%, counting each person once. These are weighted survey shares, not a promise that an Instagram account and a YouTube channel can reach nine out of ten people.[2]
There is no need to treat overlap as wasted effort. The same person may want a quick introduction on one platform and a detailed explanation on another. That is a possible content strategy, not something this survey measures. The useful distinction is whether a second channel is meant to broaden the potential audience or serve existing audiences in another way.
YouTube and Facebook stand apart on additional audience
YouTube added 38.7 percentage points beyond Instagram in the main comparison; Facebook added 29.8. WhatsApp and TikTok were much lower, at 11.3 and 10.0 points. The approximate 95% sampling intervals for the YouTube and Facebook additions are shown below, alongside the other estimates. Those intervals describe sampling uncertainty under our stated method, not all possible sources of survey error.[3]
The gap between the leading two platforms also persists when we include respondents with unanswered questions in a separate sensitivity check. Allowing every unresolved platform answer to take either value gives an overall YouTube addition of 36.3–41.4 points and a Facebook addition of 27.9–32.9 points on the full-sample denominator. These are missing-answer bounds, not confidence intervals. Their separation supports the same broad ordering without pretending blanks mean “No.”[4]
That does not make YouTube the automatic choice for every business. A team needs a useful format, a suitable audience and enough capacity to keep going. Nor should WhatsApp’s position in a platform survey be read as an equivalent public-feed opportunity. Reported app use does not establish that strangers can discover a business’s content, or that a business has permission to contact them.
Use the table to compare audience breadth. Then assess the publishing task separately. A smaller additional audience can still be relevant when the channel fits a specific customer group or a valuable kind of conversation. This dataset does not contain the commercial outcomes needed to rank those choices.
Additional reported audience beyond Instagram
Weighted percentages of the same 4,640 complete respondents. Each bar shows the share using the second platform but not Instagram. Intervals are approximate 95% sampling intervals.
Share of the complete-response cohort (%)
Platform comparison and uncertainty
| Second platform | Outside Instagram | Approx. 95% interval | Either platform | Also uses Instagram, among second-platform users |
|---|---|---|---|---|
| YouTube | 38.7% | 36.8%–40.6% | 90.9% | 55.9% |
| 29.8% | 28.1%–31.6% | 82.1% | 59.6% | |
| 11.3% | 10.1%–12.7% | 63.6% | 66.4% | |
| TikTok | 10% | 8.8%–11.3% | 62.3% | 74.2% |
| 7.2% | 6.3%–8.3% | 59.5% | 74% | |
| X | 4.3% | 3.6%–5.2% | 56.6% | 80.2% |
| Snapchat | 4% | 3.3%–5% | 56.3% | 84.9% |
| Truth Social | 1.5% | 1.2%–2% | 53.8% | 57.8% |
| Bluesky | 1.2% | 0.8%–1.6% | 53.4% | 72.2% |
| Threads | 0.9% | 0.6%–1.4% | 53.2% | 89.7% |
Age changes how much room a second channel has to add
Instagram was already used by 80.6% of respondents aged 18–29 who answered every platform question. With that much of the group already represented, any second platform has a smaller audience left to add. YouTube added 18.2 points in this age group, compared with 49.7 among ages 50–64. Facebook added 9.6 and 39.8 points respectively.
This is a useful change of perspective for a team targeting older customers. Choosing a channel by the habits of the people creating the content may overlook people the business wants to serve. The survey gives a reason to investigate the customer’s platform mix rather than assuming the team’s own routine is representative.
Read the age results as descriptive comparisons. The 18–29 cohort contains 463 respondents, while the 65+ cohort retains only 85.2% of that age group’s survey weight after the complete-answer requirement. Missing answers matter more there. We do not turn small differences among Facebook, Reddit and TikTok in the youngest group into a ranked recommendation.[3][4]
The age groups are separate groups of people observed in one survey. They do not show that an individual will change platforms as they get older. They also do not identify a brand’s customers. A business serving a narrow profession, location or interest should treat these broad groups as context.
The same comparison, by age
Weighted complete-response cohorts within each age group. These are descriptive estimates; the table includes the cohort size and retained survey weight.
| Age | Respondents | Survey weight retained | Instagram use | YouTube outside Instagram | Facebook outside Instagram |
|---|---|---|---|---|---|
| 18–29 | 463 | 95.4% | 80.6% | 18.2%Approx. 95% interval: 14.2%–23.1% | 9.6%Approx. 95% interval: 6.6%–13.7% |
| 30–49 | 1,346 | 95.8% | 63.3% | 33.2%Approx. 95% interval: 29.9%–36.5% | 25%Approx. 95% interval: 22%–28.2% |
| 50–64 | 1,185 | 91.1% | 41% | 49.7%Approx. 95% interval: 46.1%–53.3% | 39.8%Approx. 95% interval: 36.3%–43.5% |
| 65+ | 1,598 | 85.2% | 22.2% | 53.1%Approx. 95% interval: 49.8%–56.3% | 44.7%Approx. 95% interval: 41.5%–47.9% |
A small team can turn the finding into a focused test
Take a fictional home-gardening business whose customers include adults in their fifties. It already posts short Instagram demonstrations and has time for one additional weekly piece. Its question is whether another channel can introduce its advice to people outside its existing Instagram audience. The study gives it a reason to shortlist YouTube and Facebook, without predicting which will produce sales.
If the team can explain one recurring customer problem on video, it could test a weekly YouTube guide. For example, a complete demonstration of repotting a root-bound plant gives the longer piece a job of its own. An Instagram clip can show the first symptom; the longer guide can cover tools, decisions and common mistakes. Reusing the source material reduces production effort, while the two pieces still answer different questions.
If longer video would be difficult to sustain, the team could instead test a Facebook post that explains the same problem with clear photographs and a discussion prompt. That is a choice about production capacity and the customer conversation, not evidence that Facebook will outperform YouTube for this business.
Before starting, the team should write down what would make the test worthwhile: relevant enquiries, useful visits or completed purchases, depending on its existing measurement. Review the result alongside time spent creating and responding. Keep the same measurement definitions throughout. The survey helps choose a hypothesis; the business’s own results decide whether the work earns a place in its schedule.
An illustrative six-week test is a planning example, not a statistically validated duration. It should be long enough to publish useful material consistently and short enough to review before committing more resources. Do not change attribution or consent settings just to make the test appear successful.
Give the next channel a clear job
Choose the audience question
Decide whether you want to reach people outside Instagram or serve an existing audience in another way.
Choose work you can sustain
Give the format a specific customer problem to solve and plan the time needed to create and respond.
Review your own results
Use your existing measures and a defined review date to decide whether to continue, adapt or stop.
A shared audience can still have a reason to follow you twice
A team choosing TikTok or Threads may have a different objective from the gardening business. It may want to participate in a conversation, develop a different creative format or give existing followers another way to engage. High overlap does not invalidate those objectives. It does mean that adding a platform should not automatically be described as finding an entirely new audience.
The distinction also prevents a common reporting mistake. Adding follower counts or platform-level reach figures does not identify unique people across channels. This study can estimate overlap because the survey asked the same respondents about each platform. It cannot deduplicate the actual followers or viewers of a business.
The most useful planning sentence may be simple: “We are adding this channel to do this particular job.” It gives a team a clearer brief, makes content choices easier to defend, and provides a reason to stop or change direction when its own evidence does not support the effort.
How we analyzed the survey
We used Pew Research Center’s 2025 National Public Opinion Reference Survey public-use file, downloaded October 17, 2026. SSRS surveyed 5,022 US adults from February 5 to June 18, 2025 using online, paper and telephone responses in English and Spanish. We applied the released final survey weights. The 2026 download did not contain the social-platform response fields, so this analysis uses the available 2025 observations and makes no claim to measure 2026 behavior.[1]
The main comparison uses one consistent set of 4,640 respondents with a Yes or No answer for all eleven platforms. This retains 92.1% of the full sample’s weight. We excluded 176 people whose platform answers were structurally absent and 206 additional people with at least one unanswered platform item. We did not replace their answers with No. Using a complete-response cohort makes the alternatives comparable, but does not guarantee that the excluded people have the same platform habits. The 48 complete respondents without a recorded age remain in the overall results but are absent from the age breakdowns.
For each alternative, we calculated four weighted quantities: its audience outside Instagram; the share using either platform; the share of that alternative’s users who also use Instagram; and its share among Instagram nonusers. The first is our main comparison. Every calculation uses answers from the same person, rather than assuming platform use is independent. The downloadable results retain the distinct denominators.
Sampling intervals use a stratified Taylor approximation with the released final weights and sampling strata. Age and complete-response groups are treated as domains within the full sample. We retain respondent-level covariance when comparing alternatives. The calculation treats weights as fixed and does not reconstruct all uncertainty from calibration, nonresponse or self-reporting. We do not apply Pew’s full-sample margin of error to every subgroup.[3]
We checked the main result against pair-specific complete answers and bounds that let every unknown answer vary while preserving what is known. We also independently reproduced the calculations and reconciled 121 rounded published controls covering platform use, internet-user categories and age groups. Our original question specified Instagram as the baseline; the move from the unavailable 2026 battery to 2025 and its measured platform list is documented in the analysis notes. This is an exploratory secondary analysis, not a preregistered experiment.[4]
Download the derived tables to inspect the numbers, or the analysis package to reproduce them using a copy of Pew’s original dataset. We do not redistribute respondent records. Results are rounded for reading; calculations use unrounded weights and values. Small or imprecise subgroup estimates that fail our stated display rules are withheld from the public tables.
Research notes
- Pew Research Center, 2025 National Public Opinion Reference Survey. Fieldwork February 5–June 18, 2025; full sample n=5,022. The source page provides the dataset and related material.
- The primary denominator is the weighted cohort with Yes/No responses on all eleven measured platforms, n=4,640. Additional share equals the weighted count using the alternative but not Instagram divided by that cohort’s weight. Combined share equals Instagram share plus additional share. Overlap instead divides shared users by users of the alternative.
- Pew’s methodology describes the address-based design and weighting. Our analysis uses the released sampling strata and final weights for approximate domain-based intervals. Sampling intervals do not capture every possible source of bias. Display rules require denominator n and effective n of at least 100, at least 20 respondents in each outcome category, with headline estimates also requiring an interval no wider than 20 percentage points.
- Missing-answer bounds preserve known Yes/No responses and allow unresolved answers, including structural blanks, to vary. They use the full sample or full age group as the denominator and are separate from sampling intervals. Pair-specific complete-answer results and analysis decisions are included in the downloadable package. Age-group comparisons are descriptive.
- Pew Research Center bears no responsibility for the analyses or interpretations of the data presented here. The opinions expressed herein, including any implications for policy, are those of the author and not of Pew Research Center.