DoubleVerify today released "Scroll, Stop and Shop: Walled Garden Trends," the second edition of its Global Insights report dedicated to social and video platforms, built on a survey of 22,000 consumers in 21 countries and 2,020 marketers in 25 countries. One in four consumers made a purchase on a social platform in the past 12 months, according to the report, down from 30% in the company's 2025 edition. Influencers affect purchase decisions for 48% of consumers, against 54% a year earlier. DoubleVerify nonetheless describes consumers as "warming up to ads," and points to a projected $408 billion in global social media ad spend for 2026.

In Short

DoubleVerify, a company that checks where online ads appear and whether they work, asked 22,000 people and about 2,000 marketers how they use YouTube, Facebook, Instagram, TikTok and other apps. Compared with its survey from last year, fewer people said they bought something through these apps or were swayed by influencers, even though more of them said ads on the apps were acceptable. People also said they trust real creators far more than AI-made influencers, which matters for any brand thinking about paying for a synthetic spokesperson.

What DoubleVerify published

The report reached marketers by email on October 14, 2026, under the subject line "Just released: Global Insights report on social platforms." The 33-page PDF itself was produced a week earlier, according to its file metadata, which records a creation date of October 7, 2026. The email describes it as "DV's second dedicated walled garden report" and says it "draws from our annual global survey of consumers and marketing decision-makers."

DoubleVerify, listed on the New York Stock Exchange as DV, is in the middle of a change of ownership. Nielsen agreed on August 6, 2026 to acquire the company for $13.60 a share in cash, an enterprise value of about $2.15 billion, with closing expected in the first quarter of 2027. The report makes no mention of the deal.

The fieldwork was carried out by Sapio Research in February 2026 through online partner panels, according to the methodology section. Consumers were drawn from eight Asia Pacific countries, eight in Europe, the Middle East and Africa, three in Latin America, and the United States and Canada. The marketer sample spanned 25 countries, adding Austria, New Zealand, Poland and Switzerland. DoubleVerify gives a margin of error of 0.7 percentage points for the consumer survey and 2.2 points for the marketer survey, at 95% confidence. Global averages are weighted by region.

One caveat matters for anyone reading the platform tables. Questions about regional platforms were asked only in the regions those platforms serve. Line's 62% usage figure, for example, comes from Asia Pacific respondents only, Kwai's 50% from Latin America, and Josh's 55% from Asia Pacific. Those numbers sit in the same charts as YouTube and Facebook but are not comparable to them.

The headline spend figure

Both the email and the report open on money. "With global ad spend on media-sharing and social platforms expected to reach $408B this year, it's clear that marketers see significant opportunities in walled gardens," according to the email. The report gives the same $408 billion figure, labels it "Global Social Media Ad Spend, 2026," and attributes it to EMARKETER, with a 19% increase over 2025 and a growth rate twice that of spend on other digital channels. The email does not name a source, and its wording, "media-sharing and social platforms," is broader than the report's "social media." Neither document states whether video platforms such as YouTube sit inside the total.

That figure is higher than the one WARC Media published on October 8, 2026, forecasting social media ad spend gains of 21.3% to $394.6 billion. The two forecasters use different definitions, so the $13.4 billion gap is a difference of scope as much as of outlook. Both place growth on social platforms well ahead of the rest of the digital market.

Steve Mougis, Global Chief Commercial Officer at DoubleVerify, frames the spend in the report's welcome letter. "The past several years have brought many changes to digital advertising, but marketers have cemented their commitment to walled gardens," he wrote. He adds a caveat that is unusual for a vendor brochure: "These are 'walled' gardens, after all, and transparency remains a significant obstacle."

What changed between the two editions

DoubleVerify's first walled garden report came out on November 18, 2025. PPC Land's coverage of that edition recorded a sample of 22,000 consumers and 1,970 marketers across 21 countries, and several figures that the new edition revisits without spelling out the change.

  • Social purchases. In 2025, 30% of consumers said they had made purchases directly through social platforms in the past year. In 2026 the figure is 25%, presented in the report as "one in four global consumers."
  • Influencer impact. In 2025, 54% of consumers said influencers affected their purchase decisions. In 2026, 48% said influencer content had a "somewhat" to "significant" impact, while 52% chose "a little" to "not at all."
  • Audience reach. In 2025, 46% of marketers named audience reach as their biggest social media challenge. In 2026 it is 40%. DoubleVerify labels the change as "-12%"; on the 2025 figure reported by PPC Land, the relative decline works out to about 13%, a gap small enough to come from rounding in the earlier number.

The report itself frames 2026 as a year of improving sentiment. The question wording in the 2025 edition is not reproduced in the new document, so the declines above could partly reflect changes to how questions were asked or answers grouped. DoubleVerify does not make the purchase or influencer comparisons itself.

One set of 2025 numbers cannot be lined up at all. The first edition gave platform usage as 72% for YouTube, 69% for Facebook, 54% for Instagram and 45% for TikTok. The 2026 appendix shows 95%, 88%, 80% and 68% for the same four platforms. A jump of that size in a single year among a similar 22,000-person sample points to a different question or base rather than a change in behavior, and the report offers no reconciliation.

Consumers: the Big Four

DoubleVerify groups YouTube, Facebook, Instagram and TikTok as the "Big Four." Globally, 95% of surveyed consumers use YouTube, 88% Facebook, 80% Instagram and 68% TikTok. Below them come X at 59%, Pinterest at 58%, LinkedIn at 57%, Snapchat at 47%, Reddit at 46%, Twitch at 40% and Tumblr at 34%.

Expected growth follows a different order. Among users of each platform, 52% of TikTok users expect to use it more over the next 12 months, ahead of YouTube at 50%, Instagram at 46% and Facebook at 41%, according to the report. Measured across all consumers, as in the appendix, YouTube leads at 48%, with Facebook and Instagram at 36% each and TikTok at 35%. The two sets are consistent; they simply divide by different bases.

The regional spread is wide. North America has the lowest usage of most platforms in the survey: Instagram reaches 66% there against 92% in Latin America, and TikTok 52% against 85%. In Latin America, 50% of respondents expect to increase their TikTok use, and 58% their YouTube use.

There is a small internal inconsistency in how the report describes Latin America. Its opening section says Instagram and Facebook "switch places" in that region. A later page says they are "tied as the second most popular platforms," and the appendix shows both at 92%.

Age and gender

TikTok is the platform most associated with youth in the data. It ranks among the top five platforms for 18- to 24-year-olds and falls out of the top five entirely for consumers over 65, according to the age chart. DoubleVerify says seven times as many 18- to 24-year-olds as over-65s expect to increase their TikTok use, and that the younger group was "up to 13x more likely" to expect more time on walled gardens overall. Facebook moves to second place from the 35-44 age band upward. YouTube holds first place in every age group.

Gender splits are negligible on the Big Four and pronounced elsewhere. X shows the widest global gap, at 67% of men against 50% of women. Twitch follows, at 48% against 33%, and in North America the Twitch split is 43% male to 20% female. Reddit stands at 52% against 39%, LinkedIn at 63% against 52%. Pinterest is the exception: 62% of women use it, against 54% of men. In North America, men outnumber women by 50% or more on Twitch, Tumblr, X and Reddit, and by 27% on LinkedIn, according to DoubleVerify. Latin American women are described as "the heaviest social platform users in our survey."

Ad suitability and where consumers act

The section titled "Consumers Are Warming Up to Ads" rests on a single comparison: 8% more consumers said social media platforms are suitable for ads in 2026 than in 2025. The report does not say whether that is a relative change or a percentage-point change, and it does not give the 2025 base. In the 2025 edition, 16% of consumers called social platforms generally unsuitable for brand advertising.

The 2026 platform figures are as follows. YouTube is rated suitable by 72%, Facebook by 66%, Instagram by 62% and TikTok by 59%. Unsuitable ratings for the Big Four run from 9% on YouTube to 16% on TikTok. Further down, the picture is weaker: Snapchat is rated suitable by 34% and unsuitable by 25%, Reddit by 32% and 26%, Tumblr by 26% and 27%. Josh, the Indian video app, has the highest unsuitable rating in the survey at 33%.

Age shapes openness more than region does. Among 25- to 34-year-olds in Asia Pacific, 65% call social feeds suitable for ads; among over-65s there, 27% do. In EMEA the range runs from 58% to 22%, and in North America from 58% to 30%. Latin America is flatter, from 65% among 35- to 44-year-olds to 49% among over-65s. "In all regions, openness is lower among Gen Z, peaks with Millennials, then drops for every generation after that," according to the report.

Willingness to act on ads follows platform size. Half of consumers say they are likely to act on ads on YouTube, 49% on Facebook, 42% on Instagram and 38% on TikTok. Everything else falls to 16% or below, with X at 16%, Pinterest at 14%, Snapchat at 11%, Reddit at 7% and Tumblr at 4%. Some 21% of consumers answered "none of the above," a group large enough to rank fifth. Consumers in North America and EMEA are the least likely to act on any social ads, according to DoubleVerify.

Household makeup appears to matter more than income. "The presence of kids in the household is a better predictor than income level of whether consumers will act on ads," the report states. Among parents of children under three, 57% said they were likely to act on social media ads, and 65% of households with children under five expect to increase their YouTube use.

Formats

Short-form video is the format most likely to prompt action, cited by 38% of consumers globally, followed by long-form video pre-roll at 31%, search at 26%, feed at 25%, sponsored content at 24%, in-app and influencer sponsorships at 22% each, Stories at 21% and display at 19%. Short-form video moves 47% of consumers in Asia Pacific and 49% in Latin America, but only 29% in EMEA and 27% in North America.

Influencers and the celebrity effect

The influencer section carries one of the report's sharper claims. Mega-influencers, defined as creators with more than one million followers, are the group most likely to "significantly" affect purchase decisions: 25% of consumers globally, against 18% each for macro-influencers (100,000 to one million followers) and micro-influencers (1,000 to 100,000). The share reaches 31% in Asia Pacific and 32% in Latin America, but drops to 19% in EMEA and 18% in North America. Among consumers with household incomes of $100,000 or more, 31% said mega-influencers significantly affect what they buy.

"This is worth noting because, in the past, consumers put more trust in smaller-scale niche personalities," according to the report. "The trade-off between the massive reach of a celebrity and their perceived authenticity may be subsiding."

The regional split on any influence at all is stark. In Latin America, 57% said influencers affect their purchases somewhat to significantly; in Asia Pacific, 56%. In EMEA the figure is 40%, and in North America 39%.

Other research points the other way on authenticity. IAB Ireland and RED C Research found in September 2026 that 72% of Irish adults say authenticity matters in creator content, but only 17% think sponsored creator content has it. The surveys ask different questions in different markets, yet the contrast frames DoubleVerify's suggestion that celebrity reach is overtaking niche trust as an open question rather than a settled one.

Humans over AI

The section titled "Humans Prefer Humans" contains the report's most specific attitudinal data. DoubleVerify asked consumers how four AI scenarios would affect their view of a brand.

  • Human with AI assistance (a creator drafts a review with AI, delivers it in person, then uses AI to enhance the footage): 39% positive, 36% no impact, 25% negative.
  • AI influencer that looks human (a sponsored synthetic influencer realistic enough that most consumers cannot tell): 36% positive, 34% no impact, 30% negative.
  • Content is clearly AI (a creator uses an avatar or AI-voiced video): 32% positive, 35% no impact, 33% negative.
  • Influencer is clearly AI (a brand-sponsored influencer that is obviously synthetic): 29% positive, 34% no impact, 37% negative.

"For brands that sponsor AI influencers, whether realistic or not, the very existence of a fabricated spokes'person' is a trust risk," according to the report. Only the first scenario, where a human remains the face of the content, produced a clearly net-positive result. The realistic synthetic influencer scored better than the obvious one, a finding the report does not dwell on, although it implies that consumers penalize the AI they can see more than the AI they cannot.

The breakdowns show where tolerance sits. Consumers in city centers rated an obviously AI influencer positive 40% of the time and negative 31%. In suburbs those numbers flip to 19% positive and 44% negative, and in rural areas to 23% and 40%. Men split 34% positive to 32% negative; women 25% positive to 41% negative. Households with children aged one to five registered a net-positive response to all four scenarios, according to DoubleVerify.

These figures align with earlier industry data. IAB research published in January 2026 found that 82% of advertising executives believed Gen Z and Millennial consumers felt positively about AI-generated ads, against 45% of those consumers, and that 39% of Gen Z felt negatively about them.

DoubleVerify places its own product next to these findings. The report says AI SlopStopper for Social "identifies and blocks low-quality GenAI content in real time." When DoubleVerify exposed the AutoBait network of AI-generated domains in March 2026, it said AI SlopStopper was scheduled to expand to social platforms during 2026. The new report describes it as available for social, but names no platforms, coverage figures or accuracy rates. Its main competitor moved first on a named platform: IAS brought its low-quality generative AI avoidance to YouTube on September 30, 2026.

Marketers: the funnel moves to the middle

The marketer survey asked which funnel objective each platform serves. On Instagram, 40% of marketers prioritize upper-funnel awareness, 29% middle-funnel consideration, 14% lower-funnel conversion and 13% loyalty. YouTube splits 40%, 25%, 15% and 14%. Facebook is closer to balanced at 35% upper and 33% middle. TikTok stands at 35% and 22%, with 12% of marketers not using it at all.

The Big Four were the only platforms where upper-funnel mentions led, according to DoubleVerify. "Marketers buy reach in bulk from a few giants, then turn to other platforms for down-funnel momentum," the report states. In 2026, on every major platform, marketers shifted some focus from awareness toward consideration and loyalty, though the report does not give the 2025 figures behind that statement.

Smaller platforms are used differently. Pinterest, Reddit and Twitch are described as conversion buys. Tumblr had nearly twice as many lower-funnel mentions (19%) as upper-funnel ones (9%), and Vietnam's Zalo is the only other platform where conversion is the top objective, at 25%. Non-use is high across the long tail: 46% of marketers do not use Tumblr, 40% do not use Twitch or Kwai, and 39% do not use Reddit or Josh.

Consumer research habits support the emphasis on consideration. Globally, 26% of consumers research products on social media and 25% buy there. Latin America leads research at 36%, Asia Pacific leads purchases at 33%, and North America trails at 20% and 19%. Among consumers under 35, 35% made a purchase on social media in the previous 12 months.

Marketers: challenges and the tool gap

Audience reach remains the most cited challenge at 40%, followed by ROAS or ROI calculation at 36%, keeping up with content trends at 35%, and brand suitability and content alignment at 34%. Managing campaigns across multiple channels reaches 33%, handling feedback or comments 29%, measuring cross-channel performance 28%, and reporting capabilities and transparency 27%. The spread between top and bottom has narrowed. "Marketers seem focused less on finding the right audiences, and more on proving their spend worked," according to the report.

Brand suitability concern varies sharply by market. In Japan, 60% of marketers cite it as a top challenge. Canada follows at 42%, India at 41%, and Australia and New Zealand and the Philippines at 40% each. At the other end sit Indonesia at 24%, and Italy and the United Kingdom at 25% each. The regional totals are 40% for Asia Pacific, 35% for Latin America, 35% for North America and 29% for EMEA. The appendix table for Latin America labels its total line "APAC Total," an apparent labeling error, as the 35% figure matches the Latin American number in the main chart. In the 2025 edition, PPC Land reported Brazil at 86%, Canada at 79% and Japan at 78% on a related brand suitability measure; the 2026 figures for those three markets are 36%, 42% and 60%, although the two questions are framed differently.

Cross-platform management produces its own list. Different optimization levers per platform frustrate 48% of marketers, rising to 59% in Latin America. Budget reallocation complexity and lack of unified cross-platform reach each draw 41%, unified reporting and inconsistent metrics 38% each, and duplicative manual work 37%. Only 6% report no challenges.

Third-party tools

According to the report, 97% of social platform advertisers use third-party media effectiveness solutions, leaving 3% who use none. Demographic targeting and measurement is the most used at 35%, according to the appendix, followed by reach and frequency measurement at 34%, algorithmic optimization at 32%, attention metrics and ROAS attribution at 30% each, and cross-platform measurement including multi-touch attribution and marketing mix modeling at 28%. Brand suitability tools rank lower: post-bid brand suitability at 26%, pre-bid fraud avoidance at 26%, pre-bid suitability avoidance at 25% and post-bid fraud measurement at 23%. The main chart on page 19 omits demographic targeting, the top item in the appendix.

DoubleVerify sets these usage numbers against 2025 answers to a different question: which tools would be most important. In 2025, 51% named audience targeting and verification, 48% post-bid media quality measurement and 42% content alignment pre-bid controls. The report concludes that "there are significant discrepancies for brand suitability tools" between appetite and adoption. Because one survey asked about importance and the other about current use, the gap is a comparison of intentions with behavior across two different samples, not a measured fall in adoption.

The product layer

The report closes with DoubleVerify's own offerings, and several pages carry promotional panels. Those claims are the company's own. Its pre-bid content data for YouTube, Meta and TikTok "refreshes dynamically as often as every hour," according to the report, consistent with the hourly refresh DoubleVerify described when it added content-level controls on Facebook and Instagram in 2025. The company says it is certified on all major platforms and independently accredited on YouTube, Facebook, Instagram and TikTok.

Impression-level attention measurement for social, first offered on Snapchat in June 2025 with eye-tracking data from Lumen Research, is now available for Snapchat and TikTok, with "YouTube and Meta integrations on the horizon," according to the report. No date is given for either.

DV Rockerbox, the attribution suite, uses view-through integrations with Meta, TikTok, Snapchat and Pinterest to see the touchpoints before each conversion, according to DoubleVerify. The report also cites a Suntory case study with a CPM reduction of "up to -29%" across four campaigns, without giving the period, markets or baseline.

Why this matters for marketers

Two readings of the same document are available. DoubleVerify's reading is that consumers are more accepting of ads, spend is growing at twice the rate of other digital channels, and the remaining problem is proof and control. The other reading comes from its own year-on-year figures: fewer consumers say they buy on social platforms, fewer say influencers move them, and one in five says it acts on no social ads at all.

Both can be true at once. Spend can grow while per-user commercial intent softens, if platforms add inventory and advertisers keep paying for reach. The report's marketer data is consistent with that: brands still buy awareness in bulk from four platforms and then spread consideration and conversion budgets across smaller, regional or niche services where non-use often exceeds a third of marketers.

The measurement question cuts across both. PPC Land reported in 2018 that Facebook allowed 35 companies to measure its ads while YouTube limited third-party measurement to six. Eight years later, DoubleVerify's welcome letter still calls transparency "a significant obstacle." Its email asks "How much AI is too much AI?" and the report's data gives a partial answer: consumers tolerate AI behind a human face, and penalize a synthetic one they can see, most strongly among women and suburban and rural audiences.

For DoubleVerify, the timing is commercial as well as editorial. The company is selling verification, attention and attribution products into the walled gardens while it awaits completion of the Nielsen acquisition, and the report is, in part, a case for those products. Its survey numbers carry stated margins of error and a disclosed fieldwork partner. Its product claims carry neither.

Timeline

Summary

Who: DoubleVerify, the ad verification company that Nielsen agreed to acquire in August 2026, with fieldwork by Sapio Research among 22,000 consumers and 2,020 marketing decision-makers. Steve Mougis, DoubleVerify's Global Chief Commercial Officer, signs the report's introduction.

What: The second edition of DoubleVerify's walled garden report, "Scroll, Stop and Shop: Walled Garden Trends." It finds that 25% of consumers bought on a social platform in the past year, down from 30% in the 2025 edition; that 48% say influencers affect their purchases, down from 54%; that 37% view an obviously AI influencer negatively; and that 97% of social advertisers use third-party tools. It cites a $408 billion EMARKETER projection for 2026 social media ad spend, up 19%.

When: The report was released by email today, October 14, 2026. The PDF was created on October 7, 2026, and the surveys were conducted in February 2026.

Where: Consumers in 21 countries and marketers in 25 countries across Asia Pacific, Europe, the Middle East and Africa, Latin America and North America, covering YouTube, Facebook, Instagram, TikTok and eleven other platforms, including the regional services Line, Kwai, Zalo and Josh.

Why: Social platforms are taking a growing share of ad budgets, but they remain closed environments where advertisers rely on platform data and third-party vendors for proof of results. The report's year-on-year declines in social purchasing and influencer impact sit alongside rising spend, and its AI data puts numbers on the brand risk of synthetic influencers.