A study published on July 14, 2026 by WARC in partnership with TikTok puts a number on a problem the advertising industry has discussed anecdotally for two years: generative AI has multiplied creative output without a matching improvement in creative quality. The research, drawn from 400 marketers in the UK, US, Australia and Brazil, finds that 88% report higher creative volume since adopting the technology while only 45% report a significant improvement in quality.

The report, titled The New Creative Advantage: How community signals are reshaping creative success in the age of AI, was conducted by WARC on behalf of TikTok in May 2026. All respondents are directly involved in decisions about how marketing creative and content are produced, according to the research methodology. The survey was supplemented by interviews with senior marketers and industry experts, and combined with a review of WARC and TikTok global data.

The finding that anchors the study is not the volume gap itself but its diagnosis. According to the report, the shortfall in quality does not stem from the capabilities of the tools. It stems from what marketers are putting into them.

The inputs problem

Adoption is no longer the variable. According to the survey, 90% of marketers agree generative AI has become a key tool in the creative development process, 72% use it frequently as a regular part of content production, and 87% believe their organisation is using it effectively for creative development.

Usage clusters at specific workflow stages. Visual, image or video production leads at 73%, followed by content personalisation at 64% and concept development at 61%. Ideation and brainstorming registers 57%, video editing and post-production 54%, audience targeting and segmentation 53%. Scaling and adapting content for different platforms sits at 49%, copywriting and scripting at 42%, and testing and optimisation at 41%.

Against that adoption curve, the study identifies what it calls a demographic paradox. WARC's Marketer's Toolkit found that 59% of marketers agree traditional demographic segmentation is no longer effective. Yet the new survey found demographic data remains the primary input used to brief AI, relied on by 67% of respondents. Only 17% of marketers make it a rule to always incorporate community or audience insights beyond demographics into generative AI workflows.

Andy Yang, Global Head of Creative and Brand Ads at TikTok, framed the gap in the report's foreword. "The gap opening up in AI-assisted creativity is not a technology gap, it is an intelligence gap," he wrote. "Most brands are briefing the most powerful creative tools ever built with the weakest possible inputs: static demographics, legacy assumptions, data that tells you who someone was, not what they care about right now."

The report notes that the source of that quote is an interested party. TikTok commissioned the research, and the study's recommended solution routes through participatory platform signals of the kind TikTok generates. That framing warrants weighting by readers assessing the findings.

What marketers say is missing

Asked to identify the biggest gaps in inputs currently available when briefing generative AI, respondents pointed first to audience behavioural data covering how people engage rather than who they are, cited by 45%. High-quality brand and creative guidelines to feed into AI tools followed at 40%. Access to real-time cultural or community signals and tools that translate insights into effective AI briefs each registered 35%.

Historical performance data detailed enough to be useful was cited by 34%, and platform-native insights such as what is trending on specific channels by 33%. Only 10% of respondents said they do not perceive significant gaps in the inputs available to them.

The limitations marketers attribute to AI output track the input problem closely. Over-reliance on generic or familiar styles was the most cited limitation at 40%, followed by unpredictable and hard-to-control quality at 36% and a lack of originality or creative distinctiveness at 32%. Difficulty maintaining brand voice or visual identity registered 26%, inconsistency across assets in the same campaign 23%, and excessive human editing requirements 21%. One in five respondents reported no significant limitations.

Forward-looking sentiment is more consistent. According to the survey, 86% of marketers agree audience behaviour and community signals will influence creative development significantly more over the next three years, and one in three believe access to richer audience or cultural signals would do most to unlock generative AI advantage.

Effectiveness evidence cuts both ways

The report does not present AI creative as inherently weaker. It cites System1 testing finding AI-generated ads scored significantly above the global advertising average, and a large-scale analysis from Columbia Business School published in 2026 finding AI-generated display ads performed competitively in-market.

That evidence sits alongside independent research PPC Land has documented. A Taboola study across a publisher network of roughly 600 million daily active users found AI-generated creatives increased or maintained click-through rates without damaging conversion rates, with strongest results in food and drink and personal finance and more muted outcomes in categories such as education.

The counter-evidence concerns detection. Research covering publisher environments found that suspected AI content reduces reader trust by 50% and degrades brand advertisement performance by 14%. The distinction that emerges across both bodies of work is between AI-assisted creative that reads as native and AI output that reads as machine-made.

Contradictory data on marketer confidence

The WARC and TikTok figure of 87% of marketers believing their organisation uses generative AI effectively sits awkwardly against other industry measurement. A MiQ survey from November 2025, reported by PPC Land, found 72% of marketers planning AI adoption growth while only 45% felt confident. The same MIQ Global AI Confidence Curve is cited inside the WARC report itself for a separate finding: only 40% of marketers strongly believe they can apply findings from one campaign to another.

The two confidence measures are not directly comparable, since they ask different questions of different samples in different periods. But the divergence is worth flagging. A self-reported 87% effectiveness rating from a vendor-commissioned survey and a 45% confidence rating from an independent one describe the same industry.

A separate strand of research reinforces the audience-understanding thesis independently. Brandwatch survey findings reported in April 2026 identified social listening and consumer research as a top skill gap at 55%, with cross-platform content optimisation at 52%.

The Intelligence Loop and the S.C.A.L.E framework

The report structures its prescription around what it calls the Intelligence Loop, a four-stage cycle. Participation creates signals, as community engagement generates layered cultural and intent data. Signals reveal demand, with participation data surfacing leading indicators of emerging creative opportunity. Demand shapes creative, as signals are routed into the AI briefing process. Creative fuels participation, with community-informed work earning engagement that generates fresh signals for the next cycle.

Chapter three sets out five sequential principles under the acronym S.C.A.L.E. Select covers mapping objectives before tools, aligning media, creative and audience objectives before briefing AI. Connect treats creators as sources of community intelligence rather than distribution. Anchor addresses encoding distinctive brand assets and guardrails into the workflow before production begins. Lead concerns building governance, transparency and review into the process from the start. Evolve treats response data and performance learning as inputs to the next brief.

The report also supplies a two-axis matrix for deciding where generative AI belongs. Cultural proximity measures how rich the community intelligence available to inform a brief is. Creative ambition measures how much brand distinctiveness, cultural nuance and emotional specificity matter to a given piece of work. High proximity with lower ambition is designated scale with confidence, where AI leads on versioning and platform-native adaptation. High proximity with higher ambition is amplify with judgment, where AI assists and humans lead. Low proximity with lower ambition carries the label proceed with caution, with the stated risk being mediocrity at scale.

Platform data and its provenance

Several figures in the study derive from TikTok's own measurement operation rather than the survey. WARC analysis of TikTok and Emplifi data records a 200% year-over-year increase in median brand follower count on TikTok and a median engagement rate of 27.6% for content on the platform in the fourth quarter of 2025, with TikTok generating at least twice the median interactions of Instagram and Facebook.

Separate TikTok Marketing Science studies supply further numbers. A Kantar-conducted study of 1,500 users across the US, UK and Brazil found 81% of TikTok users agree that when loyal to a brand they are likely to engage with it online. A Global Future of Search Study conducted by WARC found 61% of TikTok search users say the platform's search results are more likely to motivate an action compared with other platforms, and recorded brand claim believability at 37% for the brand as message source against 45% for a creator on TikTok and 45% for a neighbour.

The report also carries case studies supplied by TikTok, with the document stating that claims related to brand products are brand-reported and not endorsed or verified by TikTok. Vucar, a Vietnamese used-car startup, used Symphony Creative Agent and reported a four to six times increase in creative output, a 50 to 75% reduction in time per draft and a three to four times faster ideation cycle. Experience Abu Dhabi used TikTok's Content Suite to identify existing user-generated content and repurpose it into paid assets, reporting 5.3 million users reached across the GCC, double the six-second view-through rate against benchmark and a 73% reduction in six-second cost per view. Dr Pepper developed flavour variants in response to TikTok trends and, during the period covered, surpassed Pepsi to become the second-largest US soft drink, with new variants achieving a 50% sales increase against previous launches.

Where the tooling sits

The report's Symphony references connect to a product line PPC Land has tracked since its launch. TikTok introduced Symphony in June 2024 as a generative AI suite spanning an assistant, video generation and translation tools. Dreamina Seedance 2.0 was integrated into Symphony Creative Studio in May 2026, alongside general availability to all logged-in TikTok for Business users.

The agentic layer arrived more recently. Symphony Agent, announced on June 22, 2026, takes a product image or campaign objective and produces scripts, videos, creator shortlists and outreach invitations across Symphony Creative Studio, Content Suite and TikTok One, generating a 12-second video in approximately three minutes and compressing creator shortlists to 20.

Content Suite, referenced in the Abu Dhabi case study, expanded globally in May 2026 as part of a TikTok One updatethat also introduced Creator AI Search. TikTok internal data science analysis from April 2024 cited in that announcement claimed Content Suite surfaces up to 44 times more results than manual search.

The governance dimension

Chapter three addresses transparency as a structural requirement rather than a compliance afterthought. Citing a TikTok Marketing Science Global Gen AI Consumer Perceptions and Usage Survey from April 2024, the report records that 66% of consumers want brands to ensure ethical use of generative AI, 60% want all generative AI content clearly labelled, 59% want the accuracy of generative AI content vetted and 42% want the prompts behind development shared.

Halil Dogan, Brand Steering Manager at Henkel Western Europe, is quoted in the report on consumer reaction. "Consumers react negatively when something feels like too much AI. Authenticity still wins, and AI is there to support it, not sterilize it."

The timing gives that section unusual weight for European advertisers. Article 50 transparency obligations under the EU AI Act become applicable on August 2, 2026, eleven days after this article's publication. The European Commission published its guidelines and finalised Code of Practice on those obligations on July 20, 2026, and the carve-out narrowing disclosure for artistic and satirical work does not extend to advertising. Google has already shifted AI ad labeling liability toward advertisers and rolled the labeling control across five ad products.

Consumer sentiment data from outside the TikTok orbit sharpens the point further. IAB research published in January 2026 found 82% of advertising executives believe Gen Z and Millennial consumers feel positively about AI-generated ads, against 45% of those consumers who actually do - a 37-point gap that widened from 32 points in 2024.

Why this matters for the marketing community

The practical consequence of the study's central finding is a reallocation question rather than a tooling question. If 90% of marketers already treat generative AI as a key creative tool and 67% still brief it with demographic data, the marginal return on further tool adoption is lower than the marginal return on improving briefing inputs. That reframes generative AI budget conversations from software procurement toward audience intelligence infrastructure.

The report supplies external corroboration for the audience-signal thesis. Kantar data cited in the study finds that customising content for context can increase campaign contribution to brand equity by 57%, while CreativeX estimates that 50% of media budgets are still spent behind ads not built for the platforms they run on. That combination describes a measurable efficiency loss independent of who commissioned the research.

The Schwartz Theory of Basic Human Values, validated across more than 80 countries, appears in the report as the theoretical basis for values-driven segmentation. According to the LIONS Advisory and Aletheia Values Influence Study published in July 2025, combining values-based segmentation with demographics improves predictive accuracy by 75 to 193% depending on segment.

There is a structural caveat worth stating. The research routes toward participatory platform signals, and the platform best positioned to supply those signals commissioned the research. Marketers evaluating the recommendations face a question the report does not resolve: whether community intelligence advantage is portable across platforms or whether it accrues primarily to the environment that generates and holds the signals. The report's own framing - that signals are generated, aggregated and connected to creative performance in the same environment - suggests the latter, which is a competitive argument as much as a creative one.

Timeline

Summary

Who: WARC, the marketing effectiveness research organisation and part of LIONS, in partnership with TikTok. The survey covered 400 marketers directly involved in creative and content production decisions. The report includes interviews with Tom Roach of Jellyfish, Becky Owen of Billion Dollar Boy, Marcos Angelides of Publicis Media, Zoe Scaman of Bodacious, cultural strategist Rachel Lowenstein and Elav Horwitz of WPP, with forewords from Andy Yang of TikTok and Lexi Wolf of LIONS Advisory.

What: A research report titled The New Creative Advantage finding that 88% of marketers report higher creative volume from generative AI while only 45% report significant quality improvement. The study identifies briefing inputs as the constraint, with 67% relying on demographic data as the primary input despite 59% agreeing demographic segmentation is no longer effective. It proposes an Intelligence Loop model and a five-step S.C.A.L.E framework.

When: The report was published on July 14, 2026. The underlying survey was conducted in May 2026. Supporting platform data covers the fourth quarter of 2025 and earlier TikTok Marketing Science studies dating to 2024 and 2025.

Where: Survey respondents were drawn from the UK, United States, Australia and Brazil. Case studies cover Vietnam, the Gulf Cooperation Council region and the United States. WARC operates from London, New York and Singapore.

Why: The research addresses a measurable gap between generative AI production scale and creative outcomes. Its stated argument is that the constraint is intelligence rather than technology, and that community and behavioural signals produce better creative briefs than demographic labels. The commissioning relationship means the recommended solution routes toward participatory platform environments, a framing that readers should weigh when assessing the findings.