Performance Creative Research Index

An open-access database cross-referencing peer-reviewed marketing science and primary field data from Darwin IQ Labs against official ad platform directives.

Editorial Governance Charter

This index operates under strict academic neutrality. Ingestion is restricted to peer-reviewed academic publications (Pillar 1), primary empirical field data generated by Darwin IQ Labs (Pillar 1), and official platform technical documentation (Pillar 2) from Meta, Google, and TikTok. Third-party agency commentary, vendor case studies, and unverified marketing blogs are explicitly excluded.

Pillar 1 · Darwin IQ Labs Empirical Study

Holiday Participation and Meta CPM Exposure During Black Friday Week

Across 41 Meta accounts (2023–2025) from clients sampled of various industries, Non–Holiday Participants—accounts whose Thanksgiving–Cyber Monday daily spend was not ≥10% above early November—still saw spend-weighted Black Friday CPM +48.6% vs early November. Holiday Participants saw larger pooled lifts (+184.7% on Black Friday). Sitting out the holiday push does not avoid auction-price inflation.

Publication Date Sep 2026
Dataset ID EMR-2026-09D
Sample Size n = 41 Accounts
BibTeX 
@techreport{darwiniq2026holidayparticipantcpm,
  title = {Holiday Participation and Meta CPM Exposure During Black Friday Week},
  author = {{Darwin IQ Labs}},
  year = {2026},
  institution = {Darwin IQ Research Index},
  number = {112},
  url = {https://darwiniq.com/research/studies/2026-holiday-participant-cpm}
}
Pillar 1 · Darwin IQ Labs Empirical Study

Seasonal Meta Auction Prices Around Black Friday and Cyber Monday

Across 41 Meta accounts from clients sampled of various industries, daily account-level metrics for November–early December in 2023–2025 show spend-weighted CPM typically crossing +10% above the early-November baseline about 8–20 days before Black Friday (median 17 days). Black Friday CPM lands about +140% to +179% vs that baseline (median +171%). Cyber Monday remains elevated (~+97% median) but below the BF peak; the following days ease further. On Black Friday, link CTR and CPC both rise vs early November (median +47% and +68%).

Publication Date Sep 2026
Dataset ID EMR-2026-09C
Sample Size n = 41 Accounts
BibTeX 
@techreport{darwiniq2026novemberauctionprices,
  title = {Seasonal Meta Auction Prices Around Black Friday and Cyber Monday},
  author = {{Darwin IQ Labs}},
  year = {2026},
  institution = {Darwin IQ Research Index},
  number = {111},
  url = {https://darwiniq.com/research/studies/2026-november-auction-prices}
}
Pillar 1 · Darwin IQ Labs Empirical Study

Within-Adset Creative Spend Splits Among Top Meta Ad Sets

Across 103 qualifying Meta ad sets (top-3 by spend per account, each with ≥4 unique ad names) from clients sampled of various industries in August 2026, the unweighted mean spend share is 46.7% on the #1 ad-name creative, 18.5% on the #2 creative, and 34.8% on the remaining creatives combined (top-2 cumulative = 65.2%). Medians are close (44.5% / 17.7% / 35.7%). Rank-2 shares are especially tight (IQR 14–20%), suggesting a familiar within-adset budget pattern rather than account-wide campaign mix alone.

Publication Date Sep 2026
Dataset ID EMR-2026-09B
Sample Size n = 103 Ad Sets
BibTeX 
@techreport{darwiniq2026adsetcreativespendsplit,
  title = {Within-Adset Creative Spend Splits Among Top Meta Ad Sets},
  author = {{Darwin IQ Labs}},
  year = {2026},
  institution = {Darwin IQ Research Index},
  number = {110},
  url = {https://darwiniq.com/research/studies/2026-adset-creative-spend-split}
}
Pillar 1 · Darwin IQ Labs Empirical Study

Creative Spend Concentration on Top Ad-Name Assets in Meta Accounts

Across 41 Meta accounts from clients sampled of various industries in August 2026, the unweighted account-mean spend share is 16.8% on the single top ad-name creative, 34.4% on the top 3, and 59.7% on the top 10—leaving 40.3% on creatives ranked 11 and below. Spend-weighted means are slightly less concentrated (top 10 = 48.6%), indicating larger accounts diversify spend across more named creatives.

Publication Date Sep 2026
Dataset ID EMR-2026-09
Sample Size n = 41 Accounts
BibTeX 
@techreport{darwiniq2026creativespendconcentration,
  title = {Creative Spend Concentration on Top Ad-Name Assets in Meta Accounts},
  author = {{Darwin IQ Labs}},
  year = {2026},
  institution = {Darwin IQ Research Index},
  number = {107},
  url = {https://darwiniq.com/research/studies/2026-creative-spend-concentration}
}
Pillar 1 · Darwin IQ Labs Empirical Study

Three-Second Video View Rates and Their Association with Link Click-Through in Meta Auctions

Across n=1,152 Meta video ads from clients sampled of various industries in July 2026 auctions, 31.9% of impressions register a three-second video view—so 68.1% drop off before three seconds. Ads in the highest quartile of three-second view rate show 56% higher impression-weighted link CTR than the lowest quartile (Pearson r = 0.35). After removing each advertiser’s mean CPM, the association between three-second rate and CPM is weak (within-account r = 0.18).

Publication Date Aug 2026
Dataset ID EMR-2026-08
Sample Size n = 1,152 Video Ads
BibTeX 
@techreport{darwiniq2026video3sviewrate,
  title = {Three-Second Video View Rates and Their Association with Link Click-Through in Meta Auctions},
  author = {{Darwin IQ Labs}},
  year = {2026},
  institution = {Darwin IQ Research Index},
  number = {106},
  url = {https://darwiniq.com/research/studies/2026-video-3s-view-rate}
}
Pillar 1 · Darwin IQ Labs Empirical Study

Click-Through and Delivery Cost Differentials for Advantage+ Creative versus Non-Advantage+ Ads in Meta Auctions

Across n=4,952 Meta ads from clients sampled of various industries in June 2026 auctions, Advantage+ Creative ads show 21% higher impression-weighted link CTR than non-Advantage+ ads (0.86% vs 0.71%; index 121 vs 100). The same pool shows Advantage+ Creative CPM 45% higher than non-Advantage+ ($15.46 vs $10.69; index 145 vs 100). Higher CTR is observed; lower CPM / favorable delivery-price treatment is not.

Publication Date Jul 2026
Dataset ID EMR-2026-07D
Sample Size n = 4,952 Ads
BibTeX 
@techreport{darwiniq2026advantageplusctrcpm,
  title = {Click-Through and Delivery Cost Differentials for Advantage+ Creative versus Non-Advantage+ Ads in Meta Auctions},
  author = {{Darwin IQ Labs}},
  year = {2026},
  institution = {Darwin IQ Research Index},
  number = {109},
  url = {https://darwiniq.com/research/studies/2026-advantage-plus-ctr-cpm}
}
Pillar 1 · Darwin IQ Labs Empirical Study

Link Click-Through Differentials for Video versus Image Ads in Meta Auctions

Across n=4,952 Meta ads from clients sampled of various industries in June 2026 auctions, impression-weighted link CTR for video creatives is 88% higher than for image creatives (1.32% vs 0.70%; index 188 vs 100). In this observational contrast, video ads do show higher CTRs than image ads.

Publication Date Jul 2026
Dataset ID EMR-2026-07C
Sample Size n = 4,952 Ads
BibTeX 
@techreport{darwiniq2026videoimagectr,
  title = {Link Click-Through Differentials for Video versus Image Ads in Meta Auctions},
  author = {{Darwin IQ Labs}},
  year = {2026},
  institution = {Darwin IQ Research Index},
  number = {108},
  url = {https://darwiniq.com/research/studies/2026-video-image-ctr}
}
Pillar 1 · Darwin IQ Labs Empirical Study

Delivery Cost and Click Cost Differentials for Video versus Image Ads in Meta Auctions

Across n=4,952 Meta ads from clients sampled of various industries in June 2026 auctions, spend-weighted delivery cost (CPM) for video creatives is 59% higher than for image creatives (index 159 vs 100). The same sample shows video CPC 16% lower than image CPC (index 85 vs 100), indicating higher auction prices to deliver video do not translate into higher cost per click.

Publication Date Jul 2026
Dataset ID EMR-2026-07B
Sample Size n = 4,952 Ads
BibTeX 
@techreport{darwiniq2026videoimagecpmcpc,
  title = {Delivery Cost and Click Cost Differentials for Video versus Image Ads in Meta Auctions},
  author = {{Darwin IQ Labs}},
  year = {2026},
  institution = {Darwin IQ Research Index},
  number = {105},
  url = {https://darwiniq.com/research/studies/2026-video-image-cpm-cpc}
}
Pillar 1 · Darwin IQ Labs Empirical Study

Visual Element Retention and Memory Decay Rates in Mobile Feed Auctions

We evaluate n=4,200 video ad assets across active Meta Reels and TikTok auctions to calculate visual memory decay velocity. Empirical tracking reveals that motion-driven text overlay hooks achieve a 34% higher 24-hour message recall score compared to purely audio-driven hooks in silent-start feeds, directly contrasting broad sound-on optimization guidelines.

Publication Date Jul 2026
Dataset ID EMR-2026-07
Sample Size n = 4,200 Assets
BibTeX 
@techreport{darwiniq2026hookdecay,
  title = {Visual Element Retention and Memory Decay Rates in Mobile Feed Auctions},
  author = {{Darwin IQ Labs}},
  year = {2026},
  institution = {Darwin IQ Research Index},
  number = {104},
  url = {https://darwiniq.com/research/studies/2026-hook-decay}
}
Pillar 1 · Journal of Marketing Research

Emotion-Induced Engagement in Internet Video Advertisements

Teixeira, Wedel, and Pieters (2012) track joy and surprise with automated facial coding, attention concentration with eye-tracking, and retention via free zapping across 28 internet video ads (~145,000 frames). Surprise level concentrates attention; joy velocity reduces zapping. Emotion measures explain 72% of zapping variation.

Authors Teixeira, Wedel, Pieters
DOI 10.1509/jmr.10.0207
Sample 28 video ads · ~145k frames
BibTeX 
@article{teixeira2012emotion,
  title = {Emotion-Induced Engagement in Internet Video Advertisements},
  author = {Teixeira, Thales and Wedel, Michel and Pieters, Rik},
  journal = {Journal of Marketing Research},
  year = {2012},
  doi = {10.1509/jmr.10.0207}
}
Pillar 1 · Journal of Marketing

Attention Capture and Transfer in Advertising: Brand, Pictorial, and Text-Size Effects

Pieters and Wedel (2004) eye-track 1,363 print ads across 3,600+ consumers. The pictorial captures attention independent of size; text capture scales with surface size; the brand most effectively transfers attention to other elements. Only text-size increases raise net attention to the full ad.

Authors Pieters & Wedel
DOI 10.1509/jmkg.68.2.36.27794
Sample 1,363 ads · 3,600+ consumers
BibTeX 
@article{pieters2004attention,
  title = {Attention Capture and Transfer in Advertising: Brand, Pictorial, and Text-Size Effects},
  author = {Pieters, Rik and Wedel, Michel},
  journal = {Journal of Marketing},
  volume = {68},
  number = {2},
  pages = {36--50},
  year = {2004},
  doi = {10.1509/jmkg.68.2.36.27794}
}
Pillar 1 · Journal of Marketing

The Stopping Power of Advertising: Measures and Effects of Visual Complexity

Pieters, Wedel, and Batra (2010) separate feature complexity from design complexity across 249 eye-tracked ads. Feature density hurts brand attention and ad attitude; design complexity helps pictorial and whole-ad attention, comprehensibility, and attitude — and is under advertiser control.

Authors Pieters, Wedel, Batra
DOI 10.1509/jmkg.74.5.048
Sample 249 ads · eye-tracking
BibTeX 
@article{pieters2010stopping,
  title = {The Stopping Power of Advertising: Measures and Effects of Visual Complexity},
  author = {Pieters, Rik and Wedel, Michel and Batra, Rajeev},
  journal = {Journal of Marketing},
  volume = {74},
  number = {5},
  pages = {48--60},
  year = {2010},
  doi = {10.1509/jmkg.74.5.048}
}
Pillar 1 · Journal of Marketing

A Meta-Analysis of When and How Advertising Creativity Works

Rosengren, Eisend, Koslow, and Dahlén (2020) synthesize 878 effect sizes from 67 papers. Creativity’s strongest link is ad attitude (r = .491); brand recall and recognition are not significant. Originality routes through affect transfer; appropriateness through signaling. Effects are stronger for high-involvement products.

Authors Rosengren, Eisend, Koslow, Dahlén
DOI 10.1177/0022242920929288
Evidence base 878 effect sizes · 67 papers
BibTeX 
@article{rosengren2020creativity,
  title = {A Meta-Analysis of When and How Advertising Creativity Works},
  author = {Rosengren, Sara and Eisend, Martin and Koslow, Scott and Dahl{\'e}n, Micael},
  journal = {Journal of Marketing},
  volume = {84},
  number = {6},
  pages = {39--56},
  year = {2020},
  doi = {10.1177/0022242920929288}
}
Pillar 1 · Journal of Marketing

Breaking through Fast-Forwarding: Brand Information and Visual Attention

Brasel and Gips (2008) eye-track DVR fast-forwarded TV ads. Attention rises but collapses to screen center. Central brand information still creates memory after 95% frame loss and no audio; peripheral branding is nearly worthless. Heavy central branding can still move attitude, intent, and choice.

Authors Brasel & Gips
DOI 10.1509/jmkg.72.6.031
Condition Fast-forward · silent
BibTeX 
@article{brasel2008fastforward,
  title = {Breaking through Fast-Forwarding: Brand Information and Visual Attention},
  author = {Brasel, S. Adam and Gips, James},
  journal = {Journal of Marketing},
  volume = {72},
  number = {6},
  pages = {31--48},
  year = {2008},
  doi = {10.1509/jmkg.72.6.031}
}
Pillar 1 · Journal of Advertising Research

Best Measures of Attention To Creative Tactics in TV Advertising

Bellman et al. (2019) map eye-tracking, skin conductance, and heart rate across 10 creative devices in 100+ TV ads with known sales results. No single attention measure is enough; heart rate links most strongly to market performance. Some “attention” devices, including packaging shots, can reduce attention and recall.

Authors Bellman, Nenycz-Thiel, Kennedy, Hartnett, Varan
DOI 10.2501/JAR-2019-002
Sample 100+ ads · 10 creative devices
BibTeX 
@article{bellman2019attention,
  title = {Best Measures of Attention To Creative Tactics in TV Advertising},
  author = {Bellman, Steven and Nenycz-Thiel, Magda and Kennedy, Rachel and Hartnett, Nicole and Varan, Duane},
  journal = {Journal of Advertising Research},
  volume = {59},
  number = {3},
  pages = {295--311},
  year = {2019},
  doi = {10.2501/JAR-2019-002}
}
Pillar 1 · Management Science

Breaking Through the Clutter: Advertisement Originality and Familiarity

Pieters, Warlop, and Wedel (2002) eye-track 119 consumers across 58 magazine ads. Original ads increase brand attention. Ads that are both original and familiar attract the most brand attention and improve brand memory — including a direct memory path beyond attention.

Authors Pieters, Warlop, Wedel
DOI 10.1287/mnsc.48.6.765.192
Sample 119 consumers · 58 ads
BibTeX 
@article{pieters2002clutter,
  title = {Breaking Through the Clutter: Benefits of Advertisement Originality and Familiarity for Brand Attention and Memory},
  author = {Pieters, Rik and Warlop, Luk and Wedel, Michel},
  journal = {Management Science},
  volume = {48},
  number = {6},
  pages = {765--781},
  year = {2002},
  doi = {10.1287/mnsc.48.6.765.192}
}
Pillar 1 · Online Information Review

Brand Recall of Skippable vs Non-Skippable Ads in YouTube

Belanche, Flavián, and Pérez-Rueda (2020) show brand-timing optima reverse by format. Skippable: pre-arousal recall 27.3% vs 8.2% during arousal. Non-skippable: post-arousal 28.6% vs 8.6% pre-arousal. Active skippable audiences process the front of the ad.

Authors Belanche, Flavián, Pérez-Rueda
DOI 10.1108/OIR-01-2019-0035
Design 3×2 timing × format
BibTeX 
@article{belanche2020skippable,
  title = {Brand Recall of Skippable vs Non-Skippable Ads in YouTube},
  author = {Belanche, Daniel and Flavi{\'a}n, Carlos and P{\'e}rez-Rueda, Alfredo},
  journal = {Online Information Review},
  volume = {44},
  number = {3},
  pages = {545--562},
  year = {2020},
  doi = {10.1108/OIR-01-2019-0035}
}
Pillar 1 · Journal of Marketing Research

Why Do Consumers Stop Viewing Television Commercials?

Woltman Elpers, Wedel, and Pieters (2003) show moment-to-moment entertainment raises continued viewing while information value raises stoppage. The two factors interact multiplicatively — their joint effect on stop probability is stronger than either alone.

Authors Woltman Elpers, Wedel, Pieters
DOI 10.1509/jmkr.40.4.437.19393
Design Two experiments · moment-to-moment
BibTeX 
@article{elpers2003stop,
  title = {Why Do Consumers Stop Viewing Television Commercials?},
  author = {Woltman Elpers, Josephine L. C. M. and Wedel, Michel and Pieters, Rik G. M.},
  journal = {Journal of Marketing Research},
  volume = {40},
  number = {4},
  pages = {437--453},
  year = {2003},
  doi = {10.1509/jmkr.40.4.437.19393}
}
Pillar 1 · Journal of Consumer Research

Visual Rhetoric in Advertising: Text-Interpretive, Experimental, and Reader-Response Analyses

McQuarrie and Mick (1999) show visual rhyme, antithesis, metaphor, and pun raise elaboration and ad attitude without making ads harder to comprehend. Metaphor and pun effects collapse when viewers lack the cultural competency to decode the ads.

Authors McQuarrie & Mick
DOI 10.1086/209549
Figures Rhyme · antithesis · metaphor · pun
BibTeX 
@article{mcquarrie1999rhetoric,
  title = {Visual Rhetoric in Advertising},
  author = {McQuarrie, Edward F. and Mick, David Glen},
  journal = {Journal of Consumer Research},
  volume = {26},
  number = {1},
  pages = {37--54},
  year = {1999},
  doi = {10.1086/209549}
}
Pillar 1 · Marketing Science

Eye Fixations on Advertisements and Memory for Brands

Wedel and Pieters (2000) eye-track 88 consumers across 65 magazine ads. Pictorial and brand fixations promote accurate brand memory; text fixations do not. Brand surface is especially important. More extracted information shortens memory latency; later ads show a recency advantage.

Authors Wedel & Pieters
DOI 10.1287/mksc.19.4.297.11794
Sample 88 consumers · 65 ads
BibTeX 
@article{wedel2000fixations,
  title = {Eye Fixations on Advertisements and Memory for Brands},
  author = {Wedel, Michel and Pieters, Rik},
  journal = {Marketing Science},
  volume = {19},
  number = {4},
  pages = {297--312},
  year = {2000},
  doi = {10.1287/mksc.19.4.297.11794}
}
Pillar 1 · AAAI 2021

Efficient Optimal Selection for Composited Advertising Creatives with Tree Structure

Chen et al. (2021) introduce AES: ingredient trees, dynamic programming, and Thompson sampling for composited creative selection under sparse CTR feedback. Beats ε-greedy, UCB, and independent greedy on convergence and CTR; ~8× faster than full enumeration on their real dataset.

Authors Chen, Ge, Jiang, Lian, Zheng
DOI 10.1609/aaai.v35i5.16516
Method Tree DP · Thompson sampling
BibTeX 
@inproceedings{chen2021aes,
  title = {Efficient Optimal Selection for Composited Advertising Creatives with Tree Structure},
  author = {Chen, Jin and Ge, Tiezheng and Jiang, Gangwei and Lian, Defu and Zheng, Kai},
  booktitle = {Proceedings of the AAAI Conference on Artificial Intelligence},
  volume = {35},
  number = {5},
  pages = {3967--3975},
  year = {2021},
  doi = {10.1609/aaai.v35i5.16516}
}
Pillar 1 · The Web Conference (WWW ’21)

Automated Creative Optimization for E-Commerce Advertising

Chen et al. (2021) introduce AutoCO: AutoML search for creative-element interactions, variational posteriors, and Thompson Sampling under sparse combinatorial feedback. Offline regret improves versus baselines; an online A/B test reports a 7% CTR lift.

Authors Chen et al. (Alibaba / USTC / UESTC)
DOI 10.1145/3442381.3449909
Online result +7% CTR
BibTeX 
@inproceedings{chen2021autoco,
  title = {Automated Creative Optimization for E-Commerce Advertising},
  author = {Chen, Jin and Xu, Ju and Jiang, Gangwei and Ge, Tiezheng and Zhang, Zhiqiang and Lian, Defu and Zheng, Kai},
  booktitle = {Proceedings of the Web Conference 2021},
  year = {2021},
  doi = {10.1145/3442381.3449909}
}
Pillar 2 · Official Platform Directive Audit

Meta Advantage+ Creative Enhancements: Platform Disclosures vs. Auction Outcomes

An index of Meta’s official developer specifications for dynamic creative assembly, dynamic aspect ratio cropping, and auto-generated brightness adjustments. Cross-referenced against Darwin IQ Labs performance data to isolate when automated visual adjustments improve lower-funnel conversion efficiency versus when they trigger visual fatigue.

Source Authority Meta Business & Developer Specs
API Reference Graph API v21.0
Classification Auction Mechanics
BibTeX 
@techreport{darwiniq2026advantage,
  title = {Meta Advantage+ Creative Enhancements: Platform Disclosures vs. Auction Outcomes},
  author = {{Darwin IQ Research}},
  year = {2026},
  institution = {Darwin IQ Research Index},
  number = {402},
  url = {https://darwiniq.com/research/platform/meta-advantage-creative-spec}
}
Pillar 1 · Literature Map

Cross-Element Combinatorial Selection for Multi-Element Creative Assembly

Index map of the exploration–exploitation problem in dynamic creative optimization under audience sparsity. Primary computational source: Chen et al. (2021) AutoCO at WWW ’21. Companion attention sources: Pieters & Wedel element maps and Teixeira et al. video engagement.

Primary source Chen et al., WWW ’21
Subject Computational Advertising
Type Literature synthesis
BibTeX 
@misc{darwiniq2026combinatorial,
  title = {Cross-Element Combinatorial Selection for Multi-Element Creative Assembly},
  author = {{Darwin IQ Research Index}},
  year = {2026},
  howpublished = {Literature map citing Chen et al. WWW 2021 AutoCO and related bandit creative-selection work},
  url = {https://darwiniq.com/research/academic/multimodal-creative-selection}
}

Index Directory