EMR-2026-10A — Meta Creative Rankings vs Conversion Cost
Dataset EMR-2026-10A is the field record behind the meta creative rankings vs conversion cost study (Index No. 113). Clients sampled of various industries.
Conversion ranking–CPA Spearman -0.16 (CI -0.30 to -0.01).
| Dataset ID | EMR-2026-10A |
|---|---|
| Publisher | Darwin IQ Labs |
| Published | October 2026 |
| Units | 59 Meta accounts; 179 ad sets; 484 new ads |
| Auctions | Meta |
| Window | 2026 field extracts (weekly / daily panels as stated on the study page) |
| Outcome | Do Meta’s free creative diagnostics (quality, engagement, and conversion-rate rankings) identify which newly introduced … |
| Reported difference | Conversion ranking–CPA Spearman -0.16 (CI -0.30 to -0.01) |
| Narrative record | Index No. 113 |
| Series | Creative early-signal Nos. 113–119 |
EMR-2026-10B — Adset Spend Breadth vs CPA
Dataset EMR-2026-10B is the field record behind the adset spend breadth vs cpa study (Index No. 114). Clients sampled of various industries.
-0.9% next-week CPA per extra ≥10% ad (CI -2.9% to +1.0%).
| Dataset ID | EMR-2026-10B |
|---|---|
| Publisher | Darwin IQ Labs |
| Published | October 2026 |
| Units | 58 accounts; 202 ad sets; 2627 weeks |
| Auctions | Meta |
| Window | 2026 field extracts (weekly / daily panels as stated on the study page) |
| Outcome | Do ad sets where several ads each hold material spend share achieve lower conversion cost than ad sets dominated by one … |
| Reported difference | -0.9% next-week CPA per extra ≥10% ad (CI -2.9% to +1.0%) |
| Narrative record | Index No. 114 |
| Series | Creative early-signal Nos. 113–119 |
EMR-2026-10C — Minimum Impressions for Early Creative Signals
Dataset EMR-2026-10C is the field record behind the minimum impressions for early creative signals study (Index No. 115). Clients sampled of various industries.
Spend-share lift ~2.0× at 500 impressions (CI 1.14–2.71).
| Dataset ID | EMR-2026-10C |
|---|---|
| Publisher | Darwin IQ Labs |
| Published | October 2026 |
| Units | 279 new ads; 30 accounts |
| Auctions | Meta |
| Window | 2026 field extracts (weekly / daily panels as stated on the study page) |
| Outcome | How many impressions does a newly introduced Meta ad need before early spend share, CTR, or CPC stops rearranging relati… |
| Reported difference | Spend-share lift ~2.0× at 500 impressions (CI 1.14–2.71) |
| Narrative record | Index No. 115 |
| Series | Creative early-signal Nos. 113–119 |
EMR-2026-10D — CPC and CTR as Stand-ins for CPA
Dataset EMR-2026-10D is the field record behind the cpc and ctr as stand-ins for cpa study (Index No. 116). Clients sampled of various industries.
Prospecting CPC share of CPA variance −9% (CI −20% to 5%).
| Dataset ID | EMR-2026-10D |
|---|---|
| Publisher | Darwin IQ Labs |
| Published | October 2026 |
| Units | 62 accounts; 237 ad sets |
| Auctions | Meta |
| Window | 2026 field extracts (weekly / daily panels as stated on the study page) |
| Outcome | If landing pages are stable, do cheaper clicks (lower CPC) or higher CTR predict lower CPA across ads in the same conver… |
| Reported difference | Prospecting CPC share of CPA variance −9% (CI −20% to 5%) |
| Narrative record | Index No. 116 |
| Series | Creative early-signal Nos. 113–119 |
EMR-2026-10E — New-Ad Launch Bonus vs Incumbent
Dataset EMR-2026-10E is the field record behind the new-ad launch bonus vs incumbent study (Index No. 117). Clients sampled of various industries.
Median first-3-day copy share 74% (CI 58–82%).
| Dataset ID | EMR-2026-10E |
|---|---|
| Publisher | Darwin IQ Labs |
| Published | October 2026 |
| Units | 67 pairs; 19 accounts |
| Auctions | Meta |
| Window | 2026 field extracts (weekly / daily panels as stated on the study page) |
| Outcome | What share of head-to-head spend does an identical creative copy receive against its still-running original in the first… |
| Reported difference | Median first-3-day copy share 74% (CI 58–82%) |
| Narrative record | Index No. 117 |
| Series | Creative early-signal Nos. 113–119 |
EMR-2026-10F — Cross-Adset Creative Rank Transfer
Dataset EMR-2026-10F is the field record behind the cross-adset creative rank transfer study (Index No. 118). Clients sampled of various industries.
Spend-share Spearman 0.61 (CI 0.49–0.69); top-1 2.9× chance.
| Dataset ID | EMR-2026-10F |
|---|---|
| Publisher | Darwin IQ Labs |
| Published | October 2026 |
| Units | 64 pairs; 51 accounts |
| Auctions | Meta |
| Window | 2026 field extracts (weekly / daily panels as stated on the study page) |
| Outcome | When the same creative assets run in two ad sets in the same account, do spend-share (and CPC/CPA) ranks agree across th… |
| Reported difference | Spend-share Spearman 0.61 (CI 0.49–0.69); top-1 2.9× chance |
| Narrative record | Index No. 118 |
| Series | Creative early-signal Nos. 113–119 |
EMR-2026-10G — Combined Early Screen for New Creatives
Dataset EMR-2026-10G is the field record behind the combined early screen for new creatives study (Index No. 119). Clients sampled of various industries.
Holdout spends 1.85× (CI 1.28–2.76); spends+performs 1.14× (CI 0.83–1.48).
| Dataset ID | EMR-2026-10G |
|---|---|
| Publisher | Darwin IQ Labs |
| Published | October 2026 |
| Units | 2046 weekly new ads; holdout split by account |
| Auctions | Meta |
| Window | 2026 field extracts (weekly / daily panels as stated on the study page) |
| Outcome | Which simple rule built from early spend share and early CPA maximizes holdout lift on later spend retention and convers… |
| Reported difference | Holdout spends 1.85× (CI 1.28–2.76); spends+performs 1.14× (CI 0.83–1.48) |
| Narrative record | Index No. 119 |
| Series | Creative early-signal Nos. 113–119 |
EMR-2026-09D — Holiday participant CPM
Dataset EMR-2026-09D is the field record behind the holiday participant CPM study, a companion to EMR-2026-09C. It classifies 41 Meta accounts (2023–2025) as Holiday Participants (Thanksgiving–Cyber Monday daily spend ≥10% above Nov 1–7) or Non–Holiday Participants. Clients sampled of various industries.
Pooled Black Friday CPM vs early November: Non–Holiday Participants +48.6%; Holiday Participants +184.7%. Non-participants still absorb material auction-price inflation.
| Dataset ID | EMR-2026-09D |
|---|---|
| Publisher | Darwin IQ Labs |
| Published | September 2026 |
| Units | 41 Meta accounts (25 participants; 16 non) |
| Auctions | Meta |
| Window | 2023–2025 Nov–early Dec |
| Outcome | CPM lift vs early November by participation |
| Reported difference | Non-participant BF CPM +49%; participant +185% (pooled) |
| Narrative record | Index No. 112 |
| Companion | EMR-2026-09C / No. 111 |
EMR-2026-09C — November auction prices
Dataset EMR-2026-09C is the field record behind the November auction prices study. It covers account-level daily Meta CPM, link CTR, and CPC from Nov 1 through Dec 7 for 2023–2025 (41 accounts per year from the same sampled set as EMR-2026-07B). Clients sampled of various industries.
Median pattern vs early-November (Nov 1–7) baseline: Black Friday CPM +171%; Cyber Monday CPM +97%; Black Friday CTR +47%; Black Friday CPC +68%. CPM typically first exceeds +10% above baseline ~17 days before Black Friday.
| Dataset ID | EMR-2026-09C |
|---|---|
| Publisher | Darwin IQ Labs |
| Published | September 2026 |
| Units | 41 Meta accounts × daily series × 3 years (2023–2025) |
| Auctions | Meta |
| Window | Nov 1 – Dec 7 each year |
| Outcome | Daily CPM, link CTR, CPC vs early-November baseline |
| Reported difference | BF CPM +171% median; rise starts ~17 days before BF |
| Narrative record | Index No. 111 |
EMR-2026-09B — Adset creative spend split
Dataset EMR-2026-09B is the field record behind the adset creative spend split study, a companion to EMR-2026-09. It covers 103 Meta ad sets (top-3 by spend per account with ≥4 unique normalized ad names) in August 2026 from the same sampled accounts. Clients sampled of various industries.
Unweighted adset-mean spend shares: #1 creative 46.7%; #2 creative 18.5%; rest after top 2 = 34.8% (top-2 cumulative 65.2%).
| Dataset ID | EMR-2026-09B |
|---|---|
| Publisher | Darwin IQ Labs |
| Published | September 2026 |
| Units | 103 Meta ad sets (≥4 unique ad names each) |
| Auctions | Meta |
| Window | August 2026 |
| Outcome | Within-adset spend share: rank 1, rank 2, rest |
| Unit definition | Normalized ad name inside ad set |
| Reported difference | ~47% / ~19% / ~35% (top / #2 / rest) |
| Narrative record | Index No. 110 |
| Companion | EMR-2026-09 / No. 107 |
EMR-2026-09 — Creative spend concentration
Dataset EMR-2026-09 is the field record behind the creative spend concentration study. It covers August 2026 ad-level spend by normalized ad name for 41 Meta accounts drawn from the same sampled accounts as EMR-2026-07B. Clients sampled of various industries. Same ad name across campaigns/ad sets counts as one creative; names containing - copy collapse to the non-copy base.
Unweighted account-mean spend shares: top 1 creative 16.8%; top 3 creatives 34.4%; top 10 creatives 59.7%; creatives ranked 11+ 40.3%. Spend-weighted top-10 share is 48.6%.
| Dataset ID | EMR-2026-09 |
|---|---|
| Publisher | Darwin IQ Labs |
| Published | September 2026 |
| Units | 41 Meta accounts; 4,942 normalized ad-name creatives |
| Auctions | Meta |
| Window | August 2026 |
| Outcome | Share of account spend on top 1 / top 3 / top 10 / 11+ creatives |
| Unit definition | Normalized ad name (case-insensitive; - copy collapsed) |
| Reported difference | Top 1 = 16.8%; top 3 = 34.4%; top 10 = 59.7%; 11+ = 40.3% (account means) |
| Narrative record | Index No. 107 |
EMR-2026-08 — Three-second video view rates
Dataset EMR-2026-08 is the field record behind the 3-second video view drop-off study. It covers 1,152 Meta video ads with measurable spend and Meta three-second video views in July 2026, drawn from the same sampled accounts as EMR-2026-07B. Clients sampled of various industries.
Impression-weighted three-second view rate is 31.9% (68.1% drop-off before three seconds). Link CTR for the highest three-second-rate quartile is 56% higher than the lowest quartile (Pearson r = 0.35). Within-account demeaned CPM association is weak (r = 0.18).
| Dataset ID | EMR-2026-08 |
|---|---|
| Publisher | Darwin IQ Labs |
| Published | August 2026 |
| Units | 1,152 Meta video ads |
| Auctions | Meta |
| Window | July 2026 |
| Outcome | Impression-weighted 3-second view rate; link CTR; CPM |
| Contrast | Quartiles of 3-second view rate; within-account demeaned correlations |
| Reported difference | 31.9% of impressions reach 3s; Q4 CTR +56% vs Q1; within-account CPM r = 0.18 |
| Narrative record | Index No. 106 |
EMR-2026-07D — Advantage+ Creative CTR and CPM
Dataset EMR-2026-07D is the field record behind the Advantage+ Creative CTR and CPM study. It covers 4,952 Meta ads with measurable spend in June 2026 (2,927 Advantage+ Creative; 2,025 non-Advantage+), from the same sampled accounts as EMR-2026-07B. Clients sampled of various industries.
Advantage+ Creative link CTR is 21% higher than non-Advantage+ (0.86% vs 0.71%; index 121 vs 100). Advantage+ Creative CPM is 45% higher ($15.46 vs $10.69; index 145 vs 100).
| Dataset ID | EMR-2026-07D |
|---|---|
| Publisher | Darwin IQ Labs |
| Published | July 2026 |
| Units | 4,952 Meta ads (2,927 Advantage+ Creative; 2,025 non-Advantage+) |
| Auctions | Meta |
| Window | June 2026 |
| Outcome | Impression-weighted link CTR; spend-weighted CPM |
| Contrast | Advantage+ Creative vs. non-Advantage+ |
| Reported difference | A+ CTR +21%; A+ CPM +45% vs non-A+ |
| Narrative record | Index No. 109 |
EMR-2026-07C — Video vs image CTR
Dataset EMR-2026-07C is the field record behind the video vs image CTR study. It is derived from the same June 2026 Meta ad pool as EMR-2026-07B (4,952 ads: 1,199 video; 3,753 image). Clients sampled of various industries.
Impression-weighted link CTR for video is 88% higher than for image (1.32% vs 0.70%; index 188 vs 100).
| Dataset ID | EMR-2026-07C |
|---|---|
| Publisher | Darwin IQ Labs |
| Published | July 2026 |
| Units | 4,952 Meta ads (1,199 video; 3,753 image) |
| Auctions | Meta |
| Window | June 2026 |
| Outcome | Impression-weighted link CTR |
| Contrast | Video creatives vs. image creatives |
| Reported difference | Video CTR +88% vs image (index 188 vs 100) |
| Narrative record | Index No. 108 |
EMR-2026-07B — Video vs image CPM and CPC
Dataset EMR-2026-07B is the field record behind the video vs image CPM and CPC study. It covers 4,952 Meta ads with measurable spend in June 2026 (1,199 video; 3,753 image), sampled from advertiser accounts that each spent at least $100,000 that month. Clients sampled of various industries.
Spend-weighted CPM for video is 59% higher than for image (index 159 vs 100). Spend-weighted CPC for video is 16% lower than for image (index 85 vs 100).
| Dataset ID | EMR-2026-07B |
|---|---|
| Publisher | Darwin IQ Labs |
| Published | July 2026 |
| Units | 4,952 Meta ads (1,199 video; 3,753 image) |
| Auctions | Meta |
| Window | June 2026 |
| Outcome | Spend-weighted CPM; spend-weighted CPC |
| Contrast | Video creatives vs. image creatives |
| Reported difference | Video CPM +59% vs image; video CPC −16% vs image |
| Narrative record | Index No. 105 |
EMR-2026-07 — Visual element retention
Dataset EMR-2026-07 is the field record behind the 3-second visual hook retention study. It covers 4,200 video ad assets observed in active Meta Reels and TikTok auctions during the measurement window published in July 2026.
The recorded outcome is 24-hour message recall, used to estimate visual memory decay velocity. The comparison is between motion-driven text overlay hooks and purely audio-driven hooks, restricted to silent-start feeds. In that comparison, motion-driven text overlay hooks score 34% higher on 24-hour message recall.
| Dataset ID | EMR-2026-07 |
|---|---|
| Publisher | Darwin IQ Labs |
| Published | July 2026 |
| Units | 4,200 video ad assets |
| Auctions | Meta Reels, TikTok |
| Feed condition | Silent-start |
| Outcome | 24-hour message recall; visual memory decay velocity |
| Contrast | Motion-driven text overlay hooks vs. purely audio-driven hooks |
| Reported difference | 34% higher 24-hour message recall for motion-driven text overlay hooks |
| Narrative record | Index No. 104 |
What is measured
- Hook modality: motion-driven text overlay, or audio-driven with no text-overlay motion.
- Delivery context: platform auction and whether the feed starts silent.
- Memory outcome: message recall at 24 hours, summarized as decay velocity across the asset set.