Video summary

L'analyse créa la plus complète de l'année dont personne ne parle (+600 000 pubs analysées)

Main summary

Key takeaways

Business

Executive summary (business/creative strategy)

Motion’s 2026 MTA creative report (Q4 recap) breaks down how to win in ad creative testing:

  • Winners are rare
  • Spend concentrates in a small percentage of creatives
  • Creative volume + diversity beat relying on “talent” or a single high-performing creative

Practical takeaway: build a repeatable creative testing system designed to reliably produce a handful of “big winners” that drive most account spend and profitability.

Key data points & KPIs mentioned

Scale analyzed

  • Creatives analyzed: 578,750
  • Total ad spend analyzed: > $1.2B (also referenced as “~$1B spend” and “600,000 ads”)

Winner rarity / spend concentration

  • Structural claim: ~95% of creations will never be spent (within the MTA model)
  • “Real winners” definition: creatives that take ~10x the median account budget
  • Top-level rarity: only ~5% of ads spend ≥10x the median account budget

Win-rate ranges (industry context)

  • ~4% to 8–9%, varying by niches/teams
  • No one consistently exceeds these ranges

Outcome distribution (by creative count)

  • 8% “true winners”
  • 38% “somewhat profitable / midrange”
  • 53% losing/underperforming initially

Outcome distribution (by spend)

  • The 8% winners drive ~60% of spend
  • Midrange spend contribution: ~26% (from referenced range)
  • Losing/test-heavy creatives: ~17% spend

Enterprise vs general volume

  • Top 25% enterprise accounts: 55 creatives avg
  • Across all accounts analyzed: 19 creatives avg

Win-rate vs scale argument (example)

  • 5 creatives @ 20% ER ⇒ 5 winners
  • 50 creatives @ 10% ER ⇒ 5 winners, but with more scaling potential due to higher volume

Account size benchmark

  • Micro-businesses: < $10,000/month
  • Micro accounts launch fewer creatives and generate fewer winners
  • Movement described as: winners increase with volume, not just “more ads” broadly

Specific examples mentioned

  • UGC win rate: 7.56% (still active; not “dead”)
  • Demo / product demo: ~8%
  • Testimony: ~6.5%
  • Podcast: ~14%
  • Offer-first hook: ~9% rate, representing ~30% of spend

Core frameworks / playbooks emphasized (action system)

1) “Winner-driven portfolio” model (anomaly hunting)

  • Goal: find 3–6 “big winners” out of ~100 creatives
  • Impact: these few winners represent ~70% of account value/spend
  • Operating analogy: treat testing like buying lots of tickets until anomalies emerge

2) “Volume beats talent” testing loop (don’t optimize ER% alone)

  • Primary metric shift: optimize for # of winners (not ER%)
  • Reason: the winner gap can vary ~2% to 8% (4x); success is rare, so more shots = more winners
  • Constraint: volume must be high-quality, diverse, and analyzed

3) “Midrange creatives stabilize CPA and feed winners”

Midrange creatives are:

  • ~40% of creations
  • They run longer (~28 days longer) than the described winners/non-winners
  • They stabilize CPA and precede/enable the big winners

4) Funnel concept: “BUFF” (bottom-of-funnel) matters early

  • Offer-first works: put the offer in the hook
    • ~9% rate and ~30% of spend
  • Recommendation: don’t rely only on high-concept storytelling—clarify the offer immediately

5) Creative element framework: Concept vs Format vs Hook

  • Concept: what’s inside (e.g., testimony, unboxing, celebrity)
  • Format: packaging/style (e.g., UGC, static, high-production images, GIFs, carousels)
  • Hook: psychological trigger (urgency/scarcity, open loop, exclusivity, confession, curiosity, price anchoring, etc.)

Recommendation: evaluate and iterate all three layers, not just the “idea.”

Concrete creative recommendations (with examples)

  • Push “Unboxing”

    • High-potential concept with relatively less common adoption, making it “extremely relevant”

  • Use “Celebrity” selectively

    • Works especially on Meta, but may require many weeks to mature
    • Prefer when it has shown spend traction in your system

  • Hook with “Now or never” scarcity/urgency

    • Put urgency/rarity in the hook, then let the script handle rational conversion
    • Avoid being only pain-point pragmatic at the hook; emotional urgency can outperform

  • Match concepts to vertical “rules”

    • Example mappings:
      • Health / VSL: strong compatibility (per report framing)
      • Fashion: carousels and TikTok-style montages
    • Recommendation: don’t copy global benchmarks blindly—validate inside your niche

  • Exploit UGC and hybrid “mash-up” formats

    • UGC is still measurable (7.56% win rate)
    • Hybrid/mash-up formats are often described as “dead,” but the report’s data suggests they still appear among top performers in the billion-spend analysis

  • Text-based formats

    • Examples mentioned: “lettre/founder,” green screen, bullet styles
    • Recommendation: prioritize styles that score well on both rate and spend (not rate alone)

How to apply the report (actionable operating guidance)

  • Stop using ER% as your sole KPI

    • ER can mislead because you must separate:
      • rate winners (high ER but limited spend)
      • rate midrange (lower ER but higher profitability/spend contribution)

  • Weekly/monthly operational goal: maximize # of winners produced

    • If winners are too low, diagnose:
      • not launching enough creatives (volume constraint)
      • not achieving sufficient quality
      • concepts/formats not well thought through
      • missing feedback loop (no learning/iteration from results)

  • Target creative pipeline structure

    • Build a system that yields high-quality diverse volume to find rare anomalies
    • Maintain midrange coverage to stabilize performance while winners are discovered

  • Benchmark by spend contribution

    • A small fraction of creatives drives most spend
    • Therefore: allocate resources to scale winners while continuing to test new variations

Presenter/source attribution

  • Presenter in subtitles: Not explicitly named (voiceover references “my notes” and “my recap”).
  • Primary source referenced: Motion (Motion’s “2026 MTA creative report” / “MTA creative recap”), including an LLM “Claude” interface mentioned in the video.

Original video