Knowledge Pack Files
Performance Marketing Knowledge Pack Files
Browse the source files that power the Performance Marketing MCP server knowledge pack.
sidebutton install marketing Video Creative Specs & Performance Data
Video ad format specifications, performance benchmarks, and creative testing frameworks for paid social and YouTube.
Platform Format Comparison
| Format | Platform | Length | Skip | Aspect | Billing |
|---|---|---|---|---|---|
| Skippable in-stream | YouTube | 12s-3min (rec 15-60s) | After 5s | 16:9 | CPV (30s or full) |
| Non-skippable in-stream | YouTube | 15-20s | No | 16:9 | CPM |
| Bumper | YouTube | 6s max | No | 16:9 | CPM ($3-4 CPM) |
| YouTube Shorts | YouTube | 10-60s | Swipe | 9:16 | CPV/CPM (~$4 CPM) |
| In-feed video | Meta | 1s-240min (rec <15s) | Scroll past | 1:1, 4:5, 9:16 | CPM/ThruPlay |
| Reels | Meta | 3-90s | Swipe | 9:16 | CPM |
| In-feed video | TikTok | 5-60s (rec 15-30s) | Scroll past | 9:16 | CPM |
| Spark Ads | TikTok | Any (organic boosted) | Scroll past | 9:16 | CPM |
| Sponsored video | 3s-30min (rec 15-30s) | Scroll past | 16:9, 1:1 | CPM/CPV |
"View" Definition by Platform
| Platform | What Counts as a View | Billing Trigger |
|---|---|---|
| YouTube | 30 seconds watched (or full if shorter), or interaction | CPV after view event |
| Meta | 3 seconds | ThruPlay = 15s or full |
| TikTok | Immediate on play start | CPM |
| 2 seconds at 50% pixels | CPV |
Performance Benchmarks
YouTube by Format
| Format | CPM | View Rate | Avg CPV |
|---|---|---|---|
| Skippable in-stream | $5-10 | 29-66% | $0.02-0.05 |
| Non-skippable | $6-10 | ~100% | N/A (CPM) |
| Bumper (6s) | $3-4 | ~100% | N/A (CPM) |
| YouTube Shorts | ~$4 | N/A | ~$0.02-0.04 |
| Connected TV | $9-10 | ~100% | N/A (CPM) |
Completion Rate by Duration
| Duration | Completion Rate |
|---|---|
| Under 15 seconds | 79.4% |
| 15-30 seconds | ~70% |
| 45-59 seconds | 74.8% |
| 1-2 minutes | 56% |
| 30+ seconds (all) | 51.8% |
Completion Rate by Device
| Device | Completion Rate |
|---|---|
| CTV (non-skippable) | 97.2% |
| Tablet | 91.4% |
| Desktop | 83.7% |
| Mobile | 82.1% |
Format Performance Multipliers
| Factor | Impact |
|---|---|
| Vertical (9:16) vs horizontal on mobile | +33.8% completion |
| Captions ON vs OFF | 18.4% skip rate vs 41.3% |
| 6-10s micro-ads | 68.7% unaided brand recall (highest) |
| Sound ON (TikTok) | +12% conversion rate |
| CTA display cards (TikTok) | +45% recall, +30% conversion |
Short-Form Video Creative Framework
Hook Rules (First 3 Seconds)
The first 3 seconds determine everything. 90% of ad recall is captured in the first 6 seconds.
Hook types ranked by effectiveness:
- Visual pattern interrupt — unexpected visual that breaks the scroll
- Direct address — "You need to see this" / "[Audience], this is for you"
- Problem statement — Name the pain immediately
- Contrarian claim — Challenge common belief
- Stat/number — Lead with a surprising data point
- Before/after — Show transformation in first frame
Hook benchmarks:
- Thumb-stop rate target: >25% (TikTok)
- Hook-to-hold rate target: >35-40% (viewers who watch past hook)
- 3-second view rate target: >50%
UGC vs Polished Creative
| Metric | UGC Style | Polished/Branded |
|---|---|---|
| CTR | Higher (+20-35%) | Lower |
| Conversion rate | Comparable or higher | Comparable |
| CPM | Lower (feels organic) | Higher (flagged as ad) |
| Brand recall | Lower | Higher |
| Best for | Direct response, TikTok, Reels | Brand awareness, YouTube, LinkedIn |
UGC outperforms polished for direct response on TikTok and Reels. Polished outperforms for brand campaigns on YouTube and LinkedIn.
Modular Video Testing Framework
Test video creative by isolating components, not replacing entire videos:
| Component | What to Test | Priority |
|---|---|---|
| Hook (0-3s) | Different opening frame/line | Highest — gates everything |
| Body (3-15s) | Value proposition, demo, testimonial | High |
| CTA (last 3-5s) | Offer, urgency, button text | High |
| Music/sound | Trending audio vs original vs voiceover | Medium |
| Format | Static opener vs motion vs face-to-camera | Medium |
| Length | 15s vs 30s vs 60s | Test after hook wins |
Process:
- Create 3-5 hook variants with same body/CTA
- Run 3-5 days, kill lowest CTR hooks
- Take winning hook, test 3 body variants
- Take winning body, test 3 CTA variants
- Final winner = optimized modular video
Optimal Length by Platform and Objective
| Platform | Awareness | Consideration | Conversion |
|---|---|---|---|
| YouTube | 15-20s (bumper/non-skip) | 30-60s (skippable) | 15-30s (action) |
| TikTok | 9-15s | 15-30s | 15-30s |
| Meta Reels | 6-15s | 15-30s | 15-30s |
| Meta Feed | 6-15s | 15-30s | 15-60s |
| 15-30s | 30-90s | 15-30s |
TikTok data: Videos 15+ seconds perform +45% better than shorter. Voiceover adds +12% conversion.
Video Measurement KPIs
| KPI | Formula | Use |
|---|---|---|
| VTR (View-Through Rate) | Completed views / Impressions × 100 | Overall video engagement |
| VCR (Video Completion Rate) | Complete views / Impressions × 100 | Full-watch measurement |
| CPV (Cost Per View) | Spend / Views | Cost efficiency |
| CPCV (Cost Per Completed View) | Spend / Completed views | Full-watch cost efficiency |
| ThruPlay Rate (Meta) | ThruPlays / Impressions × 100 | Meta-specific 15s view |
Benchmark targets: VCR >70%, CPV <$0.05 (YouTube), ThruPlay rate >15% (Meta).
Attribution Windows for Video
| Platform | Click-Through | View-Through | Engaged-View |
|---|---|---|---|
| Google/YouTube | 30 days | 1 day | 3 days (5s+ watched) |
| Meta | 7 days | 1 day | 1 day (10s+ watched) |
| TikTok | 7 days | 1 day | 7 days (6s+ watched) |
| 30 days | 7 days | N/A |
View-through attribution inflates video's reported impact. Use conservative windows (1-day VTA max) or discount view-throughs when comparing to search/social click-based channels.