Marketing Video Library Analysis

Score your entire Vimeo video catalog by emotional impact, message clarity, and behavioral effectiveness using Mavera Video Analysis

Scenario

Your Vimeo account holds hundreds of marketing videos spanning campaigns, product demos, testimonials, and event recordings — but you have no systematic way to know which are your strongest creative assets. This job pulls your entire video library via GET /me/videos, uploads each to Mavera Assets, and runs Video Analysis scoring every video on emotional impact, message clarity, and behavioral effectiveness. The result is a ranked catalog that tells you which videos to promote, which to retire, and which patterns your best-performing creative shares.

Architecture

Code

Python

import os, requests, time

VM = os.environ["VIMEO_ACCESS_TOKEN"]
MV = os.environ["MAVERA_API_KEY"]
VM_BASE = "https://api.vimeo.com"
MV_BASE = "https://app.mavera.io/api/v1"
VM_H = {"Authorization": f"Bearer {VM}", "Accept": "application/vnd.vimeo.*+json;version=3.4"}
MV_H = {"Authorization": f"Bearer {MV}", "Content-Type": "application/json"}

# 1. Fetch entire video library (paginated, 100 per page)
all_videos = []
page = 1
while True:
    resp = requests.get(f"{VM_BASE}/me/videos", headers=VM_H, params={
        "per_page": 100, "page": page,
        "fields": "uri,name,link,duration,stats,created_time,pictures.sizes",
        "sort": "date", "direction": "desc",
    })
    if resp.status_code == 429:
        retry = int(resp.headers.get("Retry-After", 60))
        print(f"Rate limited — waiting {retry}s")
        time.sleep(retry)
        continue
    resp.raise_for_status()
    data = resp.json()

    for v in data.get("data", []):
        vid_id = v["uri"].split("/")[-1]
        all_videos.append({
            "id": vid_id, "name": v["name"], "link": v["link"],
            "duration": v.get("duration", 0),
            "plays": v.get("stats", {}).get("plays", 0),
            "created": v.get("created_time", "")[:10],
        })

    if not data.get("paging", {}).get("next"):
        break
    page += 1
    time.sleep(0.6)

print(f"Library size: {len(all_videos)} videos")

# 2. Analyze each video via Mavera (sample top 20 by plays)
scored = []
for video in sorted(all_videos, key=lambda v: -v["plays"])[:20]:
    upload = requests.post(f"{MV_BASE}/assets", headers=MV_H, json={
        "url": video["link"], "name": video["name"][:80], "type": "video",
    }).json()

    analysis = requests.post(f"{MV_BASE}/video-analysis", headers=MV_H, json={
        "asset_id": upload["id"],
        "analysis_types": [
            "message_clarity", "emotional_impact", "behavioral_effectiveness",
            "hook_score", "pacing", "cognitive_load",
        ],
        "metadata": {"vimeo_id": video["id"], "plays": video["plays"]},
    }).json()

    for _ in range(30):
        time.sleep(3)
        status = requests.get(
            f"{MV_BASE}/video-analysis/{analysis['id']}", headers=MV_H
        ).json()
        if status.get("status") == "completed":
            break

    r = status.get("results", {})
    scored.append({
        **video,
        "clarity": r.get("message_clarity", {}).get("score", 0),
        "emotion": r.get("emotional_impact", {}).get("score", 0),
        "behavior": r.get("behavioral_effectiveness", {}).get("score", 0),
        "hook": r.get("hook_score", {}).get("score", 0),
    })
    time.sleep(1)

# 3. Catalog ranking via Mave
scores_block = "\n".join(
    f"- \"{s['name'][:50]}\" — plays: {s['plays']:,}, clarity: {s['clarity']}/100, "
    f"emotion: {s['emotion']}/100, behavior: {s['behavior']}/100, hook: {s['hook']}/100"
    for s in scored
)

ranking = requests.post(f"{MV_BASE}/mave/chat", headers=MV_H, json={
    "message": f"""Rank this marketing video library by creative quality.

SCORED CATALOG ({len(scored)} videos):
{scores_block}

Produce:
1. **Tier 1 — Promote** (top creative assets): which videos and why
2. **Tier 2 — Optimize** (strong but fixable): what to improve
3. **Tier 3 — Retire** (underperforming creative): replace with what
4. **Patterns**: What do Tier 1 videos share that Tier 3 lacks?
5. **Recommendations**: 3 specific creative briefs for new videos based on the winning patterns""",
}).json()

print("MARKETING VIDEO LIBRARY ANALYSIS")
print("=" * 60)
for s in sorted(scored, key=lambda x: -x["emotion"]):
    print(f"  {s['name'][:40]:<42} Plays:{s['plays']:>8,}  "
          f"Clarity:{s['clarity']:>3}  Emotion:{s['emotion']:>3}  "
          f"Hook:{s['hook']:>3}")
print("\n" + ranking.get("content", "")[:2000])

Example Output

MARKETING VIDEO LIBRARY ANALYSIS
============================================================
  Q1 Brand Campaign — Feel the Differen    Plays:  84,200  Clarity: 91  Emotion: 94  Hook: 87
  Customer Story — Acme Corp Transformat   Plays:  42,100  Clarity: 88  Emotion: 82  Hook: 76
  Product Demo — Enterprise Dashboard      Plays:  31,500  Clarity: 79  Emotion: 45  Hook: 62
  Webinar Replay — State of the Market     Plays:  18,300  Clarity: 72  Emotion: 38  Hook: 41
  Trade Show Booth Walkthrough             Plays:   9,400  Clarity: 54  Emotion: 29  Hook: 33

## Catalog Ranking

### Tier 1 — Promote
- "Q1 Brand Campaign" (composite: 91/100) — highest emotional impact in library.
  Opens with face close-up and music swell in first 2 seconds. Redistribute as
  paid social hero and homepage embed.
- "Customer Story — Acme Corp" (composite: 82/100) — strong narrative arc with
  clear problem→solution→result structure. Feature on case study landing pages.

### Tier 2 — Optimize
- "Product Demo" — clarity 79 but emotion 45. The demo is thorough but clinical.
  Re-cut with a customer voiceover narrating their workflow instead of feature
  bullet points. Hook score 62 suggests the first 5 seconds need a problem statement.

### Tier 3 — Retire
- "Trade Show Booth Walkthrough" — all scores below 55. This format doesn't
  translate to digital. Replace with a 60-second highlight reel from the booth
  with customer sound bites.

### Patterns
Tier 1 videos share: human faces in first frame, emotional audio within 2 seconds,
single-message focus (1 CTA per video). Tier 3 videos share: wide establishing shots,
multiple messages crammed in, no clear CTA.

Error Handling

Rate limit (429) handling

Vimeo returns a Retry-After header in seconds. The code respects this value. Free accounts hit ~100 req/min; Pro/Business accounts get ~600 req/10 min. Add exponential backoff for large libraries.

Video file access scope

The video_files scope is required to access download URLs for some endpoints. Without it, use the link field (public Vimeo URL) for Mavera asset uploads instead.

Large library pagination

Libraries with 500+ videos require multiple pages. Each page costs one API call. The code paginates automatically but add a 600ms delay between pages to stay within rate limits.

---

What’s Next

Vimeo Integration

Back to Vimeo integration overview

Pre-Publish Creative Testing

Gate publishing with quality thresholds

Video Analysis API

Full reference for POST /api/v1/video-analysis

Mave Agent

Full reference for POST /api/v1/mave/chat