indranilbanerjee/digital-marketing-pro / skills/context-engine
context-engine
VerifiedLoad shared marketing context: brand profile, benchmarks, compliance rules, specs. "load my brand context"
scoring-rubrics.md
View on GitHub# Scoring Rubrics — Standardized Evaluation Criteria
All content-producing modules use these rubrics to score marketing outputs consistently.
---
## 1. Content Quality Score (0-100)
Use for: blog posts, articles, guides, whitepapers, case studies, web pages.
| Dimension | Weight | 0 (Fail) | 5 (Weak) | 10 (Good) | 15/20 (Excellent) |
|-----------|--------|----------|----------|-----------|-------------------|
| **Audience Relevance** | 20% | Wrong audience entirely | Generic, could be for anyone | Addresses target persona | Deeply resonates with specific persona's pain points, JTBD |
| **Readability** | 15% | Grade level wildly inappropriate | 3+ grades off target | Within 1 grade of target | Perfect grade level for audience (B2C: 6-8th, B2B: 10-12th) |
| **SEO Optimization** | 15% | No keyword targeting | Keyword present but poorly placed | Primary keyword in title, H1, first 100 words, meta | Full optimization: keyword clusters, internal links, schema, intent match |
| **Originality** | 15% | Rehashed generic content | Some unique angles | Original insights or data points | First-hand experience, proprietary data, unique frameworks |
| **Brand Voice Match** | 15% | Completely off-brand | Partially aligned | Consistent with brand voice profile | Unmistakably on-brand across all 4 voice dimensions |
| **CTA Effectiveness** | 10% | No CTA or irrelevant CTA | CTA present but weak/generic | Clear CTA aligned with funnel stage | Compelling, specific, urgency-appropriate CTA with value prop |
| **Compliance** | 10% | Violations present (legal risk) | Missing optional best practices | All required disclaimers present | Full compliance + proactive best practices (accessibility, disclosure) |
**Scoring**: Sum weighted scores. 85-100 = Publish-ready. 70-84 = Minor revisions. 50-69 = Major revisions. Below 50 = Rewrite.
---
## 2. Ad Creative Score (0-100)
Use for: search ads, social ads, display ads, video ad scripts.
| Dimension | Weight | 0 (Fail) | 10 (Weak) | 18 (Good) | 25 (Excellent) |
|-----------|--------|----------|-----------|-----------|----------------|
| **Hook Strength** | 25% | No hook, starts with brand name | Generic hook, doesn't stop scroll | Good pattern interrupt or curiosity gap | Irresistible hook — specific, emotional, or contrarian first line/3 seconds |
| **Value Proposition** | 20% | Features only, no benefit | Benefit mentioned but unclear | Clear benefit tied to audience need | Specific, quantified benefit that differentiates from competitors |
| **Social Proof** | 15% | None | Generic ("trusted by thousands") | Specific proof (numbers, names, logos) | Compelling proof (exact metrics, recognizable names, visual testimonials) |
| **CTA Clarity** | 15% | Missing or confusing CTA | Generic "Learn More" | Specific action aligned with funnel stage | Action-oriented, urgency-appropriate, value-restating CTA |
| **Platform Fit** | 15% | Clearly repurposed from another platform | Correct specs but wrong tone/format | Native to platform's culture and format | Could pass as organic content on the platform |
| **Urgency/Scarcity** | 10% | Manipulative or false urgency | No urgency at all | Appropriate urgency (time, quantity, or exclusivity) | Natural urgency that enhances without manipulating |
**Scoring**: Sum weighted scores. 85+ = Launch-ready. 70-84 = Test with monitoring. Below 70 = Revise before spending.
---
## 3. Email Score (0-100)
Use for: marketing emails, newsletters, drip sequences, transactional emails with marketing elements.
| Dimension | Weight | 0 (Fail) | 10 (Weak) | 18 (Good) | 25 (Excellent) |
|-----------|--------|----------|-----------|-----------|----------------|
| **Subject Line** | 25% | Spam trigger words, misleading, too long | Generic, no curiosity or value | Clear value or curiosity, appropriate length (30-50 chars) | Irresistible: specific benefit/curiosity + personalized + mobile-optimized |
| **Body Structure** | 20% | Wall of text, no hierarchy | Some structure but not scannable | Clear hierarchy: headline → body → CTA, scannable | Inverted pyramid, single-column, mobile-first, perfect visual flow |
| **Personalization** | 15% | No personalization at all | Only {first_name} merge tag | Behavioral or segment-based personalization | Dynamic content blocks based on user behavior, preferences, lifecycle stage |
| **CTA** | 15% | No CTA or multiple competing CTAs | Text link CTA only | Single button CTA, above fold, clear action | High-contrast button, value-restating text, below supporting copy, repeated at bottom |
| **Deliverability** | 15% | Spam words, image-only, no text version | Minor spam risk signals | Clean text-to-image ratio, no spam triggers | SPF/DKIM/DMARC aligned, BIMI, low complaint history, proper list hygiene |
| **Preview Text** | 10% | Not set (shows "View in browser") | Set but doesn't add value | Complements subject line, adds context | Creates curiosity gap or reinforces subject line value prop |
**Scoring**: Sum weighted scores. 85+ = Send. 70-84 = A/B test the weak areas. Below 70 = Revise.
---
## 4. Landing Page Score (0-100)
Use for: landing pages, product pages, signup pages, event registration pages.
| Dimension | Weight | 0 (Fail) | 10 (Weak) | 18 (Good) | 25 (Excellent) |
|-----------|--------|----------|-----------|-----------|----------------|
| **Above-Fold Clarity** | 25% | Cannot understand offer in 10 seconds | Value prop present but unclear or buried | Clear headline + supporting copy + visual in 5 seconds | Instantly clear: what it is, who it's for, why it matters, what to do next |
| **Trust Signals** | 20% | No social proof or trust elements | Generic trust badges only | Specific testimonials, customer logos, review scores | Multiple trust layers: testimonials + logos + metrics + security + guarantees |
| **Form Friction** | 15% | 10+ required fields, no context | 5-8 fields, no progressive profiling | 3-4 fields, clear value exchange | Minimal fields + smart defaults + inline validation + privacy reassurance |
| **Message Match** | 15% | No connection to referral source | Partially matches ad/email | Headline echoes ad/email messaging | Perfect continuity: same language, same offer, same visual style as referral source |
| **Page Speed** | 10% | Over 5s load time | 3-5s load time | Under 3s, passes Core Web Vitals | Under 2s, all Core Web Vitals green, optimized images |
| **Mobile Experience** | 15% | Broken on mobile | Functional but awkward (horizontal scroll, tiny buttons) | Responsive, readable, functional | Thumb-friendly, 44px+ tap targets, single-column, fast mobile load |
**Scoring**: Sum weighted scores. 85+ = Drive traffic. 70-84 = Optimize before scaling spend. Below 70 = Rebuild.
---
## 5. Social Media Post Score (0-100)
Use for: organic social media posts across all platforms.
| Dimension | Weight | Criteria |
|-----------|--------|----------|
| **Platform Nativity** | 25% | Content feels native to the specific platform's culture, format, and tone |
| **Hook/Opening** | 20% | First line (text) or first 2 seconds (video) stops the scroll |
| **Value Delivery** | 20% | Post provides clear value: entertains, educates, inspires, or informs |
| **Engagement Prompt** | 15% | Includes a reason to engage (question, poll, challenge, opinion prompt) |
| **Brand Voice** | 10% | Matches brand voice profile adapted for this specific platform |
| **Visual Quality** | 10% | Image/video quality appropriate for platform (polished vs. raw as needed) |
---
## 6. Press Release / PR Pitch Score (0-100)
Use for: press releases, media pitches, HARO responses, journalist outreach.
| Dimension | Weight | Criteria |
|-----------|--------|----------|
| **Newsworthiness** | 30% | Genuine news value — not a thinly veiled ad. Would a journalist care? |
| **Lead Paragraph** | 20% | Who, what, when, where, why answered in first paragraph |
| **Quote Quality** | 15% | Executive quotes add insight, not generic praise. Quotable by media. |
| **Data/Evidence** | 15% | Statistics, research, or specific outcomes that substantiate the story |
| **Personalization** | 10% | Pitch tailored to specific journalist/outlet's beat and recent coverage |
| **Formatting** | 10% | Proper press release format: headline, dateline, boilerplate, contact info |
---
## 7. Brand Voice Consistency Score (0-100)
Use for: checking any content against the brand voice profile.
| Dimension | Weight | Criteria |
|-----------|--------|----------|
| **Formality Match** | 25% | Content formality level matches brand profile (1-10 scale) within ±1 |
| **Energy Match** | 20% | Energy/enthusiasm level matches brand profile within ±1 |
| **Vocabulary** | 20% | Uses preferred words, avoids restricted words per brand profile |
| **Personality** | 20% | Reflects stated personality traits (e.g., witty, empathetic, direct) |
| **Channel Adaptation** | 15% | Appropriately adjusts voice for the specific channel while staying on-brand |
---
## How to Apply Scoring
1. **Before creating content**: Reference the relevant rubric to guide creation
2. **After creating content**: Score the output against each dimension
3. **Report scores**: Include the total score and per-dimension breakdown
4. **Flag low dimensions**: Any dimension scoring below 50% of its max should be flagged for revision
5. **Track over time**: Scores saved to brand memory show content quality trends
## 8. Multilingual Content Score (0-100)
Evaluates translated and localized content quality across 5 dimensions:
### Dimensions
| Dimension | Weight | Description |
|-----------|--------|-------------|
| Translation Accuracy | 25% | Length ratio within expected range, meaning preservation, no omissions or additions |
| Cultural Adaptation | 25% | Cultural references appropriate for target market, social proof style adapted, urgency/trust signals localized |
| Brand Voice Preservation | 20% | Brand personality maintained in target language, tone consistent with market expectations, vocabulary aligned |
| Formatting Integrity | 15% | Markdown/HTML preserved, placeholders intact, merge tags unchanged, RTL handled correctly |
| Compliance Alignment | 15% | Market-specific disclaimers present, consent language correct, regulatory requirements met |
### Score Interpretation
| Score | Level | Action |
|-------|-------|--------|
| 85-100 | Publish Ready | Content can be published directly |
| 70-84 | Review Needed | Native speaker review recommended before publishing |
| 50-69 | Revision Required | Significant quality issues — re-translate or heavily edit |
| 0-49 | Re-translate | Fundamental quality problems — start translation fresh |
### Language-Specific Considerations
- **Indic languages**: Check script rendering, transliteration consistency, code-switching patterns (e.g., Hinglish)
- **CJK languages**: Verify character-based length ratios (0.5-0.8x English), honorific levels, character set consistency (Simplified vs Traditional Chinese)
- **RTL languages**: Confirm direction attributes, number formatting, layout mirroring
- **European languages**: Check formality level consistency, gender agreement, false friend detection
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| slack-gif-creatoranthropics/skillsKnowledge and utilities for creating animated GIFs optimized for Slack. Provides constraints, validation tools, and animation concepts. Use when users request animated GIFs for Slack like "make me a GIF of X doing Y for Slack." | Stars: 198k | Updated | Claude CodeCodex | Verified | Install |
| ponytailDietrichGebert/ponytailLazy senior dev mode for any coding task: the smallest change that fully solves it, and a reply a busy human understands in one read. | Stars: 194k | Updated | Claude CodeCodex | Verified | Install |
| ponytail-debtDietrichGebert/ponytailHarvest every shortcut: comment into one debt ledger, so deferrals get tracked instead of forgotten. One-shot report. | Stars: 194k | Updated | Claude CodeCodex | Verified | Install |
| ponytail-debtDietrichGebert/ponytailList every `shortcut:` comment (and older `ponytail:` ones) as a debt ledger. One-shot report, changes nothing. Use for "ponytail debt", "what did ponytail defer", "list the shortcuts", /ponytail-debt. | Stars: 194k | Updated | Claude CodeCodex | Verified | Install |