content-scoring-framework
This skill should be used when the user asks to "score content quality", "grade content", "build a content scorecard", "evaluate content performance", "content quality framework", "rate content quality", "content grading system", "assess content quality", or any variation of scoring, grading, evaluating, or building frameworks to assess content quality for B2B SaaS.
Content Scoring Framework
A content scoring framework provides an objective, repeatable way to evaluate whether a piece of content meets the quality bar before publishing and whether it's performing after publishing. Without scoring, "quality" is subjective and inconsistent — what one editor calls good, another calls mediocre.
Scoring replaces opinion with measurement. Every piece gets a number. Numbers below the threshold get fixed or killed. Numbers above get published and promoted.
The Two Scorecards
Content needs two separate scores: a pre-publish quality score and a post-publish performance score.
Pre-publish quality scorecard (score before publishing)
| # | Criterion | 0 (Fail) | 1 (Acceptable) | 2 (Strong) |
|---|---|---|---|---|
| 1 | Answer in first 50 words | No clear answer | Answer present but not in first 50 words | Direct, extractable answer in first 50 words |
| 2 | Structural formatting | No tables, no lists, wall of text | Some tables or lists | Tables, lists, and scannable elements throughout |
| 3 | Factual accuracy | Unverified claims | Most claims verified | Every claim verified with source |
| 4 | Specificity | Generic ("many companies...") | Mix of generic and specific | Every claim has a number, name, or detail |
| 5 | Voice and readability | AI-sounding, hedged, passive | Mostly readable but some AI markers remain | Clear human voice, no AI markers, direct and opinionated |
| 6 | Originality | Could have been written by any AI | 1 original element (data, POV, example) | Multiple proprietary data points or unique insights |
| 7 | AEO readiness | No schema, no FAQ, generic H2s | Some AEO elements present | Full AEO: answer-first, question H2s, FAQ, schema |
| 8 | Internal linking | No internal links | 1-2 links | 3+ relevant internal links |
| 9 | CTA alignment | No CTA or mismatched CTA | CTA present but weak | Strong CTA matched to funnel stage |
| 10 | Competitive advantage | Worse than what competitors published | Comparable to competitors | Better than any competitor page on this topic |
Maximum score: 20.
| Score | Action |
|---|---|
| 16-20 | Publish. High quality |
| 12-15 | Publish with minor fixes |
| 8-11 | Major editing needed. Do not publish |
| 0-7 | Rewrite or kill. Not worth editing |
Post-publish performance scorecard (score after 90 days)
| # | Criterion | 0 (Fail) | 1 (Moderate) | 2 (Strong) |
|---|---|---|---|---|
| 1 | Organic traffic | < 50 monthly visits | 50-500 visits | 500+ visits |
| 2 | AI citations | Not cited in any engine | Cited in 1 engine | Cited in 2+ engines |
| 3 | Keyword ranking | Not ranking top 20 | Ranking 5-20 | Ranking top 5 |
| 4 | Engagement (time on page) | < 1 minute | 1-3 minutes | 3+ minutes |
| 5 | Conversion rate | 0% | 0.1-2% | 2%+ |
| 6 | Backlinks earned | 0 | 1-3 | 4+ |
| 7 | Social engagement (if distributed) | 0 engagement | Some engagement | High engagement (top 20% of posts) |
Maximum score: 14.
| Score | Action |
|---|---|
| 10-14 | High performer. Promote and refresh regularly |
| 6-9 | Moderate. Identify improvement opportunities |
| 3-5 | Underperforming. Rewrite or consolidate |
| 0-2 | Failing. Kill, redirect, or completely rebuild |
Running Content Audits with the Scorecards
Quarterly content audit process
- Score all content published in the last quarter using the post-publish scorecard
- Categorize by performance tier:
- Top 20% → Refresh and promote. These are your best assets
- Middle 60% → Identify quick wins (better CTA, AEO improvements, content refresh)
- Bottom 20% → Rewrite, consolidate, or 301 redirect to better pages
- Calculate aggregate metrics:
- Average quality score at publish time
- Average performance score at 90 days
- Correlation between pre-publish score and post-publish performance
- Identify systemic patterns:
- Which content types score highest?
- Which stages of the scorecard consistently fail?
- Which writers/editors produce the highest-scoring content?
Using scores to improve the production process
| Pattern | Root cause | Fix |
|---|---|---|
| Low accuracy scores across all content | No fact-checking step in workflow | Add mandatory accuracy review stage |
| Low originality scores | AI-generated without human enhancement | Enforce originality gate: every piece needs proprietary data or unique POV |
| Low AEO scores | AEO not in the production workflow | Add AEO checklist to publishing process |
| High quality scores but low performance | Good content, wrong topics | Revisit topic selection — prioritize higher-demand queries |
| High traffic but low conversion | Content attracts visitors but doesn't convert | Add stronger CTAs, improve internal linking to MOFU pages |
Pre-Audit Checklist
- [ ] Pre-publish scorecard adopted by editorial team
- [ ] Post-publish scorecard metrics accessible (analytics, GSC, AI monitoring)
- [ ] Scoring threshold defined (minimum score to publish)
- [ ] Quarterly audit scheduled
- [ ] Content inventory built (all published pages cataloged)
- [ ] Action framework defined (what happens for each score tier)
- [ ] Performance data aggregation automated where possible
- [ ] Scoring responsibility assigned (who scores, who reviews)
- [ ] Historical scores tracked for trend analysis
Anti-Pattern Check
- No quality bar before publishing → Without a minimum score, quality is subjective and inconsistent. Set a threshold (12+ on pre-publish scorecard) and enforce it. Content below the threshold doesn't ship
- Scoring only at publish time → Post-publish scoring tells you whether the content actually performed. A page that scored 18/20 at publish but 3/14 at 90 days has a topic or distribution problem, not a quality problem
- Scoring subjectively ("this feels like a 7") → Use the rubric. Score each criterion independently against the specific criteria. "Feels like" scores vary by person and mood
- Never killing underperforming content → Bottom 20% pages that score 0-2 post-publish are hurting your site. They dilute topical authority and waste crawl budget. Redirect or remove them
- Scoring without action → A scorecard that sits in a spreadsheet helps no one. Every score must connect to an action: publish, edit, rewrite, kill, promote, or refresh
- Only measuring traffic → Traffic without conversion is a vanity metric. The post-publish scorecard includes conversion rate and AI citations alongside traffic. A page with 100 visits and 10 demo requests is more valuable than a page with 10,000 visits and 0 demos
Want agents that use skill files like this?
We customize skill files for your brand voice and methodology, then run content agents against them.
Book a call