📅 Day 13 of 30 · Week 2: Connect & Create · ⏱ 15 minutes
🔓 No key or cannot install? Do this instead. No MCP? Copy your recent posts into Claude.ai and run the same analysis prompts there. You gather the content by hand; the analysis is the same.
Connecting Content Analysis to Editorial Decisions
Most teams look at analytics after publication to measure performance. But the real power is using content analysis before and during creation — letting AI inform what you write, how you headline it, and how you structure your content calendar. Claude + WordPress MCP makes this practical for any WordPress site.
📌 First, know what the AI can actually see. MCP only reads data that’s exposed to the connected tools. Your content, categories, tags, comments, and publish dates live in WordPress, so Claude can analyze those directly. But pageviews, time-on-page, scroll depth, and traffic sources usually live elsewhere — Parse.ly, Jetpack Stats, or Google Analytics — and Claude cannot see them unless you connect or export those sources first. So treat the exercises below as content and structure analysis, not traffic analytics, until you’ve wired in your stats source.
Exercise 1: Content Gap Analysis
Ask Claude to analyze your content portfolio via MCP:
Look at our published content from the last 30 days.
Analyze it by category and topic:
1. Which categories have the most content? Which have the least?
2. Are there topic areas where we only have 1-2 articles but could
clearly write more? (Look for posts that mention related subtopics
we haven't covered independently.)
3. Which categories have the most comments? (Comments live in WordPress,
so you can read them. For pageviews/traffic, I'll need to connect
our analytics source separately.)
4. Are there seasonal or timely topics we should be covering now?
Give me a "content opportunity matrix" — topics where our coverage is
low but the subject clearly deserves more, based on what's in our archive.Exercise 2: Headline Performance Patterns
Look at our 50 most recent articles.
Analyze their headlines and find patterns:
- What headline structures do we use most? (questions, numbers,
how-tos, names, etc.)
- What's the average headline length?
- Which headlines could be improved based on best practices?
- Identify 10 articles that could benefit from headline A/B testing
Then use the WordPress AI plugin to generate 3 alternative headlines
for each of those 10 articles. Present originals vs. AI suggestions
in a table.Exercise 3: The AI Pre-Publish Checklist
Before publishing your next article, ask Claude:
I'm about to publish an article titled "[TITLE]" in the [CATEGORY] section.
1. What similar articles have we already published? Could this be
a duplicate or should it link to existing content?
2. Suggest 3-5 internal links to add (articles our readers would
logically want to read next)
3. Generate 3 alternative headlines using the AI plugin
4. Check: does this post have a featured image, excerpt, and proper
category/tags?
5. Generate alt text for any images that are missing it💡 Tip: These prompts work with any WordPress site connected via MCP. The WordPress AI plugin adds title and excerpt generation directly in the editor, while Claude via MCP handles the analytical heavy lifting across your entire content library.
✅ Key Takeaway: Claude + MCP can read everything stored in WordPress — content, categories, tags, comments, metadata — and surface structural patterns, gaps, and opportunities in minutes. For real traffic and audience metrics, connect your analytics source (Parse.ly, Jetpack Stats, GA) so the AI is reasoning over actual numbers, not assumptions.
← Day 12: Content Blocks with Telex · Day 14: Week 2 Review →