ai-tools

    How to Analyze Your Existing Content for Course Material Using Claude

    Upload blog posts, transcripts, and workshop notes to Claude and let it identify what's already course-ready — so you build from strength, not from scratch.

    Abe Crystal, PhD9 min readUpdated September 2026

    You probably have more course material than you think. If you've been blogging, podcasting, running workshops, or coaching clients for a few years, you've already created the raw ingredients for a course — they're just scattered across dozens of files, posts, and recordings. Claude can help you see the course that's already there.

    Quick Answer: How to Analyze Your Existing Content for Course Material Using Claude

    1. Gather your content: Execute this step in your course creation workflow.
    2. Upload to Claude: Execute this step in your course creation workflow.
    3. Run content audit: Execute this step in your course creation workflow.
    4. Identify course-ready material: Execute this step in your course creation workflow.
    5. Map gaps: Execute this step in your course creation workflow.
    6. Plan new content: Execute this step in your course creation workflow.
    1–2 hoursClaude (free or Pro)You have existing content (blog posts, transcripts, notes)
    1Gather your content
    2Upload to Claude
    3Run content audit
    4Identify course-ready material
    5Map gaps
    6Plan new content

    What you’ll walk away with:

    • An inventory of your existing content organized by teaching potential
    • A clear picture of what's already course-ready and what needs transformation
    • Gaps in your content library that require new material
    • A prioritized plan for creating or adapting content

    Why Claude for content analysis

    The biggest challenge for experienced practitioners isn't creating content — it's seeing the structure in content they've already made. You wrote that blog post about client boundaries two years ago. You answered a similar question on a podcast six months later. You ran a workshop exercise on the same theme last spring. Each piece is good on its own, but you've never looked at them as a connected sequence.

    Claude's advantage here is its context window. You can upload 50 blog posts, three podcast transcripts, and a set of workshop notes into a single conversation — roughly 150,000 words of your own material — and ask questions across all of it. "Which topics do I cover most often?" "Where do I contradict myself?" "What do I explain well in the blog but skip in the workshops?"

    That kind of cross-content analysis is nearly impossible to do manually. You'd need to reread everything, take notes, and map connections. Claude does the initial pass in about thirty seconds.

    What to upload

    Not all content is equally useful for course planning. Here's what I'd prioritize:

    • Blog posts and articles. These are usually your most structured thinking. Upload the ones related to your course topic — even posts you think are too basic or too niche. Claude will find connections you missed.
    • Podcast transcripts or episode notes. Transcripts are gold because they capture how you actually explain things — more naturally and with more examples than your writing. If you don't have transcripts, even detailed show notes work.
    • Workshop or webinar materials. Slides, handouts, exercises, facilitator notes. These show how you teach, not just what you know. The exercises especially reveal your pedagogical instincts.
    • Client emails and Q&A. The questions your clients ask repeatedly point directly to what your course needs to cover. Strip out personal details before uploading.
    • Social media threads. If you've written thoughtful LinkedIn posts or Twitter threads on your topic, those often contain your most concise, quotable explanations.

    Don't worry about uploading too much. Claude handles large volumes well — that's precisely its strength for this task.

    The analysis process

    I'd suggest working through this in three passes rather than asking Claude to do everything at once. Each pass builds on the previous one, and you get to check the AI's work at each stage.

    Pass 1: Theme identification

    Start by asking Claude what themes it finds across your content. Don't tell it what you think the themes are — let it discover them. You'll often be surprised. One course creator I've worked with discovered that she wrote far more about the emotional relationship with food than about meal plans, even though she thought of herself primarily as a meal-planning expert. That insight reshaped her entire course premise.

    Pass 2: Gap analysis

    Once you see the themes, ask Claude where the gaps are. "Based on everything I uploaded, what topics do I reference but never fully explain?" and "What would a student need to know that I haven't covered anywhere in this material?" These gaps are your new content — the pieces you'll need to create from scratch for the course.

    Pass 3: Sequencing

    Finally, ask Claude to suggest a teaching sequence. "If these themes became course modules, what order would help a beginner build understanding step by step?" This is where Claude's analysis is most helpful and most in need of your judgment — the logical sequence isn't always the best learning sequence.

    Prompts to try

    Upload your content files first, then use these prompts. Replace bracketed text with your specifics.

    Prompt 1 — Content audit:

    "I've uploaded [number] pieces of my own content — blog posts, transcripts, and notes — all related to [topic]. Analyze this material and identify: (1) the 5-7 recurring themes I return to most often, (2) the concepts I explain particularly well with clear examples, and (3) any topics I mention but never fully develop. For each theme, cite specific pieces where I cover it."

    Prompt 2 — Course-readiness assessment:

    "Based on this content, assess what's already 'course-ready' versus what needs more development. Course-ready means: I've explained the concept clearly, included examples or exercises, and a student could learn from this with minimal editing. What needs development means: I've touched on the topic but it needs deeper explanation, more examples, or structured exercises. Organize your assessment by theme."

    Prompt 3 — Module mapping:

    "Propose a 6-8 module course structure based on this content. For each module, specify: (a) the module topic and learning outcome, (b) which uploaded pieces provide the foundation for this module, (c) what new content I'd need to create to fill gaps, and (d) one discussion question or exercise that would help students apply the concept. Sequence the modules for a [beginner/intermediate] audience."

    The human layer

    Here's what I want to be direct about: Claude is analyzing your words, not your teaching. It can find patterns in your writing and identify themes, but it can't assess which of your explanations actually help people learn versus which ones just sound clear on paper.

    You know the difference. You know that your blog post about boundaries landed with every client who read it, while your workshop handout on the same topic always needed verbal explanation. That knowledge — what works in practice versus what looks good in text — is something Claude can't detect.

    Use Claude's analysis to see your body of work from above, then apply your teaching experience to decide what belongs in the course and what doesn't. Some of your best content might need to be rewritten for a course format. Some of your roughest notes might contain the most important insights. Only you know which is which.

    What it gets wrong

    It overvalues frequency.

    Claude identifies "recurring themes" by counting how often you mention something. But your most important insight might be something you wrote about once, deeply, rather than something you've referenced twenty times. Don't let frequency be the only measure of importance.

    It misses your pedagogical choices.

    If you deliberately leave certain topics for later in your client work — building up to complex ideas gradually — Claude may flag those as "gaps" rather than recognizing them as intentional sequencing. You know your teaching logic; the AI doesn't.

    It flattens voice.

    When Claude summarizes your themes, the summaries sound like Claude, not like you. Your blog post might say "I've watched clients sabotage their own progress because they mistake busy for productive." Claude's summary might say "A recurring theme is the distinction between productivity and busyness." Both are accurate. Only one sounds like you. Build your course from your original language, not the AI's summaries.

    Context window isn't infinite.

    While Claude handles large volumes well, uploading hundreds of thousands of words can still degrade quality at the edges. If you have an enormous content library, batch it by topic area rather than dumping everything in at once. Quality of analysis matters more than comprehensiveness. For grounded analysis of your own uploads, NotebookLM is also worth trying — it restricts answers to your source material.

    From analysis to course

    Once Claude has helped you see the course hiding in your content, the next step is building it. You've got your themes, your sequence, and a clear sense of what's ready versus what needs work.

    Ruzuku makes the transition from plan to course straightforward — create your modules, drop in the content that's already course-ready, and note where you need to create new material. You can start building with a free account and invite students as soon as your first module is solid.

    Related guides

    Topics:
    claude
    content analysis
    course planning
    ai tools
    repurposing content
    blog to course
    transcript analysis
    existing content

    Frequently Asked Questions

    How much content do I need before this approach works?

    You can get useful results with as few as 5-10 blog posts or a couple of workshop transcripts. The more you upload, the better Claude can identify patterns and recurring themes. But even a small body of work contains teachable structure — you just may not have seen it yet because it's scattered across different formats and dates.

    Why Claude instead of ChatGPT for content analysis?

    Claude's main advantage is its context window — it can hold roughly 200,000 tokens (about 150,000 words) in a single conversation. That means you can upload dozens of blog posts or an entire book manuscript at once and ask questions across all of it. ChatGPT has a smaller effective context window, so you'd need to analyze content in smaller batches.

    Will Claude plagiarize my content when creating the course outline?

    No. Claude analyzes your content to identify themes, gaps, and structure — it doesn't republish your material. The output is a structural analysis (which topics recur, what's missing, how to sequence things) that you then use to build your course in your own voice. Your original content remains yours.

    Related Articles

    ai-tools

    How to Analyze Competitor Courses Using ChatGPT

    Use ChatGPT to systematically review competitor sales pages, curricula, and student reviews — then find the gaps your course can fill.

    Read more
    ai-tools

    How to Research What Your Audience Wants Using ChatGPT

    Use ChatGPT to analyze Reddit threads, book reviews, and forum posts to uncover what your audience actually needs before you build your course.

    Read more
    ai-tools

    How to Create a Course FAQ Using ChatGPT

    Use ChatGPT to anticipate student objections and questions before they ask them. Build a FAQ that reduces support load and increases enrollment.

    Read more

    Ready to Build Your Course?

    AI handles the first draft. You bring the expertise. Start free on Ruzuku — unlimited courses, zero transaction fees.

    No credit card required · 0% transaction fees