Last Updated: April 15, 2026
Status: Flagged for batch UX pass
Issue: Serif font looks great on landing page but becomes difficult to read at smaller sizes in admin screens.
Action: Audit all typography hierarchy; consider switching to sans-serif for dense admin UI while keeping serif for marketing/public-facing pages.
Priority: Medium (impacts daily admin workflow)
Status: Phase 2 feature
Request: Process books in chunks (e.g., chapters 1-2 now, chapters 3-4 later) rather than all-or-nothing.
Context: Users may want to process a few chapters, review quality, then do more.
Implementation:
- Add
startChapterandendChapterparams to pipeline job - Modify segmentation logic to extract only specified range
- Auto-calculate maxChars based on chapter range
Priority: High
Complexity: Medium (requires DB schema extension, pipeline refactor)
Status: Phase 2 feature
Request: Instead of dropdown with fixed options, allow users to specify exact chapter ranges (or entire book).
Current: Limited to predefined maxChars options
Desired: "Process chapters 5-12" or "Process entire book" or "Preview first 3 chapters"
Implementation:
- Fetch chapter metadata from Gutenberg metadata API or parse from book text
- Build UI to show chapter list + checkboxes or range selector
- Pass chapter indices to pipeline
Priority: High
Complexity: Medium–High (requires book structure analysis)
Status: Phase 2+ feature
Request: Export user highlights, notes, and commentary to PKM systems.
Use Case: Readers want their study notes to flow into their personal knowledge base.
Possible Integrations:
- Notion: API-based export (highlights as database, commentary as linked pages)
- Obsidian: Markdown export with wikilinks to episodes
- Logseq: Outline format with backlinks
- Apple Notes: iCloud sync
- Generic: Markdown dump that user can import anywhere
Implementation Approach:
- Add
/api/export/notion,/api/export/obsidian, etc. - Format user's notes + highlights + AI commentary as destination format
- Handle auth (Notion token, Obsidian vault path, etc.)
- Provide one-click "Export to Notion" in UI
Priority: Medium (nice-to-have, differentiator)
Complexity: Medium (requires learning each platform's API)
Value: Increases stickiness; turns Strata into part of user's workflow
- Strata uses ElevenLabs TTS (single voice,
eleven_multilingual_v2) - Narrates at ~30–50 words per minute depending on voice
- Quality is good but sounds synthetic
- Professional audiobook-grade narration (likely human-quality)
- Character voices (different speakers per character in dialogue)
- Pacing & inflection tuned for literary content
- Premium pricing (presumably higher cost than TTS API)
Pros:
- Dramatically better audio quality (human-like narration)
- Industry-standard for publishing
- Differentiates from generic TTS
- Makes audiobooks sellable on other platforms
Cons:
- Higher API costs
- May not support per-episode generation (might require full book at once)
- Less control over voice selection
- Vendor lock-in to ElevenLabs
Recommendation: Good if you want premium distribution-quality audiobooks.
Strata for Free Users: Fast TTS narration (current)
Strata Pro / ElevenLabs Audiobook: Premium narration via ElevenLabs Generator
Then Distribute: Export both versions to ElevenReader marketplace
Pros:
- Scalable (free users get quick preview, paid users get pro narration)
- Monetization layer (charge for premium narration)
- Flexible (user chooses quality vs. speed tradeoff)
Cons:
- More complex pipeline
- Higher support burden
Recommendation: Best if you want to offer tiers.
- Process books in Strata (commentary, segmentation, themes)
- Narrate with ElevenLabs Audiobook Generator (professional quality)
- Publish to:
- ElevenReader marketplace (60% revenue)
- Apple Books, Audible, Scribd (through aggregators like Draft2Digital)
- Your own Strata app (subscription model)
- Own the metadata & relationship — you control the experience
Pros:
- Maximize revenue streams (SaaS + marketplace + wholesale)
- Use ElevenLabs as a service tier, not the main product
- Leverage your annotation work across multiple channels
- Could become a full publishing operation
Cons:
- More complex (requires aggregator integrations)
- Cash flow timing (payments take 90+ days via aggregators)
Recommendation: Long-term play for scale.
Replace the current TTS with ElevenLabs Audiobook Generator as Phase 2.
Here's why:
- Your current TTS is fine for MVP but won't compete with real audiobooks
- ElevenLabs Audiobook Generator is better because that's what it's designed for
- The better the audio, the more you can charge/distribute
- You're not building an audio engine; you're building a learning platform
Don't try to be better at narration than ElevenLabs. Use their engine. Your edge is:
- Episode structure + commentary
- Highlights + annotation
- Learning journey (not just listening)
- Community features (chat, notes)
Suggested Roadmap:
- ✅ Now: Ship MVP with current TTS (it works)
- Phase 1: Chapter selection, partial processing
- Phase 2: Switch to ElevenLabs Audiobook Generator for better quality
- Phase 2: ElevenReader marketplace export
- Phase 2–3: PKM integrations (Notion export)
- Phase 3+: Direct distribution (Apple Books, Audible via aggregators)
The key insight: You're not in the narration business. You're in the literary curation + learning design business. ElevenLabs should handle audio.
ElevenLabs recently launched ElevenReader — an audiobook marketplace where creators can:
- Upload audiobooks
- List them on ElevenReader marketplace
- Earn 60% of revenue
- Leverage ElevenLabs TTS infrastructure
Question: How does Strata differentiate, and can we integrate?
| Aspect | ElevenReader | Strata |
|---|---|---|
| Core Function | Audiobook distribution & hosting | Content segmentation + guided literary experience |
| Content Processing | Upload pre-existing audio | Cut books into thematic episodes w/ commentary |
| Context Layers | Just audio | Episodes + notes + highlights + AI commentary + quote cards + chat |
| User Experience | Listen & navigate | Structured learning journey with interleaved study aids |
| Monetization | 60% revenue share | Direct subscription to readers (+ marketplace option) |
| Customization | Publisher decides structure | Strata algorithmically segments + AI annotates |
| Target Audience | General audiobook listeners | Deep readers, learners, Great Books fans, Catholics |
| Workflow | Upload final product | Source text → pipeline → annotated episodes → publish |
- Structured Learning — Episodes aren't arbitrary; each has commentary, themes, character arcs
- AI-Powered Annotation — Chat, quote extraction, thematic abridgments aren't available elsewhere
- Niche Focus — Religious, classics, Great Books positioning (not generic audiobooks)
- Direct Relationship — You own the reader experience; ElevenReader is just distribution
- Data Leverage — User highlights, notes, chat give you insight into what resonates
- Flexible Monetization — Subscription app + selling individual titles to libraries + ElevenReader marketplace
- User processes a book in Strata (e.g., Brothers Karamazov)
- Strata generates annotated episodes with narration
- Export to ElevenReader: Package as EPUBs/PDFs with audio links + metadata
- List on ElevenReader marketplace
- Earn 60% from listeners who discover via ElevenReader
- Reach: ElevenReader brings discovery (you handle production)
- Revenue: 60% is pure upside on top of Strata subscription model
- No Conflict: You're not competing with ElevenLabs; you're using their distribution
- Differentiation: Your commentary + curation makes your books better than generic TTS audiobooks
- Phase 1: Ensure Strata pipeline exports clean EPUB + audio files
- Phase 2: Build ElevenReader export connector (package episode metadata + links)
- Phase 3: Wire ElevenReader earnings dashboard into Strata admin
- Phase 4: Marketing: "Books on Strata + ElevenReader marketplace"
Keep building Strata as the primary product. Your core USP is the guided learning experience, not just audiobooks. ElevenReader becomes a distribution channel, not a competitor.
| Channel | Content Type | Model | Monetization | Primary Use |
|---|---|---|---|---|
| Strata App | Audio + text + learning layer | Freemium subscription | Direct (free trial → $X/mo) | Core product |
| ElevenReader | Audio + metadata | Free listing | 60% revenue split | Discovery + audiobook market |
| YouTube | Audio + transcript overlay | Free content | Ad-supported or sponsorship | Discovery + marketing |
| Podcast | Curated audio commentary | Free feed | Ad-supported or sponsorship | Discovery + audio format |
| Substack | Weekly written commentary | Free newsletter | Email list building | Top-of-funnel marketing |
| Beehiiv | Weekly written commentary | Free newsletter | Email list building + potential affiliate | Top-of-funnel marketing |
| PKM Exports | Written commentary + notes | API/integration | Value-add to Strata Pro | Workflow integration |
Substack could work, but only if positioned correctly:
- Another paywall (fractures audience)
- Competes with Strata Pro (confusion)
- Audio/video are secondary on Substack (not ideal)
- Don't do this — you'd be splitting revenue and attention
- Free weekly "commentary digest" (e.g., top 3 quotes + themes from the book)
- Drives discovery to Strata app
- Builds email list (owned asset)
- Better than podcast/YouTube for written content + email-native audience
- Low effort (repurpose commentary already generated)
Priority 1: Strata App (core product, direct revenue)
Priority 2: Podcast (free, audio-first, easy distribution)
Priority 2: YouTube (free, algorithmic discovery, ads/sponsorship)
Priority 3: Email Newsletter (free Substack digest, drives to Strata)
Priority 3: ElevenReader (60% splits, passive revenue)
Priority 4: PKM Exports (Notion/Obsidian, value-add)
- Audience fragmentation: You'd have Strata subscribers + Substack subscribers (confusing)
- Substack doesn't do audio well (their platform is text-first, not audio-first)
- Substack takes 10% of revenue (vs. 0% on your own app)
- Email is a weak substitution for app (notification, features, learning structure)
- Owned email list (Substack makes it easy, but it's your data)
- Low-effort repurposing (commentary already exists, just digest it)
- Top-of-funnel marketing (free weekly → "Want full guided experience? Try Strata Pro")
- Complements podcast/YouTube (different format, email-native audience)
- Drives traffic to Strata (not competing for subscription revenue)
Why run both?
- Different audiences: Some readers prefer Substack ecosystem, others prefer Beehiiv
- Same content: Single source (your commentary digest) → auto-publish to both platforms
- Redundancy: If one platform has issues, you still have reach on the other
- SEO/discovery: Beehiiv has built-in sponsor marketplace (potential affiliate revenue from book recommendations)
- Low marginal cost: Once you digest the commentary, distributing to both is minimal extra effort
Implementation:
- Write once (weekly digest)
- Publish to Substack and Beehiiv simultaneously
- Include CTA in both: "For full guided experience, listen on Strata"
- Track clicks/conversions from each to see which drives more Strata signups
Note: Both should remain FREE (no paywall comparison between platforms).
Later (Phase 3+): Launch free weekly Substack digest that mirrors your podcast/YouTube to capture email-native audience. Subject line: "Top 3 insights from [Book Chapter]" → links to full guided experience on Strata.
Now: Focus on podcast + YouTube + Strata app.
If free channels (ElevenReader, YouTube, podcast) have poor audio quality, listeners won't even consider Strata. Audio quality is your acquisition hook, not your paywall.
Conversely, if they experience premium narration for free, they want to share it — organic growth.
| Channel | Audio Quality | Added Value | Model |
|---|---|---|---|
| ElevenReader Marketplace | Premium (Audiobook Generator) | None | Free / 60% splits |
| YouTube | Premium (Audiobook Generator) | Transcript overlay | Free / Ad-supported |
| Podcast | Premium (Audiobook Generator) | Curated episode selection | Free / Ad-supported |
| Strata App (Free Trial) | Premium (same) | Audio + basic features | 2-week free trial |
| Strata App (Pro Subscription) | Premium (same) | + Guided commentary + highlights + chat + notes + structure | Paid |
The paywall is NOT audio quality. It's the learning layer.
Free users get: "Wow, this audiobook is beautifully narrated."
Paid users get: "Wow, AND I understand it deeply with guided commentary and community."
- Everywhere (free + paid): Premium narration via ElevenLabs Audiobook Generator (or professional narrator)
- NOT a paywall lever
- Differentiation: Your commentary + structure + chat + annotations = the paid experience
- Discover → ElevenReader / YouTube / podcast (premium audio, no paywall)
- Download Strata → Try free trial (same premium audio)
- Hit study feature → "Want guided commentary?" (paywall)
- Subscribe → Unlock full learning experience (same audio + structure)
- Word-of-mouth: "This audiobook is incredibly well narrated" (true, no TTS artifacts)
- No friction: Audio is the same paid/free
- Conversion happens on learning value (commentary, structure, community), not on audio gatekeeping
- Scalable: You use premium narration everywhere; differentiation is curation + learning design
- Competitive: Your position is "Great audio + Great learning", not "OK audio locked behind paywall"
- ElevenReader is a commodity marketplace — success depends on content quality & marketing
- Strata is a unique platform — no competitor offers structured learning + AI annotation for classics
- Strata subscription captures direct value — ElevenReader is gravy on top
- You own the brand & data — readers getting smarter is your asset; ElevenReader just helps distribute
✅ Now (MVP): Perfect the Strata experience on Brothers Karamazov
✅ Next (Phase 1): Add chapter selection, partial processing
📋 Later (Phase 2): ElevenReader export + marketplace listing
📋 Eventually: Native Strata publishing (books → readers directly without ElevenReader)
Status: Phase 2+ feature
Request: Add a preamble to each chapter (audio or text) directing listeners to deeper experience.
Example: "You're listening to Chapter 5 of Brothers Karamazov on [Platform]. For guided commentary, notes, and deep discussion, listen with Strata.com or subscribe to our podcast/YouTube channel."
Implementation:
- Add optional
introMessagefield to episodes table - Template: "Listen with Strata at [link] | Subscribe to podcast: [link] | Watch on YouTube: [link]"
- When exporting to ElevenLabs Audiobook Generator, prepend intro to episode narration
- Also export in text formats (EPUB, Markdown)
Priority: Medium (good for conversion but not blocking)
Complexity: Low (just text field + template logic)
Value: Cross-promotes your other channels; drives conversion from audiobooks → Strata subscription
Status: Architecture decision for Phase 2
Current: Pipeline does everything end-to-end (segment → commentary → abridgments → audio)
Proposed: Two-step workflow
Step 1: Text-Only Export
- Generate commentary, abridgments, highlights, quote cards
- Export as: JSON metadata, Markdown, EPUB, plain text
- Output: Ready for external narration services (ElevenLabs Generator, human narrators, etc.)
- Use cases:
- Distribute commentary separately from audio
- Archive text layers without audio
- Pass to ElevenLabs Audiobook Generator (bring-your-own-narration)
- Publish as ebooks independently
Step 2: Audio Generation (Optional Later)
- Still in pipeline but now decoupled
- Can apply to any text export
- User chooses: use ElevenLabs TTS, swap in Audiobook Generator, or skip entirely
Admin Portal Flow:
- Upload/fetch book
- Trigger: "Generate Text Only" → commentary, abridgments, notes
- Review & export as JSON/EPUB/Markdown
- (Optional later) Trigger: "Generate Audio" → choose narrator (TTS / Audiobook Generator / other)
Benefits:
- You own the text layers (commentary, structure, curation)
- Audio is now a separately monetizable/distributable asset
- Can use different narrators for different channels
- Supports multiple export formats for PKM/publishing
- Faster initial processing (no waiting for audio)
Database Changes:
- Add
textProcessingStepstatus topipelineJobs - Add
audioProcessingStepstatus (separate) - Export formats stored in S3 (JSON, EPUB, Markdown alongside MOBI/PDF)
Implementation Complexity: Medium–High (requires pipeline refactor)
Priority: High (unlocks ElevenLabs Audiobook Generator integration + PKM exports)
- Typography audit: serif → sans-serif for admin, keep serif for marketing
- Add
startChapter/endChapterto pipeline jobs schema (migration) - Build chapter selection UI for pipeline trigger
- Research ElevenReader API/export requirements
- Investigate Gutenberg chapter metadata API
- Spec out partial processing logic
- Design text-only export workflow + formats (JSON, EPUB, Markdown)
- Refactor pipeline: decouple text generation from audio generation
- Add intro/discovery messaging template to episodes