Discovery & Recommendations
Narratorr’s Discover feature analyzes your library and suggests new audiobooks you might enjoy. It looks at the authors, series, genres, and narrators you already own and finds related titles.
Configuring Discovery
Section titled “Configuring Discovery”Discovery is enabled by default — Narratorr generates suggestions on a configurable schedule out of the box. To turn it off, or to adjust its interval and limits, go to Settings > General (Discovery section).
| Setting | Description |
|---|---|
| Enabled | Master toggle for the discovery engine |
| Interval | How often (in hours) to refresh suggestions. Range: 1–168. Default: 24. |
| Max Per Author | Maximum suggestions generated per author. Range: 1–50. Default: 5. Prevents a single prolific author from dominating the feed. |
| Expiry Days | Suggestions expire after this many days if not acted on. Range: 1+. Default: 90. |
How Scoring Works
Section titled “How Scoring Works”Each suggestion gets a score based on why it was recommended:
| Reason | What It Means |
|---|---|
| Author | You own other books by the same author |
| Series | Completes a series you’ve started (see below) |
| Genre | Matches genres in your library |
| Narrator | Narrated by someone you’ve listened to before |
| Diversity | Introduces variety — prevents a filter bubble by surfacing titles outside your usual patterns |
Each reason has a base weight:
| Reason | Base Weight |
|---|---|
| Series | 50 |
| Author | 40 |
| Genre | 25 |
| Narrator | 20 |
| Diversity | 15 |
Scores get bonus points for recency (+10 if published in the last 2 years), duration match (+5 if close to your library’s median), and series position (+20 if it’s the exact next book you need). Final scores are clamped to 0–100.
Weight Multipliers
Section titled “Weight Multipliers”Each reason category has a weight multiplier (0–1), but these are not user-configurable — Narratorr computes them automatically from your dismissal patterns on each refresh cycle.
If you consistently dismiss genre-based suggestions (more than 80% dismissed with at least 5 suggestions), that category’s multiplier is automatically reduced — down to a minimum of 0.25. This prevents the engine from repeatedly suggesting things you’re not interested in.
Series Completion Intelligence
Section titled “Series Completion Intelligence”When you own books 1, 2, and 4 of a series, Narratorr notices the gap and suggests book 3 with a boosted score. It also suggests the next book in a series you’re actively following.
Series suggestions are flagged with a series reason and include context like “Book 3 of The Expanse — you own 1, 2, 4.”
The Discover Page
Section titled “The Discover Page”The Discover page shows your suggestion feed, sorted by score. Each card shows:
- Cover art, title, and author
- Why it was suggested (reason badge)
- The score
Actions
Section titled “Actions”| Action | What It Does |
|---|---|
| Add | Opens a popover that adds the book to your library as “wanted”, with an optional Search immediately checkbox (defaults to your Quality > Search Immediately setting) |
| Dismiss | Removes the suggestion. Feeds into the dismissal tracking system. |
Diversity Factor
Section titled “Diversity Factor”To prevent filter bubbles — where you only see more of exactly what you already have — the engine includes a diversity signal. A portion of suggestions intentionally come from outside your dominant patterns. This is controlled by the diversity weight multiplier.
Suggestion Lifecycle
Section titled “Suggestion Lifecycle”- Generated — engine creates suggestion with a score and reason
- Pending — visible in the Discover feed
- Added — you chose to add the book to your library
- Dismissed — you dismissed it (tracked for weight tuning)
- Expired — automatically removed after the expiry period
Only pending, added, and dismissed are stored status values. “Generated” (the suggestion being created) and “Expired” (it being auto-removed) are lifecycle transitions, not statuses a suggestion sits in.
Suggestions are refreshed each cycle. If a suggestion already exists (matched by ASIN), its score is updated and its refresh timestamp is reset. Expiry, however, is measured from the original creation date — not the refresh timestamp — so a suggestion that keeps reappearing will still expire the configured number of days after it was first generated.