How It Works

Audible Performance Rating vs Story Rating: What Each Star Means

Audible shows Overall, Performance and Story stars. Here is what each measures, how counts work, and how to use the performance rating to pick a recording.

On Audible, Overall is the listener's judgment of the audiobook as a whole, Performance is the narration and the recording, and Story is the book underneath. A listener can give the same title five stars for Story and fewer for Performance. Audible's review form asks for separate stars for "Overall, Performance, and Story", per its Manage your reviews help page.

What each of the three stars means

Audible's form names the three ratings. Cameron Knott's 2022 guide to reviewing Audible books supplies the plain-language versions. He describes Overall as whether the listener would recommend or buy the audiobook, Story as the quality of the underlying material, and Performance as the listening experience, including the recording quality and how the narrator handles the voices.

In ordinary use, then, the three stars answer three different questions. Overall asks whether the audiobook was worth it. Story asks whether the book itself is good. Performance asks whether the person reading it, and the way it was recorded, made it a pleasure or a chore to listen to. That reading of the labels comes from Knott, not from an Audible rulebook, so treat it as a sensible way to read the numbers rather than an official definition.

The split matters because a great book can be badly served by its recording, and an ordinary book can be lifted by a strong narrator. Two stars that disagree tell you more than one star that blends them.

Three clear glass jars on a wooden shelf, each filled with a different amount of smooth pebbles
Audible splits its stars three ways, and each number answers a different question.

Who can leave a rating

Audible says: "To leave a review or rate a title, you must have listened to at least 20% of it." That comes from the same help page, and Audible's UK version of it carries the same guidance. The page talks about listening, not buying, so the requirement it states is time spent with the title.

So, a rating does not have to come from someone who finished the book. A listener who is 20% in can rate it, which is worth remembering when you read a verdict on the ending.

A comfortable armchair with a pair of headphones draped over one arm and a closed plain book resting on the seat
Who can leave a rating shapes what the average can tell you.

What the stars attach to

Ratings sit on one audiobook listing. That means one recording, by one narrator or one cast, from one publisher. They are not a score for the book in every format, and they are not a score for every recording of it.

Here is a real listing as an example. On October 1, 2026, the Audible US page for Andy Weir's Project Hail Mary, the Audible Studios recording narrated by Ray Porter, showed these figures:

  • Overall: 5 out of 5 from 197,704 ratings
  • Performance: 5 out of 5 from 180,944 ratings
  • Story: 5 out of 5 from 180,391 ratings

Those are storefront figures on that date, not a standing total. They will move.

Look at the counts as well as the stars. The same listing shows 197,704 Overall ratings, 180,944 Performance ratings and 180,391 Story ratings, so each line is its own pool. Read the star and the count together, and treat them as what that storefront shows today rather than a number you could recompute yourself.

Two pairs of headphones on a wooden table, one hanging on a stand and one lying flat, with a small blank card between them
When two recordings compete, the Performance score is the number that separates them.

How to use the three numbers to choose a recording

Different decisions lean on different stars.

  1. Two recordings of the same book. Compare the Performance stars and their counts. Story describes the same book either way, so it will barely separate them. Performance is where a full-cast production and a single narrator, or an older recording and a newer one, will pull apart.
  2. Two different books. Story and Overall carry more of the signal. Performance tells you about the delivery, which matters less when you are still deciding what to read.
  3. A big gap on one listing. If Story sits well above Performance, listeners are reacting to the narration more than the book. If Performance sits well above Story, the reverse is likely true: the recording is working harder than the material.
  4. A tiny count. A handful of ratings is a thin read. Lean on the written reviews and the audio sample, and give the stars less weight.

Where the evidence is limited, it helps to say what it covers. A 2017 analysis by NYC Data Science, titled Drama vs. Words and dated February 20, 2017, looked at roughly 150,000 audiobooks. It reported that many titles scored similarly across the three categories, and it saw variation associated with narrators. That describes a pattern in one dataset. It does not show Audible ratings are reliable, and it does not show they are unreliable.

Narrators do not have one Audible score

Performance belongs to a listing, not to a person. Audible shows it per title, so a narrator has no single Audible rating that follows them from book to book.

Some third parties build one anyway. The independent site Finelyspun, in its about page, says it ranks narrators by averaging each narrator's title-level Performance ratings, weighted by rating count. That is Finelyspun's roll-up, not an Audible figure. If you follow a narrator, look at the Performance stars on the specific title in front of you.

Each Audible store keeps its own ratings

Regional stores keep separate pools. On October 1, 2026, the Audible UK listing for the same Ray Porter recording of Project Hail Mary showed Overall 4.9 from 44,024 ratings, with 41,113 Performance ratings and 41,056 Story ratings, according to Audible UK's search results. The US listing showed 197,704 Overall ratings on the same date.

A second example is Trevor Noah's Born a Crime, narrated by Trevor Noah and published by Random House Audio. On the same date, Audible US's chart page showed Overall 4.8 from 4,290 ratings. Audible UK's Trevor Noah listings showed Overall 4.9 from 15,333 ratings.

The practical step is simple. Check the store you buy from, and never add the two pools together. A 4.8 from one store and a 4.9 from another are separate readings, not one combined score.

A brass balance scale with a plain book on one pan and a pair of earbuds on the other, the book side sitting slightly lower
Stars are a signal, not a verdict, so weigh them with the sample and the credit.

How far to trust the stars

Audible publishes Customer Review Guidelines. They say reviews that break the guidelines can be removed, including compensated reviews, undisclosed conflicts of interest, and reviews posted by an author, narrator or publisher posing as an unbiased listener. Any customer can report a review, and Audible decides whether to keep it. A submitted review can also take two business days to appear, per the help page.

On how high the numbers run, one named comparison is the 2024 paper "Measuring Literary Quality" by Aleksandar Ristovski and colleagues, in the conference reader. In its corpus, Audible averages came out higher than Goodreads averages, and the authors suggested the smaller Audible rating population might contribute. That is a finding about that corpus, so it supports a modest habit: expect Audible stars to sit toward the top of the scale, and read the differences between listings rather than the absolute number.

Ratings on other stores work differently

Other places to check a recording do not use Audible's three-part split.

  • Apple Books uses one-to-five stars and written reviews, and Apple's support guide limits rating to customers who purchased or redeemed a code. There is no separate narration or story score.
  • Goodreads uses one general star rating per edition, according to its help article on reviewing multiple editions.
  • Libro.fm and StoryGraph: use the written reviews on the listing for the recording you are considering, and look for comments about the narration in the review text.

Questions listeners ask

Does Audible tell you the percentage of books you read?

The one percentage Audible's help page gives is for rating: you must have listened to at least 20% of a title to rate or review it. For how far along you are in a given title, look at your own library and listening history in the Audible app.

What are the downsides of using Audible's ratings?

Each store keeps its own pool, so US and UK figures differ. Newer titles can have small counts that make the stars a thin read. Reviews are moderated under Audible's guidelines, and the displayed average is best read as the storefront's current figure, since Audible does not publish a formula for it. So, compare listings inside one store.

Can I rate an audiobook without finishing it?

Yes. Audible says you must have listened to at least 20% of a title to rate or review it, so you can rate before you reach the end.

Is Performance the narrator's score for all their books?

No. Performance is attached to a single audiobook listing. A narrator who has recorded fifty books has fifty separate Performance scores on Audible. Finelyspun is one example of a third party that rolls them into a single narrator ranking.

Where to go next

If you want to see how one specific recording sounds in practice, my Project Hail Mary audiobook review walks through the Ray Porter recording, and it is a good place to see the Performance question applied to a real title.

By Cal Hewitt. Published .

Back to all Notes on how the audiobook market works