Have you ever noticed how the same books keep appearing everywhere?
On Amazon. On TikTok. On Goodreads. On YouTube.
Different platforms. Different creators. Same titles.
It starts to feel like discovery… but something about it feels off.
How much of what we read is actually chosen by us—and how much is shaped by algorithms?
The quiet influence of recommendation systems
Most of us assume that book recommendations reflect our taste.
You click on something you like, and the system responds. It feels personal. Almost intuitive.
But research suggests something more complex is happening.
A recent paper on recommender systems highlights that these systems “influence the types of narratives and ideas to which users are exposed” — meaning they don’t just reflect preferences, they actively shape them. (Source: Recommender Systems and Narrative Exposure, arXiv)
In other words, algorithms don’t simply help you find books. They quietly define the boundaries of what you’re likely to encounter.
Why you keep seeing the same books
If you’ve ever felt like your feed is repeating itself, there’s a reason for that.
Researchers refer to it as “popularity bias.”
A study on recommendation systems found that popular items are disproportionately recommended, while less-known works receive far less exposure—even when they might be equally relevant. (Source: Popularity Bias in Recommender Systems, arXiv)
This creates a loop:
books that get attention early → get recommended more
books that get recommended more → gain more attention
And so on.
The result? A relatively small group of books dominates visibility across platforms.
The hidden side of the algorithm
Another study found that recommendation systems can increase “homogeneity” in what users consume over time—meaning people are exposed to a narrower range of content as the system learns their behavior. (Source: Algorithmic Confounding in Recommendation Systems, arXiv)
And if your taste leans slightly outside the mainstream, the problem gets worse.
Research from CWI (Centrum Wiskunde & Informatica) shows that users with niche interests often receive less accurate and less diverse recommendations than mainstream users. (Source: The Unfairness of Popularity Bias in Recommendation, CWI)
So while it feels like you have endless options…
You may actually be seeing a narrower slice of the literary world.
What kinds of books get left behind
Without giving too much away from the episode, we explore several types of books that tend to struggle in algorithm-driven systems.
These often include:
slower, more subtle stories
books that don’t fit neatly into a single genre
stylistically complex or literary writing
indie books without early traction
translated or culturally specific works
These books aren’t rejected outright.
They simply don’t gain the visibility needed to enter the recommendation loop.
Readers are starting to notice
This isn’t just theoretical.
Readers themselves are beginning to push back.
In a recent article, one reader described algorithm-driven discovery like this:
“The algorithm… gives you the same books within the genre.”
(Source: The Guardian, 2024)
That sense of repetition is becoming more visible—and more frustrating.
A different way to discover books
In the episode, we explore practical ways to step outside that loop—without making discovery harder or more time-consuming.
Because better reading doesn’t require more effort.
It requires better inputs.
Algorithms are very good at showing you what’s already winning.
But the most interesting books—the ones that feel new, strange, or unexpectedly brilliant—often sit just outside that system.
You just have to know where to look.
Listen to the full episode to explore how algorithms shape your reading—and how to take back control of your discovery.
