A Review on Bike Reviews

A data-driven deep dive on bike reviews

I ❤️ bikes. It’s why I write this newsletter and I enjoy exploring new places and routes, following the UCI racing scene, and reading the latest news and trends.

But one trend has me asking why? Every bike review I read is positive. There are never glaring concerns and the ride quality and components seemingly resemble one another. So I crunched some numbers to test my hunch.

My Hunch: All bike reviews are positive because all bikes are pretty damn good.

Ok, let’s dive in 👇🏽.

Approach

I looked at 30 bikes equally split across three segments; road, gravel, and mountain. I kept the reviews current, 2023 and onwards, and maintained a price of $2-8K per bike which effectively covers the low and mid tiers.

For each review, I ran a sentiment analysis and designed a scoring rubric (out of 100) to align bike scores across the reviews. If a review already had a score system, I translated that to my 100-point scoring rubric.

What I Found

All the bikes in my analysis are good to great. Of the 30 reviews, 19 were labeled as Great, a score of 80+, and 11 were labeled as Good, a score between 70-79. No reviews scored below a 70 or Good review.

Road bikes reviews prioritized refinement and rider comfort. Gravel aimed at versatility and value. Mountain bikes favored all-out trail competence.

Stepping back (graph below), we can see that price is not strongly correlated to review scores. The shaded section shows the best value, but as the analysis proves, you can’t actually go too wrong on choosing a bike since they rank similarly with price being the exception.

Franken Bike

Following the understanding that all bikes reviewed/performed well I wanted to understand if there was a best bike based on all of these different configurations. So I built up the best bike for each segment, breaking it down by the major components that received great feedback in a review.

Wrapping Up

My hunch seems to be supported by the data that all bikes are pretty damn good. Not surprising if you’re going to shell out on average $6K for a bike which is where most non-cyclists start looking at you sideways.

My take is that R&D is uniform across brands and driven by technology; computer design, wind tunnels, etc. Because of this bikes are similar because the approach is data driven. Any manufacturing, design, or material advancement is quickly replicated across the field and parity reached across a stable of bikes for that vintage (year).

Do you agree? Share your thoughts in the comments below.

Here’s the full list of bikes and their reviews for those that want to dive in. Enjoy!