On Medium, Fans Still Matter the Most
Reading time counts, but fans are what’s important
Back in October of 2019, Medium announced a dramatic change in calculating writer earnings through the Medium Partner Program.
Prior to this change, Medium based its writer earnings through the claps, a “Like” system where a reader could endorse an article on a scale of 1 to 50. There were a variety of challenges with this model including not having to read the article to clap for it and the number of claps being somewhat arbitrary.
Regardless of implementation and results, most writers’ earnings have decreased, the idea of prioritizing subscriber reading time over claps made sense. Starting in December, I collected data on my views, reads, claps, and daily earnings. Fast forward five months and I can confidently say that fans/claps still matter the most. In this article, I’ll go over my methodology and the data as well as the conclusions I’ve drawn from the analysis.
My Daily Medium Statistics
Each morning, I check my Medium Partner Program page to record my updated running total for the month. Since Medium does not provide a day-to-day breakdown of earnings, it’s necessary to update the spreadsheet every day. Along with the running total, I record the number of views, reads, and fans. The following stats are then calculated:
- Daily earnings: current running total minus previous running total.
- Read ratio: reads divided by views.
- Earnings per View/Read/Fan: daily earning divided by views/reads/fans.
Just looking at the numbers is not particularly insightful, so I graphed each of the stats. Aside from my daily earnings steadily waning, what I found most intriguing was that the correlation between fans and earnings is significantly closer than fans and reads. Calculating the coefficient of determination (r squared), we see that fans (0.775) are a markedly better predictor of earnings compared to reads (0.5).
There are some important factors that may influence the quality of the metrics. While I do not believe they would flip the results, it is responsible to acknowledge the limitations of the analysis.
- There is no way to differentiate between member and non-member reads.
- The length of the article will impact how meaningful each read is.
I’ll continue to update this spreadsheet, I’m hoping to automate it in the future, but for now, it’s safe to say that despite Medium changing their earnings metric to member reading time, claps still reign supreme as the primary indicator of earnings. I’m interested to see if anybody else has kept similar stats and what their results look like. If you have, share your experiences below.