Last month I decided to start writing on a semi-regular basis. I wasn’t so sure what I should write about. I didn’t even know where I should reach people that would be interested in the work that I do. During the day, I work on my site which is an online app to learn Japanese where I also provide simple tools like my new hiragana to romaji converter. But, these topics often aren’t that exciting to the bigger audiences on Medium. Hence, I was wondering what I should write about. Being the software engineer that I am, I had to write a program to come up with a story.
I started by taking a look at one of the most popular publications called Hackernoon where fellow software engineers write about all kinds of tech related content. From Bitcoin to GraphQL to finding a job as a developer, you name it. Given the mix, I wondered what the average claps in 2018 were, wrote a web scraper for the Medium platform and analyzed this years posts. Then I published my first Medium post in a publication and it found modest popularity, still under Hackernoon’s average claps though. After that, I told a friend about my new endeavor to start writing and showed him my recent post. Him being an avid reader of Towards Data Science, he urged me that I should do the same for this publication. Because, he thinks that it would prove itself more popular. And that’s what I did.
First I was taking a look at the distribution of posts per claps in Towards Data Science. As you can see in the graph below, the distribution is very unsurprising with no peculiarities.
The data suggests that the average posts receives 515 claps but looking at the data this applies only to 22% of 2018’s posts. If we calculate the amount of claps of 4 different percentiles, the story suddenly looks differently:
- 10%: 27 claps
- 25%: 73 claps
- 50%: 185 claps
- 75%: 449 claps
After analyzing the posts average claps, I was wondering: “What’s the best day to post on Towards Data Science”. I simply grouped the posts based on their weekday and calculated their averages. In the graph below we can see, that there is a high chance of going popular on Sundays with the lowest competition. After that, Saturdays follow and the data suggests that it’s unwise to publish from Tuesday till Friday with the highest competition.
In the end, I decided to publish this post on a Sunday to have the odds on my side. I hope this post gives you enough courage to also start contributing to Towards Data Science and helped you decide when it’s the best time to publish. Now I’m curious whether this post will alter the distribution of posts for the rest of the year. So I have set a reminder for January 2019 where I’ll take a look at the data again and see if it made a difference.
This post has also been published on Medium.