Dataset
For this project I explored my personal Netflix viewership data. Netflix allows users to download their personal viewership data via the account settings page on Netflix.com. Seeing as we are well into the streaming era there has been a ton of study done to pull as much insights as possible from users viewership habits in efforts to better understand consumer viewership habits and improve the respective product offerings available. Since there are many firms competing to offer compelling services I thought it would be interesting to take a look at it at the individual level by looking at my own personal viewership data.
I believe by better understanding the data at the individual level I can better understand the types of insights that can be pulled from users at scale once the study moves from individual level to a collective level(user segmentation) or full to scale(user base).
The dataset I downloaded included data from other people that have access to my account. Luckily, we use this Netflix account under different usernames so it was easy to filter the other users on the account out of my working dataset. Once I filtered and cleaned the data I was left with my viewership data from 2010-2020. Over 15,000 watch sessions.
Overview
I was really curious to see if it would be apparent in the data that my streaming viewership habits had changed during the 2020 segment of the COVID-19 Pandemic. I, like millions of others, found myself at home a lot more- doing my best to process the uncertainty that was in the air as the pandemic was unfolding. By mid-march my district was under strict stay at home orders- which lasted, to some degree, throughout the year. This is my research effort to quantify the impact.
Wow! My 2020 watch time increased significantly from 2019.
141% watch time increase from 2019 to 2020
2019 12 days 05:00:53
2020 29 days 22:06:10Although my 2020 watch time started ahead of my 2019 watch time, by February they were near exact. The Covid-19 stay at home order started in my district in the 3rd week of March. This is visible in the data and can be seen as a clear inflection point by what is observed in the chart below.
Looking at the 2019 trend line- we can see what many call the summer drop-off. There was a decrease in watch time during the summer/warmer months of the year when people are more likely to be outside enjoying the weather.
This trend was not witnessed for me in the summer of 2020. I exhibited some of my highest watch times ever during this time. I remember the first few months of the stay at home order being the most uncertain. I must have found comfort indulging in more streaming content at the time.
Sunday remains my biggest streaming consumption day.
I wonder if Netflix has a favorite day to release content on. Is it a better move for them to release content on the day that averages the most viewers or is it a better strategy to release content on one of the least popular viewing day to try to increase watch time on that day?
Sunday remain my highest streaming day of the week. Interesting to see in my 2020 watch time that there was an increase in watch time, relatively, on Monday and a decrease on Saturday. I would not have guessed that. I wonder if this trend is observed in many other users. I suspect there are enough users that share the same habits even with the 2020 habit shakeup- I wonder how the streaming providers adjusted to the influx of consumption.
Multi-season TV shows is where I spent most of my watch time.
Looking at the actions by the streaming companies- this is clearly a data point that has been used to influence business strategy many times over. I, like millions of others, found myself binge watching the available TV shows. I can see why the service providers meaningfully invest in TV shows to cater to every demographic and interest point.
In the words of Mr. Spock, "Live Long and Prosper".
Star Trek kept me going in 2020.
Telling from the data- my interest in the show had been building for the last four years, but my interest truly went parabolic in 2020.
I wonder if this was observed in a significant number of other users. It is certainly possible and likely that users enjoyed new content as well(ie.. the viral Tiger King), but I wonder if users frequently opted to find new tv shows or if they preferred to retreat in to the safety of familiar shows.
Comments