What percent of my social media content are ads?



I collected this data by using each companies’ official android app. Using the home page of the app, I scrolled at a consistent moderate speed for the first 100 posts/elements and counted how many elements were advertisements or suggested content. Advertising behavior is a fickle thing and might change depending on how frequently you use the app, how and what advertisements you engage with, or even how quickly you scroll through the information. As such, your mileage may vary, but these were my personal results—take them for what they are worth. 
I made the graphic in Excel and PowerPoint. I have provided more information about the collection for each app.

Facebook: I use Facebook lite under 30m a day but nearly every day. A few times when scrolling (2-3/100 posts), Facebook “suggests” content or posts where I have not liked anything from the page nor had my friends engaged with it. I counted these suggestions as advertisements.

Instagram: I use Instagram a few times a month and have not posted in years. Interestingly, I encountered 0 ads in the first 100 posts on Instagram. I wondered if this was because I didn’t use Instagram frequently and they did not present me with ads in an effort to give me a premium experience and come back. I logged into my wife’s Instagram (used daily) and quickly found ads.

YouTube: I use YouTube for at least an hour every day. I never scrolled the front page of my YouTube to count ads. This estimate does not account for ads you encounter when clicking on videos. Interestingly, no ads were presented to me after the 50th post. My guess is some kind of restriction such as preventing ‘x’ ads in a short period of time. For comparison, I decided to use the first 50 posts as a representation of what the average user would experience (18% ads rather than 9% ads).

LinkedIn: I use this a few times a month for a few minutes. The ads here were the most difficult to detect. LinkedIn had the exact same behavior YouTube--ads were presented across the first 50 posts and none were presented after. Like YouTube, I used the first 50 posts to estimate the advertisement rate.

Twitter: I use twitter a few times a week for 20-30 minutes at a time.

Reddit: I use Reddit for about an hour a day and used the front-page (Canada) for this experiment.


Comments

Post a Comment

Popular posts from this blog

Boy vs Girl Marvel Movies?

Image
Purpose: To investigate what, if any, Marvel (MCU) movies males prefer more than females.     Source(s): I used the Wikipedia list of MCU films  Wikipedia page . I scraped most IMDB ratings mid-2019, but manually added in the post-2019 movies last week. When you have an IMDB account, there is a page that shows you gender demographics and age demographics. IMDB gets these demographics from their IMDB members who have registered with their gender and age. Excel Data  | Annotated R Code | Disney version | Pixar version   Highlights and Considerations: •           IMDb ratings are rated on a 10-point scale. The average IMDb rating (across the 26,000 most rated movies) is 6.92 with a standard deviation of 0.96. This means most average movie ratings (68%) across   movies are within 1 point of each other. Therefore, while a decimal point difference between gender ratings might seem negligible on a 10-point scal...
Read more

COVID-19 Canadian Dashboard

Image
Canadian Covid 19 Dashboard Social isolation can really change one's behavior. For me, it has meant spending a lot more time compiling and exploring different datasets in R. As I'm sure many others are, I have recently been playing with COVID-19 data. I was curious about how my home country of Canada has been handling COVID-19 so I created this dashboard to monitor the progress. Each graph hosted here is generated by my publicly available R code 1 and updates whenever I run the code (which tends to be daily). The data comes from official Canadian government reports 2 . If you have any questions, do not hesitate to ask. A few notes on the data... As one might expect, data during a pandemic isn't always accurate and up-to-date. There are several considerations when interpreting this data: New cases and deaths are reported on a 14-day rolling average. As day-to-day statistics can be volitile and influenced by systematic varaibles, I used a rolling ...
Read more