Can Social Media Metrics Predict American Idol Results?

From the archive. Originally published on migmol.com on 2021-04-10, recovered and republished here in 2026 at its original address. Unedited except for formatting.

Part of a three-post series testing whether public social metrics could predict a competition result, published with a mid-season accuracy update. The method behind it later became Konteksto.

American Idol is one of the biggest, longest-running musical contests in the US. It appeals to a broad demographic of Americans which is evident in the range of musical styles of the contestants (and even the hosts).

In the past few years, there has been a bit of talk about how usually a WGWG (white guy with guitar) ends up winning the competition. In fact, this has been the case for 56% of the winners (10 out of the 18 seasons). In 11 out of 18 seasons, a guy has won. When you look at the numbers, it doesn’t seem so bad, but if you consider recency since the trend started in season 5, 71% of the time the stereotype prevails.

I’m not sure how I feel about the above-mentioned stat, but I will just pass it off because being the show’s ultimate winner doesn’t really hold that much weight in the real world e.g. Jennifer Hudson, Katharine Mcphee.

American Idol usually gets the most buzz during the auditions. In the earlier seasons, it was sometimes because of funny auditions like William Hung, but more recently, auditions go viral because of pure talent and sometimes the contestants’ personal stories. As the show progresses though, a contestant will need more than a story and their natural skill to succeed.

I’ve always wanted to see if social media metrics could predict the success of contestants. This season, I’ve finally decided to do it. I’ll be comparing several metrics and making hypotheses on what’s going on behind the numbers. This should be interesting!

To begin with, three questions: 1) is there a correlation between Instagram followers and ranking, 2) how well do audition video views compare to how contestants progress, and 3) how well will the video views of the top 24 predict who will end up in the top 16.

Since the top 24 will be cut to 16 this weekend, let’s see who might get booted out if we base it on YouTube views.

[image: table/chart of contestant YouTube views, in thousands]

Looking at the above list, I already have some personal picks in the bottom 8. No surprise with Cecil Ray being at the bottom because he looked like he just woke up from a nap and decided to perform a song.

We’ll find out next week if the above ranking can predict who makes it to the top 16. But since we don’t know have that data yet, another question we can check is whether Instagram followers correlate to YouTube views.

[image: table/chart of contestant Instagram followers, in thousands]

As you would expect, the contestants with the most IG followers are also in the YouTube top 8. Notable outliers are Ava August and Alyssa Wray, who are underperforming on Insta, and Jason Warrior and Chase Beckham who are underperforming in YouTube views. Alana could work on her Instagram game as sleepy boy Cecil Ray has 3k more followers than her. To be honest, I had a hard time finding her profile because her stage name is a common first name, bad idea right there.

Which metric is a better prediction of success: Instagram followers or YouTube views? We shall see.

Now let’s take a look at how the audition videos stack up. This is an interesting metric because it can show how much raw talent or personal story translates to success on the show.

[image: table/chart of audition video views, in thousands]

Major underperformer is Anilee List. I think her duets performance was actually better than her audition, but her ‘story’ was featured in the audition video so that may be a reason why she performed better in that metric. Ava August’s audition video was actually shared with another artist so that may be the reason why she’s at the bottom of the list. She doesn’t have such an interesting story also (she’s just the youngest contestant) but her talent shines through. Madison Watkins was labeled as having the ‘total package’ during the auditions but the views might not have agreed with that judges’ statement. Admittedly, I didn’t remember her audition video also but took note of her when she performed in the top 24. Perhaps, true to that total package label, she really shines more on stage. Last, Hannah & Cecil Ray take a big tumble from audition views to duets numbers, but let’s see how America votes.

Can we make any bold predictions this early in the contest? Let’s look at the numbers.

From the YouTube views from the all-star duets, Willie Spence seems to be the favorite. He also had the 3rd most views for the audition video perhaps aided by his personal story. I have my money on Ava August.

In the following weeks, let’s see if anyone can catch up to Willie Spence. It would be interesting to continue to see how the contestant’s stories (as featured in their audition videos) rank up against their final rankings. It would also be nice to see how their Instagram numbers change as the show progresses. Right now, the WGWG trend doesn’t seem to be appearing, but let’s wait for America’s vote. That’s the other thing I wanted to compare: Instagram & YouTube’s global audience vs America’s votes.

We have great talents this season so this should be interesting. Updates same time next week!