“There is no formula in writing a hit song.” But, perhaps there are patterns.
Finding a “hit song formula” has been on the minds of researchers for some time. Back in 2011, the researchers from Bristol University in the U.K created “ScoreAHit”, a website that tracks hits using a “hit equation”, formalised by analysing the Top 40 singles charts over the last 50 years. And back in 2006, researchers from Emory University in the U.S. used an MRI scanner to record brain activity in teenagers while they listened to music, hoping this would provide insight into music that “performs well” on the brain and thus provide clear indications on future hits songs. It did find significant correlations between brain response and hit songs, but did not go on to change the industry or how music is made.
We wanted to explore this question ourselves. What patterns could we derive by tagging Billboard UK’s Top 100 Charts from 1990 to 2019? Are there common trends that songwriters could take a leaf of? This was a question for our A.I. tagging product.
The Experiment
We compiled these 300 songs (100 songs per decade - 90’s, 00’s and 10’s) and ran it through our AI for the list of tags. Our tags cover 8 core classifiers, including quality of music, vocal presence, vocal gender and more; but for the sake of this experiment we’re only going to focus on genres, energy, moods, key and tempo.
Here are a select few to serve as a refresher course on the hits of each decade:
Results Are In
Our results showed that of the 300 chart toppers, Pop, Electronic and Hip Hop are the most popular genres. Pop leads with a wide margin of 120 tracks compared to Electronic and Hiphop.
Medium energy (neither too relaxed or too upbeat) dominates the charts, having it make up 60% of the 300 hits.
Chart toppers are also most likely to be written in the tempo between 125 to 130 beats per minute,
and in the key of F# minor or A major. Of all the classifiers, Key has the most even distribution and therefore the least important element when it comes to crafting the next Thank U, Next. However it does seem that are a handful of them are more prominent, like Eb minor, F major, G minor and A minor.
As moods go, there’s a high chance that it is energetic, powerful or romantic.
And if you combine all these characteristics with Mood Variance - whether the music sounds positive, neutral or negative - the data looks something like this.
Y-axis - Genres
Colour - Mood Variance
Size - Energy Level (the larger the circle, the higher the energy level)
We can see that hits are mostly positive or neutral, as negative-sounding songs (in dark green) only came up 4 out of 300 times!
Through this graph you could also see 2 major hot spots:
Hot Spot 1 are songs that are Pop or Electronic, within 115-140 bpm, positive-sounding and in medium energy while Hot Spot 2 includes more variety such as Hiphop, Indie and RnB, a slower BPM of 80-110bpm, with a more neutral-leaning soundscape.
Do these descriptions match your favourite hits? What do you think about these results? And most importantly - are we any closer to finding that perfect “hit formula”?
We’ve got loads more where this came from - stay tuned for Musiio’s insights into hit songs and chart toppers. And try running your own experiment! Take a list of your most popular songs and tag them using our tagging demo.
