Can 5200 Songs Tell You Where Viral Music Is Going?

We used our AI to analyze 5,200 viral songs from 104 weekly charts in the USA from 2019-2020 to see if we could discern useful patterns in Genre with actionable takeaways for artists and labels.

SURFACE-LEVEL RESULTS

On average, Hip Hop emerged as the clear winner, accounting for 26.14% of all tagged Genres, with Pop trailing a very far second at 12.5%, providing a very strong indicator that artists who do not already create Hip Hop music may want to start collaborating with Hip Hop artists to maximize their viral reach. 

This is something that highly successful Pop artists such as Dua Lipa and Camilla Cabello have already incorporated into their strategy, by collaborating with Hip Hop artists like DaBaby. 

With his 50+ recent collaborations, DaBaby has certainly made out well. Other Hip Hop artists may also want to leverage their Genre to secure similarly successful collaborations.

A LITTLE DEEPER - HOW DO GENRES BEHAVE OVER TIME?

Trends change from week to week, so it’s important to be able to read those trends.

We took a look at how each of the top 12 Genres behaved in relation to each other. From week to week, Genre scores would either increase or decrease. We observed that pairs of Genres would move in parallel (increase-increase) or opposite directions (increase-decrease/vice versa) to different degrees.

This graph shows the likelihood that a pair of genres will trend in an opposite direction.

For example, Latin and Reggaeton only exhibited opposite trends 9.71% of the time (meaning that they trended in parallel 90.29% of the time). In contrast, Hip Hop and Pop exhibited opposite trends 67.96% of the time.

In general, the further away from 50% we get, in either direction, the more likely you are to be able to predict the trend of one Genre by observing the other Genre in the pair.

Below 50%, as we get lower in score, we will see an increasingly predictable relationship between Genres. Above 50%, as we get higher in score, we will see an increasingly predictable opposite relationship between Genres.

As you can see in the graph, an increase in Reggaeton score is correlated with an increase in Latin score. The opposite holds true - a decrease in Reggaeton score is correlated with a decrease in Latin score.

On the other end of the spectrum, a change in Pop score is likely to be correlated with an opposite movement in Hip Hop score.

As you can see, the relationship is not as strong as in the previous example, but there is still a clear relationship.

The last example shows that an increase in Pop Score is not particularly likely to be correlated consistently with an increase or decrease in Reggaeton score.

In other words, these Genres are not particularly related.

SO WHAT DOES THIS MEAN?

If you are releasing music in the USA and hoping to ride the wave of a trend or predict when the next one is going to happen so that you can plan to release your new music before it dies down, don’t just look at the Genres that do well on average. That’s important, but understanding how the tastes of your target audience ebb and flow so that you can be in the right place at the right time is also important.

This is a combined graph of all Hip Hop genres tagged (Trap, Pop Rap, UK Grime etc) in this time period. At one point, combined Hip Hop accounted for 57% of all music listened to, and at one point, it accounted for just 18% (less than a third of that peak). 

Releasing new music before a peak or before a dip could make a huge difference to an artist’s career - working off the average (40.59% in combined Hip Hop) can be useful, but it does not take timeliness into account. 

CONCLUSION

Data is good, but how you look at it can tell many different stories. If you are using data to determine your strategy as an artist or a label, don’t leave any stone unturned.