Not able to get a good accuracy score for the classification problem

I have taken a music popularity dataset which has five classes based on the popularity of the songs.I have made a Random forest model to predict the popularity of a given song(given its features).I have done feature engineering like feature scaling and feature selection. Without doing feature selection, the accuracy that I got was around 0.625
But even after doing feature selection and feature scaling the accuracy did not improve even a bit.It was still around 0.62.
What have I done wrong?
Is there any other way by which I can increase my model’s performance?