I have an e-commerce dataset for a month consisting of five columns where the fifth column is "Revenue". The other four columns are different factors for the revenue such as "Page-Views","Visitors","Bounces" and "Click Through Rates". Each row is basically a particular day of the month(hence 30 rows in total).
So, if I would like to find out the contribution/importance of each factor for the revenue, will correlation values between the variables be sufficient or are there better metrics ?
You could try implementing a predictive model, like linear regression or a decision tree and then measure feature importance using the model. However, measuring feature importance is a slippery concept. This question illustrates that feature importance is not trivially measured in regression models. This paper shows how to measure importance of features in a decision tree. To me, these are richer measures of just linear correlation. Distance correlation is another measure that does not need predictive models, and more meaningful than just linear correlation.
To explicitly get the percentage of contribution for one variable to another feature, you can take their (pairwise Pearson) correlation and square it. That is the percentage of variation in one explained by variation in the other. (This is the R-squared of a linear regression, if you just regressed one variable on the other.)
Similarly, if you want to control for other factors (e.g. you want to control for day of the week), you can run a linear regression with just the control variables, then one with those control variables and another one of your predictors. The difference in adjusted R-squared between those models will be how much of the variation in Revenue is explained by the other predictor, beyond what's explained by your controls, and with a slight overfitting penalty for adding another variable.
