My company wants to find a correlation between website visits and marketing spend between sales. Essentially, are website visits or spend a good predictor for sales? In addition to that, we hypothesize that desktop visits/spend is a better predictor than mobile visits/spend when predicting sales. We want some way to prove this somehow.

My coworker who came to me with this problem mentioned that regression analysis may be the way to go. And the reason she came to me was because I have a degree in statistics. I've been out of school for over 4 years and have never used my stats degree in a real job. I am an analyst, but more of a systems analyst than a data analyst.

That being said, I'm a little rusty when it comes to regression analysis, so I wouldn't want to mislead anyone to thinking that regression analysis is the route we should be taking when it may certainly not be. Which really leads to my first question, would multiple regression analysis be the correct statistical model in this case? Or are there any other models that make more sense?

Also, if there are any general suggestions on resources and/or software for similar problems, I would greatly appreciate it.

I may have more questions depending on responses, but hopefully you have enough to response.