# How to quantify impact of one data on another?

I work with an airline company trying to quantify if google search do lead to ticket booking.

I have two dataset: (1) Search data = basically date of search, fly-date for search (2) Booking data = date of booking and fly-date for booking.

Is there a mathematical way to prove or dis-prove the correlation between search data or booking data. Ideally booking happen even without search so not sure of how to establish correlation.

• Compared to what? I am sure you will find at least one match of someone googling and "shortly" after booking a ticket. Does this mean that there is a relationship? Would this person have booked a ticket even without googling? How would you know? Suppose 20 % of people google and then book, is this a relationship? How many people book a flight without googling it first? If you are going to match the two data sets to find which people googled and then booked, and you will declare that everyone else did not google, what if they binged? Jul 12 at 11:33
• @user2974951: Fair points. To make things simple: I would like to test if there are significant portion of ppl booking after google search. (Here: significant could be 20% to begin with). Does this makes sense? (If they binged or not is difficult to answer as the data is not capturing how long they searched) Jul 12 at 13:28
• What do you mean by 'fly-date' for search and booking?
– mkt
Jul 12 at 20:08
• fly-date = travel date,,, booking date = date on which ticket was booked. Jul 13 at 5:02