Using Flows to identify friction and deviation in the customer journey

How can I use the Adobe clickstream data and then run correlation analysis or segmentation to identify factors causing friction?

Idea is to define deviation as moving to any path that doesn’t lead to an app start( product application page), I thought of using application page as one focus node and backtracking and capturing the data in csv to define frictions. Does anyone have any other approach which is efficient as well, mine would be brute force looking all the way down.