For a paper about a new marketing tool that replaces predictions based on information about individual users (their age, gender, height, etc.) with both individual and social-network information (their age, gender, height AND also how many friends they have on social network, what their friends buy, how much they spend, etc.) we wrote an article to a marketing magazine ("World Marketing") and we now need to generate a graphic to go with the article.
The visualization essentially needs to take a table that we attach to the this project description and visualize it (either by making the table easier/clearer to understand, adding illustrations to drive the points, or suggesting a new way to express the data, your call).
The key message that we are trying to communicate is that when you use the 'new method of marketing' (the lower row in the table) you have more information from the get-go (you know something about a user who joins the service as soon as they joined just by observing their friends' behavior), and that throughout their experience (i.e. from the day the joined Uber and throughout the time they were an Uber customer until they left) you always have more information about them. Also, you can _predict_ when they might leave you by observing the behavior of their friends (if all your friends stopped using Venmo you are less likely to have anyone to transact with and may leave soon yourself), so we can do things before you leave to prevent you from leaving before you churn.
So the bottom line - new marketing research method is better than old one in knowing things about you _earlier_, knowing _more_, predicting _before_ you do things, and is simply better.
We are happy to have visuals replace all the text, not have a table at all, or find many creative ways to communicate this message that there are 2 paths and one is better.
Thanks, in advance, for your creative ideas.
UPDATE: (Sept. 3, 2017: 09:05 EST).
One comment that was not clear in my brief is that we want to illustrate that the 'network-based approach' (the better approach) is to do analytics during the user's usage of the service and PREDICT that s/he is about to leave/churn, so chronologically, we know about their departure before they actually quit the service. So in graphics the chronology should be illustrated such that one understands that we knew they were going to leave before they did.
Added a PDF "graphic_comments_alex.pdf" with comments on the first submission we received to help designers see what things required changes.
UPDATE: (Sept. 4, 2017: 12:45 EST).
One comment that we may have not been clear about so wanted to clarify: The 'better solution' (the network-based one) should have one additional step before the user quits suggesting that 'We can predict churn before it happens." should appear at the section BEFORE the "user quits service."
What we are trying to imply is that we can know 'before' that a user is _about to quit' and we can still do something about it. The text could read: "proactively do something to prevent the user from quitting (i.e. offer promotions)" after we learned that they are about to churn.