Crunching the Data When Your Post Goes Viral

Singer-songwriter Marian Call wanted to write about our changing relationship to work. She sent out a quick tweet to her followers asking what their first jobs had been before she went to sleep.

What were your first 7 jobs?   Babysitting, janitorial, slinging coffee, yard work, writing radio news, voice-overs, data entry/secretarial   — Marian Call (@mariancall) August 5, 2016

Call woke up to find her tweet had gone viral and she got replies from many people including some celebs like Buzz Aldrin (Dish washer, Camp counselor, Fighter pilot, Astronaut, Commandant, Speaker, Author) and Sheryl Sandberg (1. Babysitter -twice – Office receptionist, Salesperson in clothing store, Aerobics instructor, World Bank health team, Children’s Defense Fund).

Marian did not use a hashtag but tag emerged and that made it easier to see responses. Unfortunately, different versions were used and are still active, like #firstsevenjob or #firstsevenjobs and #first7jobs. And the query and tags also appeared in other networks like Facebook.

She was interviewed on the Make Me Smart podcast  and she explained that then needed a way to to crunch the data from all the Twitter responses.  She was contacted by the social product manager at IBM who had heard her interviewed and they put the data into their Watson supercomputer and then were able to produce an infographic of the data.

infographic

The data shown is interesting and shows commonalities across the world. Of course, we can’t manipulate the data or request other queries. Call said she would have preferred a spreadsheet she could sort and search.

This little exercise points out one flaw with Twitter and many other social sites – no easy way to pull user data and draw conclusions about it. There are paid programs and people who can do those things for you, but a free, built-in way to do those two tasks is not reality.  Most of our posts will not go viral, but even gathering the data from a normal social media campaign can be difficult.

Marian’s experience did get others to try their hand at the task without a supercomputer. One example is at blog.monkeylearn.com/analyzing-first7jobs-tweets-monkeylearn-r/