Here’s something that I’ve been thinking about all day after seeing some interesting data revealed by British Journalist and graphic designer David McCandless, so I thought I would braindump and get it all out in a post.
David and his team scraped 10,000 status updates on Facebook for the words ‘break up’ and ‘broken up’ and discovered trends in the days and months in the year that people choose to end their relationships. The data showed that break ups peak before social occasions like Spring break and Christmas (tightwads).
One slight problem with this data, it’s rubbish. The trouble with using keywords in status updates is that unless its dead specific, you’re just going to end up with junk. For example, how many people use the word ‘break up’ in their status before SPRING BLOODY BREAK? And before summer, and before xmas, and before easter. “Can’t wait to break up!” or “Broken up for da summer Wooowwoo” or just simply “Broken up!…lol” etc etc etc.
Anyway lets forget about the data that David scraped for a second because Facebook CAN tell when people break up – but sadly that info is only available to Facebook themselves.
How? Relationship status updates.
Facebook knows when people end their relationship or ‘get complicated’, also when people hook up. Times that by 500 million people and you’ve got a pretty good picture of when the world just is going through break ups and make ups.
I love using data to find out this sort of stuff, that’s probably why I’m so fascinated with Facebook, once you get to critical mass like Facebook has you can literally find out how the world feels, which brings me nicely on to –
Google Flu Trends
This.is.amazing – Google can predict the outbreak of flu epidemics in cities and regions 1-2 weeks earlier than federal centres for disease control and prevention.
We have found a close relationship between how many people search for flu-related topics and how many people actually have flu symptoms. Of course, not every person who searches for “flu” is actually sick, but a pattern emerges when all the flu-related search queries are added together. We compared our query counts with traditional flu surveillance systems and found that many search queries tend to be popular exactly when flu season is happening. By counting how often we see these search queries, we can estimate how much flu is circulating in different countries and regions around the world.
So Facebook knows when people are love sick, Google knows when people are actually sick, how would LinkedIn know if a company is in trouble?
Well think about how most people use LinkedIn, you pretty much treat your profile like your CV, when things are stable you tend to leave it, never updating or checking it reguarly to improve it. However when it looks like you need a new job you’ll keep logging in, add new career experience, tidy up summary, education etc.
Now if your data sample is a 100 people in one company, then of course you might not get any solid information out of that.
However, lets say you take a company with a few thousand workers, if a significant amount of those goes from not logging in within 3-6 months to logging in frequently AND updating their profiles, then you might just have a company that’s about to cut staff .
What makes this so powerful is that staff know about cuts and redundancies way before internal and public announcements are made, whispers and rumours spread like wildfire and that’s when people start logging in and updating, meaning that LinkedIn knows about cuts and redundancies too – suddenly making this data extremely valuable.
The value of this information for people playing the stock market is massive, so on that basis, I erm, had a quick look at the LinkedIn APIs to see if I could grab it, I can’t think of a way to do it (although it is 1am) so looks like its only available to LinkedIn.
Imagine plugging this into a automated share buying/selling system for the FTSE 500?
Is that even legal?
Anyway some food for thought there