Anatomy of a Viral Lottery Hoax Sharer
You probably saw it or heard about it – Nolan Daniels posted a photoshopped picture of his supposed $239 million-winning Powerball ticket, along with the generous offer that he would give $1 million to a random person who shared it. If you missed it here is a screenshot as of the afternoon of Monday December 3rd 2012, along with the kind of question always nagging at us about so many things we see in social media: who are these people? As Dan Zarrella points out in his piece about chain letters that applies here is the power of social proof:
Perhaps the most subtle and powerful viral element of chain letters in email is the social proof that comes with many of them. Every time someone forwards one to his or her address book, another list of recipients and senders is attached to it, creating essentially a list of people who implicitly give authority to the message. If one person sends an email to another, the source may or may not be cited, and the sender’s reputation is the only real social authority the email carries, with a huge list of hundreds of others attached it, with popular viral emails, it suddenly appears that the message is common knowledge and the receiver is perhaps the only person left on the internet who wasn’t warned of the danger.
Different from eras past when we would wonder who was sharing this stuff, there’s actually a way to get some insights on a sample of these people. We thus ran a report to profile these sharers in aggregate via Optimal’s Audience Matrix. This self-service analytics tool is available to Optimal’s analytics customers, but we’ve also made the report available to anyone for free who creates an account at AudienceMatrix.com (along with a bunch of other neat analytics! See important NOTE below on this report). By the way, some of the comments are very colorful – this analysis was NOT of the people who liked the post or those who commented on it, but is based on those who shared it. Below is a no-judgement, tongue-in-cheek sketch of the average user who fell for it…
A few small observations – many Californians seemed to skip the sharing party. Perhaps that is because California offers only Mega-Millions and doesn’t provide access to Powerball to its residents (that’s changing in 2013, by the way). While still large in absolute terms (Californians were the largest group that shared), Californians on average under-shared relative to the size of that group on Facebook. Market researchers commonly provide “index” values to indicate propensities among population groups, so in this case there are 11.4 million women on Facebook in California, about 6.8% of Facebook’s total US population, but only 3.9% of the “Fake Lottery Sharers” (“FLS”) were women from California, meaning a 58 index. If 6.8% of the FLS were women from California, they’d have an index value of 100. The default sort order for most of the data in an Audience Matrix report, is by this index value.
Here’s two different ways to look at how the post fared based among US audiences by gender and age index, (screenshot from the self-service report) to give you a good sense of where the demographic sweet spot for sharing of this post is (red is women, blue is men):
Here’s the view by combined age bands which will be familiar to many media people:
The US wasn’t the only place for this lottery sharing – there were people from 117 other countries participating, from Aruba to Vietnam!
Another thing to keep in mind, is that most of these sharing users over-index for a LOT of liking and sharing activity, period. The average person who shares a hoaxy photo with hundreds of (on average, probably at least 220+) friends is probably expressing their excitement and affinity for things a good deal on an ongoing basis! So, wonder no more who is sharing this photo far and wide on their social network. You’ve probably received an email from them before.
NOTE: the total audience figures are estimates of a two-week trend based on a 48-hour window of activity. Although the post continues to accumulate shares, this extrapolation to a 14 day number is hence likely to overstate the total consumer figure, although we believe that the relative incidences of various features of this audience should be accurate. As with any analytics, the larger the sample size, the stronger the data.