Manufacturing Popularity by Lying


A few years ago, Duncan Watts — a network-theory pioneer at Yahoo! Research — wanted to test a theory. He ran a series of tests to see if popular music could be made popular again or if the original success was just plain luck. Watts wondered, can you browbeat people into thinking something is popular when it isn’t? (Sounds like some advertising out there.)

Watts partnered with Matthew Salganik to come up with the right experiment to sort this all out. Their concept? To lie.

They tested some popular music with a sample group that was not familiar with the songs and asked them for their ratings. They then inverted those ratings so that the worst was the best and the best was now the worst, and gave the songs, complete with false ratings, to a different group.

The unsuspecting new group was duped into believing the original ratings were accurate and snubbed the fake-low ranking ones. Apparently, flat-out lying works!


Not so fast.

Over time, the previously top-ranked songs crept back up the charts amongst second group and the previously lower-ranked ones dropped. However, despite the fact that the “good” music eventually did rise to the top, people participating  in this topsy-turvy experiment ended up consuming less of everything.

Astroturfing (fake grassroots) is the worst crime you can commit in social marketing. Although lying in this space may get you a short-term gain, the long-term forecast is lower sales and probably (hopefully) a damaged reputation.

• Learn more about this experiment here.