Modern Metrics Lessons from Fat-Fingered Dialers

When I was a data-entry grunt at an ad firm in Minneapolis (now Saatchi & Saatchi), I was tasked with compiling stats from a company that provided phone systems to businesses to determine how often people mis-dialed.  Presumably, there was a new feature to reduce dialing errors, so we wanted to scope how big the problem was.

It was big.

Fifty-five percent (yes, 55%) of the touches to a phone at the time (this was in the late 80's) were mistakes.  This didn't mean they necessarily fully dialed the wrong number, but pushed numbers they didn't mean to push.  These are Oops-I-have-to-dial-9-to-get-out or Oops-that-was-a-2-not-a-3 moments. 

I was astounded by the number so I went to my supervisor, "these numbers are saying that people dial more WRONG numbers than right ones.  Can that be true?"  The supervisor shrugged it off and said, "numbers don't lie."

OK, so why does that matter now?  Well, think of the intricate data we collect about our clicks, views, and conversions from our social media ad or email campaigns.  I know I've accidentally clicked an ad or two--or seventeen.

Need I say more?