Data-driven still needs human decisions

Unless you’re blind and deaf (or not a techie), you’re aware that in the past week or so reddit had a meltdown. During the raging debate that followed re: firing Ellen Pao (and why Alexis Ohanian chose to tell the world on Friday afternoon), a redditor linked our blog post on worst days and times to send a press release for context on why Friday afternoons are coveted when you have bad news to shove out the door. Have any doubt about the power of reddit? That one link resulted in 10,000 new visitors in a single day to said post.

Data would say to us, syndicate this, talk with people on reddit about it, take action. Machines reading all of these signals would have seen a clear indication to promote/engage. As an experienced marketer and sometimes lurking redditor, my response to said machine decision is “Whoa, whoa, whoa, not touching – reddit is a tricky beast of a thing. Besides it’s not positive. We value positive conversations here at SHIFT.”

Data makes humans smarter, provides direction, and keeps pushing us to question what we think we know.

We don’t deny that data is valuable when it comes to making human-driven decisions. However, being data-driven does not leave in entirety the decision-making up to machines. Machines aren’t yet capable of inferring good judgement as they lack the understanding of context and positive and negative sentiment. And good judgement says don’t engage with angry people on the internet. Even on a Friday afternoon.

Chel Wolverton
Account Manager, Marketing Technology

[cta]