Automation. It’s a dirty word when it comes to social media. And while it’s not always a bad idea to automate some aspects of your social media presence (like scheduling, for example), brands have run afoul of automated interactions in the past.
Here are three examples of brands that found themselves in hot water because they automated the worst thing possible on Twitter: replies.
You might be surprised to see Oreo on a list of brands that messed up on Twitter. After all, their smart response to last year’s Super Bowl blackout and their usual dominance over all things social should be enough to shield them from this list.
Not so, unfortunately. Even the greats can make a mistake every now and then.
Oreo’s auto-reply slip-up happened earlier this week. It contained words that are offensive to many, so if you have delicate senses, avert your eyes now.
So how did this vulgar username get tweeted from Oreo’s official account? The most likely scenario is that somebody on the team set up auto-replies that were triggered when a Twitter user sent them a tweet. Most of the automated replies were sent to innocuous accounts. But this one wasn’t – and it got a lot of (negative) attention from the Twittersphere.
Another big brand that fell victim to the auto-reply was Bank of America. This summer, an Occupy activist tweeted a picture of himself being chased away from a Bank of America building by New York police, after writing protest slogans in chalk on the bank’s sidewalk. And he tagged @bankofamerica in his tweet.
After being retweeted a few times by fellow Occupy supporters, the tweet earned itself a reply from what was presumably a Bank of America robot. Tweets like “I work for Bank of America. What happened? Anything I can do to help?” would make sense if this protester was looking for some customer service, but they came off as tone-deaf and ignorant given his original tweet.
The incident quickly spiralled out of control as more protesters tweeted to the bank and received cheery, generic customer service verbiage in reply.
The auto-replies continued for a surprising length of time. You can read them all here (but be prepared for some coarse language on the part of the protesters).
Following this embarrassing gaffe, a Bank of America spokesperson insisted that there were real people behind each and every one of these vapid tweets, and that they didn’t use an auto-responder. But the fact that angry tweeters were answered by offers to help with user accounts makes this a good example of when not to use a bot – whether that bot is a program or just a zonked out customer service rep.
Marketing messages gone wrong are one thing, but when a corporation blithely tweets about a tragedy using auto-replies, it leave a bad taste in the mouth.
In the summer of 2012, Matt Fisher took to Tumblr to call out Progressive for not only failing to pay insurance on his sister’s automobile death, but for actually defending her killer in court to save themselves some money.
Whether or not these allegations were true isn’t the issue here. What is the issue is how Progressive responded.
As Gawker reports, Progressive tweeted the same, generic auto-reply to anyone who sent a tweet criticizing them in their handling of Mr. Fisher’s sister’s death. The auto-reply was sent apparently indiscriminately, at least 16 times before someone hit the brakes on the bot.
These three examples show that even the biggest brands struggle with how to properly incorporate automation into their social media strategy. A good rule of thumb is to not automate interactions – something that these brands had to learn the hard way.