Filtering Technology Fails New England Patriots Social Media Team

Posted by Crisp on

A light hearted marketing initiative from the New England Patriots backfired when an automated system posted an offensive tweet to the team's 1 million followers at 9pm last night.  The offensive tweet said: "Thanks for helping us become the first NFL team with 1 million followers!#1MillionPatriots" but included an image of a team jersey with the follower's Twitter handle emblazoned across it - nice idea, but this particular handle was an offensive comment featuring the N word.  

Patriotstweet

An apology followed at 3.30am but by then the image had gone viral and gained even more exposure from articles across global news outlets.  Patriots_apology

If the Patriots' marketing team had had a team of people reviewing this exercise overnight it wouldn't have happened, or at least the impact would have been minimized. If they had known faster they could have deleted the offensive tweet before it gained so much exposure.  

How Crisp's service would have helped

Our team of people would have detected the problem - and the reaction to it - and made sure the digital and PR team were aware straight away so they could work to minimize the impact. We could even have reviewed and escalated the auto-tweet within minutes before or after it was posted depending on the service they'd requested.

The social media team might have been reviewing their promotion, but probably not every minute, overnight. Crisp's team review a brand's content 24/7 so it wouldn't matter what time the reward had been triggered, we'd have identified the problem and woken someone up if necessary.  

Speak to a risk monitoring expert

Crisp

Written by Crisp

Crisp’s mission is to achieve complete Trust and Safety for everyone on the Social Web. Through world-leading AI and a global team of digital risk experts we will achieve complete online Trust and Safety by protecting brands, social platforms, advertisers and kids from the bad actors that exploit, extort, distress, offend and mis-inform.

Read more posts from Crisp »