Pivott

Music + an AI toolbox + social networks = the perfect solution to catch millennials' attention. Find out how Pivott created a new dynamic campaign for Proximus.

Music Maistro

Millennials love to tweet but most of their tweets don’t get noticed in the mass. With that strong insight, Pivott wanted to give millennials the feeling of being noticed on Twitter with a personalised reply, and this not in a dull automated way where they get a reply that doesn’t fit, but with a real appropriate personalised tweet reply! So the Pivotters used one of the biggest passion points of the Proximus audience: music, and created music mAIstro, a awesome and surprising way to make the millenials get noticed.

Music mAIstro is a kind of toolbox that provides AI-driven personalized tweet replies. The reply contains a part of a music lyric that matches the mood of the tweet. It also includes a Spotify link so that people can immediately listen to the song, and a link to the full lyrics. And the cherry on the cake, if the artist played on one of the festivals sponsored by Proximus, a hashtag of that festival is added next to #Proximus and #CloseToMusic. Pivott could have hired people to do all this manually but it would have been a titanic job and thanks God they used the ultimate technology mix! The different technological tools of Music mAIstro work together in concert: one checks the politically-correctness of the language, a second one makes sure the reply is in the right language, a third one constantly scans the top hit artists database, and the last one looks for tweet-keywords matching lyrics-keywords.

To kick off the campaign, we replied to tweets of influencers and celebrities, like Eden Hazard or Donald Trump, to get the buzz going. It was a simple yet effective way to create awareness around music mAIstro. Then, during one month, we had a social campaign with Facebook, Twitter, and Instagram image and video ads focused on awareness and consideration with the message: Music mAIstro brings you #CloseToMusic. In the end, we got 317.000 impressions, analysed 139.680 tweets and replied to 6.504.