Opentopic is teaming up with Dove and Razorfish to launch the #MyBeautyMySay campaign, encouraging the media and the general public to focus on the athletic ability of women in sports, not their looks.

Objective

The campaign launched in Times Square in New York City via a billboard that will broadcast real-time media comments that put the focus on the looks of female athletes rather than their performance.

“I feel like our world today is pushing beauty over athleticism for young girls, I want to be an advocate to change that.” – Shawn Johnson, Former gymnast and Olympic gold medalist

Digital billboards will also be running in Los Angeles and Toronto throughout the month of August and aim to engage users via social media using the hashtag #MyBeautyMySay and online at www.dovehaveyoursay.com.

When a user clicks on a quote, they are able to automatically send Dove’s tweet to the media outlet or person it came from. This real-time engagement gives consumers the feeling that they are able to make a real difference.

On Twitter Dove is sharing graphics, facts and quotes from media sources that are judgmental of female athletes. Dove is also inviting women to share their own motivational stories via the hashtag #mybeautymysay

This social media campaign comes full circle at the “Have Your Say” website, which offers interactive content, statistics, and a list of negative tweets that visitors can click on and reply to.

Purposeful messaging. Successful social media campaigns. Audience empowerment.

Approach

  • Understand Natural Language: Using Speech-to-Text, Natural Language Processing, and Taxonomy capabilities, we were able to analyze media mentions from 300,000+ daily sources and score the relevancy of each mention for 5 defined categories: Hair, Body, Age, Clothes, Beauty.
  • Extracts Concepts and sentiment:
  1. Concept Expansion: Expose media mentions that specifically talk about a female athletes.
  2. Sentiment Analysis: Applied to Concept Expansion to expose media mentions that specifically talk about a female athlete in negative sentiment. 
  • Highlights mentions in real time.

Outcomes / Insights

The Technology behind it implied 4 stages:

1Collect data (use jutro and magazyn to source articles related to sports)
2Understand volume of content that can be analyzed per day/month (volume of media that commented on female vs male athletes)
3Create custom classifiers for appearance mentions (add training data to help Watson find similar appearance mentions)
4 Reflect on training process (how content is classified, how much training was necessary, what mistakes the algorithm was making, ways to improve on the process)