Updated: Jan 3
Written by: Megan McDowell
In a recent article entitled We Made a Facebook Messenger Bot to Track People’s Feelings about the Election looks at how The Washington Post created a bot to track the emotions of a sample group leading up to the election. The bot named Feels was created to monitor and compile data regarding shifts in the news cycles during election times and how they were impacting this group.
How the study took place was the newspaper asked readers each night how the election was making them feel that day. The participants were asked to respond using five emoji’s with the option to explain in one or two sentences. The emojis ranged from a happy face to sad face to angry face. The following morning the bot would produce a report charting the emotions of the respondents from the day before, as well as some of the most interesting responses given by the group. Following the election the respondents were given a personalized report charting their feelings throughout the whole process. This idea of charting feelings in an online research community is a great way to gather data on emotions regarding any facet of your company.
When asked why this approach was taken, the response highlighted the need to have a fresh approach to readers, “we’d launched several bots this year and wanted to apply what we’d learned. In particular, we were eager to try an approach that engaged with users rather than simply delivering news to them.” Having used online research communities before, the paper knew that creating positive spaces was challenging especially in regards to politics but that it could lead to exposure to new ideas and create a sense of community.
The Washington Post spent a lot of time in preparation and planning for such a study. In order to show the participants emotional journey the questions had to be framed so that there was consistency in what was being asked over time. “We decided to ask readers, “How is the election making you feel today?” This phrasing allowed the question to be consistent, but for the readers’ answer to be variable, and dependent on the events of the day.” By narrowing the focus to feelings and their fluctuation over time, the study was able to produce a clear understanding into a very complicated topic.
Research Incorporating Modern Lingo (Emoji’s)
The bot was programmed to ask the questions and allow for an option if the user wanted to share more information about the emoji they chose. When a new participant signed up, they would click Get Started and Feels the bot was able to help them with the on-boarding process. If someone needed help they could click for help with an option of commonly asked questions. The programmers also planned for trolls or unhelpful responses. Should respondent’s type profane language into the comments section Feels would show an emoji of disapproval.
The response to the study was very high with a third of respondents participating every day. An impressive 85% of users stayed subscribed through the whole study. The study showed that during key times of the election there was some large emotional changes in the participants. The example given was when Hillary Clinton’s emails were brought up again in late October. This event had users going from to ? to ?.
What is interesting about this study was that the Washington Post recognized that the needs of the readers are changing, and that these needs are shifting to a growing desire to be connected. While individuals see phone calls to survey political opinions is becoming more intrusive, it does not mean that people do not want to share their opinion. It seems that now more than ever individuals want to share thoughts, feelings and ideas but in a way that they deem appropriate, and on their time schedule. Whether you are a news agency or a consumer goods company we believe that market research online communities can be a great way to interact with your customers.
If you would like to talk more, feel free to contact us.
Follow us on Twitter!
Check us out on LinkedIn!