New: Sentiment & Emotive Analysis on ACTIVATE
Applying Machine Learning to Measure Consideration, Intent, Social Proof and Skepticism in Influencer-Driven Conversations
When people tell stories about your brand on social media, the entire world is invited to participate. As audiences react and comment to media, a narrative begins to unfold. Understanding the sentiment of your audience empowers you to shape that narrative as the story evolves.
Introducing Sentiment Analysis
ACTIVATE is excited to announce a new suite of Sentiment Analysis tools to help you track and manage audience sentiment in your influencer marketing collaborations.
“As you work with influencers at scale, your brand is talking to many separate audiences, all of which are having different conversations. By monitoring the sentiment generated by different influencers, you can understand which influencers are telling the right kind of stories and deepen those relationships.
We have been internally testing this with a number of partners over the last few months and are excited about the level of accuracy when it comes to detecting emotion as well as the ability to train the technology to better detect new expressions of emotions”
— Nicholas DiSanto, CTO of ACTIVATE
We aggregate and summarize the sentiment generated by each post, by each influencer, and overall on each collaboration. These summaries give you clear visibility into they reactions of your audiences to different influencers and content. You can also drill down to see how we are analyzing each individual comment.
A Spectrum of Emotions
Our sentiment analysis feature can analyze the sentiment of Instagram, YouTube, and Facebook content to provide deeper, more holistic insights surrounding each campaign’s success. When it came to choosing emotions to focus on, we wanted to focus on topics, categories, and emotions that were deeper than the usual binary emotions of “positive” and “negative”. We also require that the campaign has a minimum of 25 comments for analysis before results are displayed.
“We’ve found from hundreds of influencer activations that sentiment is not just positive or negative. In analyzing the conversations around influencers’ content, we find that the emotive points are much more nuanced and we wanted to incorporate as such into our technology. We help you detect not just positivity, but love or consideration and intent, and not just negativity, but subtler concepts like fear or skepticism. We continue to train the technology through machine learning to better understand comments expressed on social media to map it to up to 8 emotive axis”
— Lauren McGrath, Vice President, Head of ACTIVATE Studio
We currently score each comment on 8 different axes: Positivity, Negativity, Fear, Hate, Love, Skepticism, Damage, and Intent & Consideration.
Based on our data, most sentiment is positive, but at times the narrative can turn in a direction that you might want to steer back on track. Occasionally, one bad actor might derail an entire conversation. Being able to detect this in real time and react to stay in control is key to protecting your brand.
Every time we detect a possibly negative comment, we send a notification to the campaign manager and the influencer with the analysis of the comment, it’s text, and a link to the comment on its corresponding social platform.
Measuring Sentiment in a Recent Global Influencer Program
A global pasta brand & ACTIVATE collaborated with 31 food and lifestyle influencers from around the world to promote and draw attention to a large scale event in Italy. The brand was interested in how the audience of the creators responded to the content, so our team ran sentiment analysis on all 300+ audience comments. The results demonstrated just how much the audience loved the pasta content that they engaged with, ultimately allowing the client to gain a better understanding of the audience’s sentiment towards their product and brand.
31 Influencers, 300+ comments analyzed:
- 50% of posts had 100% “positive” sentiment
- 14.2% of posts displayed feelings of “intent & consideration”
- 77.4% average “positivity” rate across all posts
- 5.4% of posts demonstrated “love”
With these new Sentiment Analysis tools, ACTIVATE continues to unlock more insights than ever before, positioning your brand for success in the ever changing social media landscape.
See this feature in action today!
ACTIVATE is a fully end-to-end influencer marketing platform, covering influencer discovery, program workflow, measurement, and analytics. Our ACTIVATE Studio team also helps brands build out and execute their influencer strategy. Last year, the ACTIVATE platform enabled over 75,000 influencers to collaborate with brands.
Want to learn more about Bloglovin’ or ACTIVATE? Just drop us a note at research@activate.social.
