By Justin Tenuto, February 17, 2015

Register for our Webinar: Make People-Powered Sentiment Analysis Work for You


Every brand wants to know what its customers actually think about it. Are they enjoying the new soda flavor they just released? Does their online customer service actually make their customers happier? Did they like the Super Bowl commercial where that dad cried a lot?


To uncover these opinions, innovative companies and PR agencies do something called “sentiment analysis” or “opinion mining.” Put simply, that just means analyzing language, text, and conversations for the attitudes around a given topic.

On Thursday, February 26th, at 10 a.m. Pacific, we’ll be joined Edelman, the world’s largest public relations firm, to show you how they use CrowdFlower to understand the sentiment around their brands and why they prefer people-powered sentiment analysis to strictly machine-based approaches. The webinar is free and you can register here.

What makes people-powered sentiment analysis different?

Most sentiment analysis tools are natural language processing (or NLPs). They search for keywords or strings of words and bucket those into positive and negative based on fairly complicated linguistic algorithms. What makes a people-powered approach different is that the conversations (or tweets, comments, posts, etc.) are distributed to fluent speakers who can analyze the data quickly but with a higher degree of accuracy.

Why is that? Well, the simple answer is that language is complex. Algorithms and sentiment analysis tools traditionally have a tough time with linguistic modes of the disinterested barista: things like sarcasm or irony. For example, a tweet that says “good job!” might be a compliment on work well done. Or it could also be a snide remark about an especially idiotic behavior.

Contronyms, emojis, bad jokes, images used to convey feeling, and slang are the sort of things humans just “get.” It’s hard to train a machine to understand what an image of two people high-fiving is, let alone understand the context of that picture. People can do that, though. We know that sometimes “sick” means “good” or that sarcasm can be used to convey a compliment or an insult. Having a fluent crowd of millions looking through text means that you won’t be just hunting for keywords: you’ll actually be understanding what’s being talked about.

What you’ll learn at the webinar

If you’d like to learn how people-powered sentiment analysis works, this free webinar is a great place to learn. Edelman will show you how to:

  • Locate your biggest advocates (and critics)
  • Understand the real opinion about your brand
  • Analyze thousands of conversations in hours
  • Determine if a recent product launch was successful
  • Leverage the power of the crowd to get actionable insights
  • Take your brand’s temperature, year after year

…and plenty more. The webinar is free and we’d love it if you joined us. There will be a Q&A at the end with both Edelman and CrowdFlower CEO Lukas Biewald. Here’s a brief introduction to the speakers. We hope to see you there.


Meet the Speakers


Christopher Lightner, Executive Vice President, Measurement and Insights at Edelman Digital

Chris focuses on establishing measurement guidelines that help Edelman’s clients clearly define measureable success from social media efforts that align back to their business objectives. He enjoys cooking, spending time with friends and family and, of course, rooting for Bay Area sports teams.


Jocelyn Swift, Senior Account Executive, Measurement and Insights at Edelman Digital

Jocelyn is a member of Edelman’s Los Angeles team, specializing in Insights and Measurement. She provides analysis of digital and social media including ROI, KPI forecasting, and real-time reporting. In her free time, Jocelyn loves hiking and camping but is just as happy sitting at home playing video games with her fiancé Blake.

Lukas Biewald | CrowdFlower

Lukas Biewald, Founder and CEO at CrowdFlower

Lukas is the CEO at CrowdFlower. Following his graduation from Stanford University with a B.S. in Mathematics and an M.S. in Computer Science, Lukas led the Search Relevance Team for Yahoo! Japan. He then worked as a senior data scientist on the Ranking and Management Team at Powerset, Inc., acquired by Microsoft in 2008.