Posts Tagged ‘text analysis’

Social Media Sentiment: H8ers and

Monday, August 15th, 2011

H8ers! <3ers! Isn’t social media just full of people who have radical opinions? It’s been a while since we first shared information about the distribution of opinions/sentiment in social media so we thought it was about time we conducted our little experiment again.

For six different sets of data, we gathered hundreds of thousands of sentiment scores and prepared frequency distributions of the results. As you can see below, some brands have more positive (A, B) sentiment while others have more negative (C, E, F) sentiment. You can also see that some brands have more flat (E) or peeked (C) distributions, or longer tails (A, B). No matter which particular feature of a brand’s chart interests you, it is clear that all of the distributions are reasonably normal, they are generally bell shaped.

So is social media full of haters and lovers? Most definitely not. Most social media data consists of lots of moderate like and dislike, plus a healthy representation of haters and lovers. Now the only puzzle is determining which of these charts reflects the sentiment of 1) autism, 2) Lady Gaga, 3) Obama, 4) Steve Jobs, 5) Toyota, and 6) Walmart.


.
.
Related links
Are Only Crazy People Commenting About Brands in Social Media?
Article in the Vue: Words I’ll Live to Regret
Cell + Survey + SMR: A Social Media Mashup #MRIA2011 #MRA_AC #MRX
There is no question but the research validity question

.

Upcoming ESOMAR 3D presentation: Tell me what you want, what you really, really want

Tuesday, July 26th, 2011

ESOMAR 3D 2011
3D DIGITAL DIMENSIONS 2011
(ONLINE + SOCIAL MEDIA + MOBILE) RESEARCH
MIAMI / 26 – 28 OCTOBER

.
TELL ME WHAT YOU WANT, WHAT YOU REALLY, REALLY WANT
CREATING DESIRED RESULTS FROM SOCIAL MEDIA RESEARCH

Annie Pettit, Conversition and Research Now, Canada
.
This presentation will teach you how to generate the social media research results you desire regardless of what the true results are. I will demonstrate how to gather social media data from the internet using inappropriate sampling methods, and how to select the wrong pieces of data and code it incorrectly. The topics of sampling, weighting, data quality, sentiment analysis, and text analysis will be highlighted so that you can understand the full range of options for mistreating data. The ultimate goal will be to create set of data that reflects our predispositions towards a topic as opposed to reality.

Attendees are required to come prepared with a sense of humour  (i.e., I will be speaking in jest!)

Social Media Sentiment: H8ers and <3ers

Friday, July 8th, 2011

H8ers! <3ers! Isn’t social media just full of people who have radical opinions? It’s been a while since we first shared information about the distribution of opinions/sentiment in social media so we thought it was about time we conducted our little experiment again.

For six different sets of data, we gathered hundreds of thousands of sentiment scores and prepared frequency distributions of the results. As you can see below, some brands have more positive (A, B) sentiment while others have more negative (C, E, F) sentiment. You can also see that some brands have more flat (E) or peeked (C) distributions, or longer tails (A, B). No matter which particular feature of a brand’s chart interests you, it is clear that all of the distributions are reasonably normal, they are generally bell shaped.

So is social media full of haters and lovers? Most definitely not. Most social media data consists of lots of moderate like and dislike, plus a healthy representation of haters and lovers. Now the only puzzle is determining which of these charts reflects the sentiment of 1) autism, 2) Lady Gaga, 3) Obama, 4) Steve Jobs, 5) Toyota, and 6) Walmart.


.
.
Related links
Are Only Crazy People Commenting About Brands in Social Media?
Article in the Vue: Words I’ll Live to Regret
Cell + Survey + SMR: A Social Media Mashup #MRIA2011 #MRA_AC #MRX
There is no question but the research validity question

.

Dear Brand, you are never to talk to me even if I talk to you first: The Social Media Puzzle

Tuesday, July 5th, 2011

Ok, so in the last couple of posts, we agreed that not everyone feels comfortable about brands responding to comments they make online. Sometimes, people want to make a comment and then just be left alone. But when is it acceptable for brands to respond?

We conducted an online survey (based on the e-Rewards survey panel) to determine how people feel about companies responding to them about a comment they made in the social media space. We surveyed a census representative sample of 1000 Americans and 1000 Canadians to find out in which situations they thought it was appropriate for a company to respond to them about comments they’ve written in social media.

First of all, it’s interesting to see that there are no situations where there is 100% (or even 90% or 80%) agreement that it is appropriate for a brand to respond. In fact, even when people write a comment on a brand’s website, only about 54% of people think it’s acceptable for the brand to respond to them. And, it doesn’t matter if they casually mention a brand or tweet directly to them, about a quarter of people indicated “it is never appropriate for a company to respond to you about comments you have written in social media.” Wow!

So who are these people who think it is never appropriate for a company to respond to them? Perhaps it is some odd, demographically unique group of people? No. It’s very slightly more men, slightly fewer younger people, slightly more uneducated people, and slightly more people without children. Just remember, when you generalize to say that most people are ok when a brand responds to their comments in social media, you’re ignoring the preferences of a lot of people. 26% of people.

Commenting in social media means you want a response, right? Wrong!

Thursday, June 30th, 2011

If you talk to people about social media engagement, you’ll hear a common thread – people want to be responded to when they make a comment about a brand or company in social media. People like it, they expect it, it should happen. This always makes me wonder about people who make comments online but don’t necessarily want to chat with the company. Do such people exist?

We conducted an online survey (based on the e-Rewards survey panel) to determine how people feel about companies responding to them about a comment they made in the social media space. From a census representative sample of 1000 Americans and 1000 Canadians, we identified 152 people who said a company had responded to them when they made a comment online.

Given a multiple choice question listing a number of positive, negative, and neutral feelings, we asked survey responders to select as many items as appropriate to describe how they felt (which means these numbers will add up to more than 100%).

About 41% of people said that they liked being responded to and about 40% of people said they appreciated being responded to. That is a nice, healthy, positive number. However, about 10% of people were annoyed and about 10% felt that they were being stalked. Would you be comfortable annoying 1 out of every 10 people you talked to?

Now, if about 41% of people liked or appreciated it when the company responded to them, that left a lot of people who did NOT necessarily like or appreciate the response. Perhaps they didn’t care one way or the other, or they actually disliked it. Either way, they did not feel the need to indicate that they liked the response. What are the demographics of the group of people who didn’t like the interaction?


.
So now let’s think again about the generally agreed upon idea that people want to be responded to when they write something about a brand online. Is that really true? I don’t think so.

.
.
Related links
Are Only Crazy People Commenting About Brands in Social Media?
There is no question but the research validity question
Cell + Survey + SMR: A Social Media Mashup

.

6CFBPK6DVC9T

Are Only Crazy People Commenting About Brands in Social Media?

Monday, June 27th, 2011

We know who uses social media to comment about brands. They are crazy people who have nothing better to do with their time.  Really? Is this still true?

We recently conducted an online survey (based on the e-Rewards survey panel) to determine who participates in social media. We talked to a census representative sample of 1000 Americans and 1000 Canadians and asked people about their social media usage.

Specifically, we asked them “Over the last month or so, have you written any comments or questions about a brand or company in social media?” I’m sure a lot of survey gurus will have fun pointing out alternate wordings of that question which would have generated far more accurate and precise results, but let’s take the answers for what they are: An indication of who is using social media to talk about brands.

22%. Let that number sink in.
.
22% of people have knowingly written something about a brand in the social media space. Add to that even more people who don’t realize they occasionally chat about brands and we’re working with a lot of people. It’s not just men, it’s not just educated people, it’s not just people with kids, it’s not just employed people. The only demographic that skews a bit away from average are older folks, but even a good percentage of them share their voices online.

So what kinds of people comment about brands in the social media space? All kinds of people.
.
.
Related links
Social media monitoring vs social media research: Can you see the difference?
The Conversition Hierarchy of Social Media Insight
Observational Research – The Original Research Method

Article in the Vue: Words I’ll Live to Regret

Friday, June 10th, 2011

This article was originally published in the Vue, June 2011.  When people answer research surveys, they don’t sign their names or write their email address. This degree of anonymity gives them the power to share their true feelings with as little pressure as possible. But what happens in the social media space? Read along to find out. It can be pretty scary!
.

.
.
Related links
Social media monitoring vs social media research: Can you see the difference?
The Conversition Hierarchy of Social Media Insight
RW Connect: Privacy and Ethics in Social Media Research #MRX
ESOMAR Launches Consultation on Social Media Research Guidelines

openview: e-Rewards acquires Conversition

Wednesday, May 18th, 2011

e-Rewards acquires Conversition: A Marriage of Social Media and Market Research

Tuesday, May 17, 2011 by Faria Rahman
.
Last week, e-Rewards acquired social media market research agency Conversition. Conversition applies scientific principles to the collection and analysis of social media data. Its solutions include TweetFeel, MatterMeter and EvoListen. Its product EvoListen, which is currently under development, collects data from online social media outlets; cleans, filters and weighs it, and then formats the information into quantitative data sets. This will allow e-Rewards’ clients to listen and analyze what customers are saying on social media.
.
Read the full article here.

There is no question but the research validity question

Tuesday, May 17th, 2011

As a budding field, social media research is a magnet for questions from both experts and novices alike. People are curious about the processes and methodologies used to accomplish the various aspects of the research. Some of the more common questions we field on a regular basis are as follows:

  • What sentiment analysis system do you use?
  • How do you carry out the text analysis process?
  • What is your method for identifying and eliminating spam?

In fact, each of these questions is one and the same. They have nothing to do with sentiment, text analysis, or spam.  They have nothing to do with processes or methods or systems. In fact, they have everything to do with validity.

Validity refers to truth. Is the sentiment scored accurately? Is the text analyzed accurately? Is the spam identified accurately? Is the entire process valid? Among all the pieces of the puzzle, this is the one question that must be answered.

Unfortunately, there is no single method that automatically identifies a sentiment analysis, text analysis, or spam detection system as being the most valid one. You simply have to evaluate a large, representative sample of data and determine the answer for yourself.  Are your results valid?

.

Also see
The Sharks and Icebergs of Social Media Measurement
5 Ways to Fool An Automated Sentiment Analysis System
A Formula for Perfect Sentiment Analysis

e-Rewards, Inc. Announces Agreement to Acquire Conversition Strategies

Tuesday, May 10th, 2011

Plano, TX – May 10, 2011 – e-Rewards, Inc., a global leader in digital data collection and reporting, today announced it has reached an agreement to acquire Conversition Strategies.

Founded in February 2009 by former IPSOS executives, Conversition is a pioneer in the Social Media Research industry, and was formed to capitalize on the emerging trend to perform market research within the social media channel. The primary product offered by Conversition is EvoListen®. A patent-pending technology, EvoListen collects data sourced from online social media outlets, cleans and filters the data collected, utilizes scientific sampling and weighting to report the results, and then formats it into quantitative data sets.

“With over half a billion consumers worldwide engaging in social media platforms today, we believe social media listening will continue to become an increasingly important way for businesses to collect valuable insights around the needs and wants of their customers,” said Chris Havemann, President and Chief Executive Officer of e-Rewards. “Our acquisition of Conversition will allow us to further accelerate our social media strategy while enhancing the research capabilities we offer. We remain committed to equipping our clients with innovative products that help with the collection and analysis of digital data.”

Commenting on the announcement, Tessie Ting, Co-founder of Conversition said, “We are thrilled about combining the benefits of more traditional online survey research with those of social media research. Through this acquisition, e-Rewards will continue to strengthen its position as the leading global digital data collection provider with an expertise in social media research.”

.
About e-Rewards, Inc.

e-Rewards, Inc. is a global leader in permission based digital data collection and reporting. With over 1000 employees worldwide, it operates a strong portfolio of brands including Research Now, e-Miles, and Peanut Labs. For more information, visit www.e-rewardsinc.com.

Press Contact:
Ashley Harlan
VP, Corporate Communications
+1 214-365-5000
aharlan@e-rewards.com