How to Take Action with Social Insights — Free Webcast ›
Popular Today in Social: All Popular Articles

Measuring Social Media Sentiment without Keywords & Smiley Faces

Measuring Social Media Sentiment without Keywords & Smiley Faces image Social Media Brand SentimentAccording to Social Media Examiner in its 2012 State of the Social Media Marketing Industry report, “How do I measure the effect of social media marketing on my business” was the most asked question by more than 3,800 marketers. This has never been a hard question to answer here in the Kuno labs. It’s quite simple really; measure the trend line of website traffic, conversions and customers from social sources while paying attention to the appropriate ratios. Where the challenge lies for many is measuring overall brand sentiment and trust.

The Social CRM

There are many enterprise-level social media monitoring tools available that try to measure social media sentiment. They achieve this by tracking loosely defined negative, positive or neutral keyword phrases in social media posts.

Not Ready for Primetime

Most of the software allows community managers to tighten the filter by creating and defining their own keyword lists. Even with the filter, many posts are improperly labeled because of the nuances of human language. For those posts the community manager is expected to click on a smiley or frown face to signify the correct sentiment. The end result is an untrustworthy brand sentiment grade and hours of invested time.

The Social Sentiment “Klout Score”

Spending time chasing a social media sentiment “Klout score” is a waste of time because it is not accurate and never will be. Sentiment is not an individual number, but a trend over time reflected in action (or lack thereof) metrics and not measured with keywords or smiley faces. Sentiment itself is intangible and can only truly be measured when it becomes tangible through an action. Any good community manager that’s actively listening can communicate a sense of sentiment, but relying on an intangible number is misleading.

What happens if the the smiley face score is trending down, but website traffic, conversions and new customers are trending up? Is there really a brand sentiment and trust problem?

Tried & True Sentiment Analysis

This intangible approach to measuring brand sentiment and trust is not necessary. It’s much easier and more accurate to examine real tangible web analytics over time. By no means does this alleviate the community manager’s responsibility for listening, engaging and handling negative comments, but it does save dozens of hours from smiley face clicking and keyword filter management.

Measuring Social Media Sentiment without Keywords & Smiley Faces image Macro Social Media Sentiment Analysis

Generally, all of the non-paid sources of traffic trend lines can reflect brand sentiment and trust. However, pay close attention to social and referral sources for this analysis.

Positive Measurement

Positive brand trust and sentiment is reflected when website conversion rates increase with or without an increase in website traffic. However, an increase in website traffic can reflect the effectiveness of a community manager over time (as shown above). Ultimately though, if there appears to be a positive correlation between the number of people visiting a website and the percentage of them converting, you can be confident that sentiment is positive.

Neutral Measurement

Neutral sentiment can be identified when there appears to be no correlation between traffic and conversion rates.

Negative Measurement

On the other hand, negative sentiment can be shown when a negative correlation between traffic and conversion rates is identified. Also, if traffic remains consistent from social media and referral sources, but conversion rates continue to drop, negative sentiment may be the culprit.

It is true that many other factors can play a role in the correlation between traffic and conversions, but for the seasoned inbound marketing manager ruling out other factors should be easy. So rather than spending all of your time clicking on smiley faces and managing keyword filters, you can use this new-found time creating more content and building community online. For more help with social media, watch this expert panel with our guests Jay Baer, Laura “Pistachio” Fitton and Jim Kukral.

Image: tumbleweedmarketinganalytics

Allocate Your Marketing Resources Accurately

Measuring Social Media Sentiment without Keywords & Smiley Faces image budgetcalculator blog

RSS Comments Feed

Comments on this Article: 1

Add a Comment
  1. Good post. Sentiment Analysis is a extremely challenging for most tools out there. No matter how sophisticated the algorithms, nothing can replace a human being interpretation. No amount of smiley faces or scoring systems.
    Some tools like Actionly have a learning algorithm that if you edit sentiments, over time; the algorithm can start picking it up. But I guess for now social managers and analysts need to review sentiments in addition to social monitoring tools.

Add a Comment:


Thank you for adding to the conversation!

Our comments are moderated. Your comment may not appear immediately.