Social Media Monitoring Grows Up: Analytics for Big Boys


Share on LinkedIn

I love some of the really cool names out there like Filtrbox, Buzzstream, Viral Heat, Synthesio, Mediamiser, and Inuda Innovations, to name just a few of my favorites.  One wiki [] lists 99 vendors linking themselves to this space, with seemingly new ones popping up every day.

With all these companies and all this buzz, you would think there were billions of dollars in the social media monitoring space.  Truth is, there are only a few companies that are actually making money.  But based on conversations I’ve had with customers, even A-list vendors struggle to deliver value beyond the PR and Community teams, which not surprisingly, keeps the average price point for social media monitoring vendors down. It also keeps the analytical value down. So what’s the deal?

My belief is that there are 3 problems with tools in this space but that each can be overcome by taking a slightly different approach:

On my way to class I went to Burger Place to meet Sally and have a Big Burger.  I can’t believe how expensive her new outfit was compared to mine.



On my way to class I went to Burger Place to meet Sally and have a Big Burger.  I can’t believe how expensive a Big Burger is compared to a Whopping Burger.” 


Both sentences have similar words, but the 2nd provides real value to Burger Place, while the first doesn’t.   Social Media Analysis tools need to automatically remove SPAM, filter content, and classify not just the entire post, but also the specific issues, experiences and opinions within a post into buckets or categories that make business sense.


  • Problem #2: A focus on one channel creates biased market understanding.  Social media content is real-time, high volume and great, but what about internal sources, like call center notes, emails, and surveys?  What about branded websites like  Combining those other sources with social media feedback provides richer customer perspectives. 


General Information
Specific Insight


 On my way to class I went to Burger Place to meet Sally and have a Big Burger. I can’t believe how expensive a Big Burger is compared to a Whopping Burger.





I just ate at your store on Michigan and Main, and you are too overpriced for such a pathetically sized meal


Every source has limitations and benefits.  How timely is the information?  Is it biased?  Is it targeted and specific?  Can you tie feedback to specific customer transactions or demographics?  How fast will the information spread?  Social Media analysis tools should integrate all these diverse channels of solicited and unsolicited feedback as well as integrate with other systems and data to provide a more holistic view.


  • Problem #3: Existing solutions have limited analytical value beyond the social media team.  Right now, functionality in monitoring tools serve a single business audience: social media teams. Other business functions like product marketing,  customer support, and operations need to go beyond surface-level analysis into the nature of the content itself. 

In the example above, Burger Place needs to understand not just that 1,356 people mentioned Big Burger’s last week.  The more interesting thing is to know that 10% of those conversations involved discussion of Big Burger prices and that 75% of those were negative on price.  And that nearly all of those comments were from college students.  Chatter about price has increased 200% since the new college-focused advertising campaign.  


This deeper analysis would turn into an action item for the owner of Burger Place’s college segment.  My example is fictitious, but Clarabridge customers confirm their need for details to serve more than one business audience: What are the specific issues and praise?, how is it changing over time?, how does that compare across various channels/demographics/product lines?, what’s the root cause of a particular issue?, etc.   An enterprise needs social media analysis that integrates sophisticated natural language processing, classification, and sentiment engines to truly understand the context and syntax of what is being said and distribute this information to product, marketing, support, or operations folks enterprise-wide.
Clarabridge looks to expand our presence in the social media world, and I look forward to hearing other challenges and opinions on how they can be overcome.  Let me know what you think.

Republished with author's permission from original post.

Tony Lopresti
As Vice President of Sales and Marketing for Clarabridge, Tony is responsible for the integration of our clients' feedback into our product offerings, marketing activities and business development plans. Tony is an entrepreneurial executive with progressive experience in enterprise software sales, product management, marketing, strategy, and operations.


  1. Hi Tony,

    Thanks for your insight into the subject; you raise some interesting points here. One of the interesting points of differentiation between social media monitoring vendors is the mix of automated vs. manual service to the customer. As you pointed out, many conversations online are not worth the time spent reading them. A post written by an influential blogger, however, could become significant for a company quite quickly. We address this by analyzing influencers at post level and at site level, and also by analyzing the inbound and outbound links of each site, their audience, the volume and frequency of their publications, the type of site, and their Google PageRank.

    Your second point brings up a couple of points, actually: integration of internal and external data and the creation of niche monitoring. We customize our dashboards to the specific industries that wish to monitor conversations, but internal information would be an interesting next step.

    As for your final point, there is certainly a need for human involvement in social media monitoring, analysis, and engagement, be it a member of a PR/communications agency or a marketing director at any given company. Nathan Gilliatt actually wrote some great articles about this need for human involvement after the automation to make sure that the noise is filtered out, and that the data can provide valuable insights.

    I would love to talk more with you on the subject: michelle(.)chmielewski(@)synthesio(.)fr


  2. Not only can’t you trust the specificty and validity of the metrics from vendors, the metrics related to the market are not trustworthy.

    That Pear Analytics study about “pointless babble” is a great example. It’s pointless babble to a corporate marketer who wants business information. It’s meaningful communication to the individuals involved. The Pear article is an example of the low quality of information in this space. Even the methods are laughable.

    Some good points above, but you still have to go back to the source of the information: why did this individual type this particular information? Understanding this provides the proper context to analyze the text. Automating that analysis is very hard to do.


Please use comments to add value to the discussion. Maximum one link to an educational blog post or article. We will NOT PUBLISH brief comments like "good post," comments that mainly promote links, or comments with links to companies, products, or services.

Please enter your comment!
Please enter your name here