Skip to Content

Sorry, it’s been a while and things have been busy around here!

There has been quite a bit of buzz lately about Voice of the Customer analysis — taking text that customers have posted on blogs like this one, on forums, as well as text entered in customer surveys, and transforming that text into actionable information.

Voice of the Customer analysis falls into two main categories — buzz (external) analysis, which is provided by service companies like Omnicom and Newstin (both of which are Business Objects OEM partners) as well as general analysis, which is provided by players like Attensity, Clarabridge, and Business Objects. (Of course, we think Business Objects Text Analysis, with its truly extensible platform and multi-lingual capabilities, is the perfect choice, especially for existing SAP or Business Objects XI customers!)

Voice of the Customer analysis can tell you not only what your customers have done, but what they would like you to do in the future – shedding light on desireable new product features and service issues, helping you better understand what really matters to your customers, and exploring that information in a multi-variate environment.

Do your customers in Florida value different things than your customers in Berlin? Do you have a particular issue with the French version of your software that doesn’t exist in other versions? Is your service desk in Belarus causing more problems than your desk in Botswana? Did that new change in policy help improve customer satisfaction over the last three months, or are customers angrier than ever? These are all issues that Voice of the Customer analysis can help answer.

SAP itself is looking in the future to better mine comments on sites like this one by using BusinessObjects Text Analysis, as well as use Text Analysis to recommend additional content of interest to you.

We are also working on adapters to SAP CRM systems that will allow you to seamlessly plug in BusinessObjects Text Analysis to your SAP CRM installation to begin mining and using information in that system.

To report this post you need to login first.

3 Comments

You must be Logged on to comment or reply to a post.

  1. Richard Hirsch
    Hi,

    Interesting stuff. I have a few questions.

    1) How do you view “tag clouds” based on user-provided tags. Are these more subjective whereas text-analysis is more objective?
    2) An interesting correlation would be between what users view (“lurkers”) and where they comment / collaborate? Can text-analysis work in this area as well?
    3) Can you describe how the tools can determine the “emotionality” of contributed content?
    4) You mention CRM. What about using other types of back-end systems (SRM, PLM, etc.) as well?
    5) Can you place users into certain categories based on their content? Creating social networks automatically based on what they contribute?

    Look forward to your answers.

    Thanks.

    Dick

    (0) 
    1. Anonymous
      1) How do you view “tag clouds” based on user-provided tags. Are these more subjective whereas text-analysis is more objective?

      You have hit the nail on the head. Without a controlled vocabulary, user-provided tags are a challenge (one person’s “dog” is another person’s “canine”). In addition, even with a controlled tag vocabulary like that on SDN, opinions differ. I might tag my article “advanced analytics” while others might argue “advanced analytics” only refers to predictive modeling, etc. I think both user-generated information as well as automatic tagging are useful.

      2) An interesting correlation would be between what users view (“lurkers”) and where they comment / collaborate? Can text-analysis work in this area as well?

      Yes, great idea!

      3) Can you describe how the tools can determine the “emotionality” of contributed content?

      Again, it’s based on linguistic patterns. Emotionality is tricky; for example even humans have trouble with written sarcasm, for example: “Well, THAT was a GREAT idea.” So a lot of work needs to be done there. Some cues are simple (a sentence ending in “!!!!” is clearly emotional, as is the phrase “I HATE this”.) So you start with the base and refine as you go.

      4) You mention CRM. What about using other types of back-end systems (SRM, PLM, etc.) as well?

      Yes, anything with freeform text comments is a perfect candidate.

      5) Can you place users into certain categories based on their content? Creating social networks automatically based on what they contribute?

      This would be a great application of the technology!

      (0) 
  2. Richard Hirsch
    Hi,

    Thought you might want to take a look at the Community Project “Enterprise Social Messaging Experiment” (ESME – The Demo) which is being submitted for this TechEd DemoJams. It is an enterprise messaging system ala twitter but with functionality that is relevant for the enterprise environment.

    There are at least two functionalities -creation of a tag cloud based on message content and dynamic network creation based on message content – that appears to be a perfect match for “Voice of the Customer” functionality.

    Could you take a look at the screencam and see if Inxight might be a good fit. Another option appears to be TREX.

    Really curious to hear your comments. If you wish, you can email directly. I tried to access your email but couldn’t find it.

    Thanks.

    Dick

    (0) 

Leave a Reply