Etuma - Affordable Text Analytics

Like a lot of people who help businesses with customer feedback, these days I'm frequently being asked about Text Analytics.  Sentiment analysis in particular seems to be the buzzword du jour. 
Clearly, the growth of social media as a customer connecting technology has made text analytics a more important technology for categorizing and assessing large volumes of free text feedback.  More importantly, as more and more companies actively solicit feedback via e-mail and web surveys more open text comments need to be evaluated.  Text analytics, of course, is very valuable in this regard as well.  This is particularly true if your survey tools can't filter responses on the basis of sentiment based answers (as in net promoter surveys for instance).

As usual, in emerging markets and technologies, the leading vendors are typically quite expensive and often underperform when objectively compared to solutions from less well known vendors.  A situation that would appear to be occurring again with text analytics.

Having said that, I think I've found a text analytics vendor that is both reasonably priced and very effective at processing text into usable insights.  The company is Etuma.  They are from Finland - and impressively - can assess textual feedback in 10 different languages with 90%+ accuracy and categorize that feedback seemlessly into a single report in a single language (very useful for US companies doing business in Europe).  They also do sentiment analysis at a more granular level than the leading text analytics vendors. 

To my mind Etuma offers really great tool for customer support organizations that operate worldwide, or for US companies looking to do a better job assessing their foreign sourced (non english language) customer feedback.

With 10 languages "out of the box", the ability to pull all feedback into a single report and granular sentiment analysis, all in an affordable tool, Etuma would seem to be a vendor that lots of North American companies would want to talk to.

Check them out at http://www.etuma.com/

Customer Feedback and the "Dialogue Dilemma"

Just published my first post on QuestBack's "Friends of Feedback" blog site.  If you're interested you can check it out at:

http://www.friendsoffeedback.com/customer-feedback-and-the-dialogue-dilemma/

An excerpt follows:

Generally speaking, businesses seek customer feedback for purposes of identifying customer issues and taking actions that improve their satisfaction and loyalty. Yet, the majority of solicited customer feedback isn’t acted upon. And worse, there are indications that if feedback isn’t acted upon, negative customer experiences result. An exact opposite effect to the one intended. I call it the “Dialogue Dilemma”.


There's a bunch other good information on customer feedback management on the F-of-F blog too....

My 7 "Rules" For B2B Customer Surveys - updated with some expert input

I was recently surfing the web and found myself at http://www.netpromoter.com/.  So, visited their Blog and found a post by Satmetrix CEO Richard Owen which talked about my Rule #7 for customer surveys and discussing the effects on him of a survey he took that failed to incorporate known information (thus violating the rule).  Rule #7 states: "Never ask a customer a question you already have the answer to." Richard did a great job documenting why this is great rule for customer surveys.  So, I thought I'd share his post here.
http://www.netpromoter.com/netpromoter_community/blogs/richard_and_laura/2011/04/16/hello-old-friend-who-are-you-again

For anyone who want to read my original post.  It's here on the site.

New EFM Blog from QuestBack

QuestBack has a cool new Blog site called: "Friends of Feedback".  The name is a little hokey, but it has pretty good content designed for people who use or plan to use enterprise feedback management technology in their businesses. 

The content, so far anyway, is very much focused on value add of feedback, tips for doing feedback correctly, etc. Not a site that "flogs" QuestBack. 

I encourage people to have a look...

www.friendsoffeedback.com

Feedback vs Research - Thoughts on doing both better using EFM

Customer Feedback and Customer research are two phrases that used to mean the same thing but now are differentiated because EFM technology allows dialogue to be spawned from electronic surveys.  But, like many change agents, users of EFM have created some new business problems by their use of the technology.  With customer research, virtually anyone today can freely and easily overburden their customers with survey after survey.  All delivered over the internet or via e-mail.  So, there is a new business problem: Negative Customer Experiences-stemming from survey reponses that are not followed up.

I've been thinking about how companies can manage around the divergent goals of customer research and customer feedback so that negative customer experience does not result.  I'm not sure I've got great answers to the problem, but thought I'd jot down some ideas.

Fundamentally, a customer feedback survey is distinct from a research survey in that its feedback solicitation is an invitation to dialogue.  Whereas, a research survey is an information gathering exercise.

Most companies' customer survey processes share three characteristics:

• Both types of surveys target many of the same people
• The same internal entity (market research) often develops and administers the surveys 
• And the same survey tools are often used

It's not surprising to me, therefore that feedback and research efforts would "look and feel" a lot alike to the customers being surveyed.  As a result, I think there is a significant amount of confusion amongst customers as to when their feedback will produce a dialogue and when it won't.  My guess is this confusion is at the heart of the negative experiences issue.

Furthermore, research shows that even today most customer feedback doesn't result in a dialogue.  And, survey participation rates are generally declining, except for companies who go to great pains to "close-the-loop" on customer feedback. 

"Fixing" the problem is, of course, easy.  Just always follow up on any survey response given by any customer.  Needless to say, that "solution" is probably impractical for most companies.  With most companies not even able to consider reponding directly to all the feedback they solicit.  I think though, that with the right tools and processes they could easily respond to a lot more of it than they do now.

So, what I think companies need to do is create a set of "Customer Survey "rules of engagement".   Rules that would govern how they should approach doing both customer feedback and customer research surveys.  My ideas about what those rules ought to be are outlined below:

Customer Survey “Rules of Engagement”

For Customer feedback:

• Try to integrate surveys with CRM, ERP, HRM and other systems where practicable. This can work especially well for short transaction based surveys, which are relatively easy to integrate.  Just be sure to impose rules on the users of those systems to pursue follow up on all feedback received.

• For customer relationship surveys (usually done once or twice per year) create a follow up plan for all customers and response scenarios that you can reasonably think of.
  - Make sure your survey system generates an “alert” or “notification” to a specific person for each foresee-able response scenario.
  - Any feedback that doesn’t “fit” a response scenario should be directed to the most senior available company officer.
  - This ensures an action plan will get created for almost everything that can fit into a response bucket.
  - Keep in mind that response scenarios will need to involve people other than the sales or customer support staff. Ensure those other groups understand the obligation to customers to respond to feedback alerts that will be generated.

• Relationship surveys tend to be difficult to integrate with other systems, as they tend to change over time and often blend-in research questions. So these surveys should only be done in systems that provide follow up mechanisms that are independent of CRM, ERP or HRM systems.

• Keep in mind that if you are soliciting information from all of your customers, or all customers within a segment, you’re probably seeking feedback, not doing research. So plan for a dialogue.

For Customer Research:

• Try to use anonymous surveys for research. Preferably deliver them to a sample subset of your customers or customer segments. People don’t expect follow up on anonymous surveys and don’t feel guilty about abandoning them in mid-stream.

• Try to employ “research panels” containing customer volunteers. People who you have a profile for, who know that surveys may not be followed up and don’t mind acting as research fodder.

• Label research surveys as RESEARCH. Avoid using the word “FEEDBACK” which co notates dialogue. Be sure to tell research survey participants that no follow up is planned. And, that the research won’t result in a dialogue.
  - This may hurt response rates
  - And, may force you to incentivize your research surveys

• Apply strict contact rules for all research oriented surveying. My guess is that no customer should receive more than one research survey in a given year

These are ideas and thoughts.  I'd welcome feedback from readers on other "rules of engagement" with customers being surveyed.