Discourse analysis, sentiment analysis, text analytics:  whatever you call it, you’re looking at a huge volume of information.  Unstructured information.

When CRM started in the 1990s, the challenge was to link customers to behaviors and make recommendations to keep them happy/buying more.  Think of the Amazon “recommendation” feature:  “people who looked at this, bought this” or “if you liked this, may we recommend…”  They were able to “mine” purchase data to make better shopper suggestions.

Text mining is infinitely more complicated.  The longer a blog post, Facebook comment, or product review, the more difficult it is to categorize.  Twitter posts and other microblogs are theoretically easiest because they’re so short.

How It’s Done.  Comments are filtered based on keywords (e.g., brand name).  Next, as I understand it, using proprietary algorithms (depending on the software), comments are grouped into three buckets:  positive, negative, and neutral.

The longer the post, the more likely the comment is to be classified as neutral.  This isn’t surprising.  Think of the complexity in nuance, sarcasm, slang, humor and differences in word usage across different cultures speaking the same language.  (Anyone watch the Emmy’s Sunday night and see Ricky Gervais?  He was laughing at Bucky Gunts’ name and it’s taken me 10 minutes to figure out why – so bad I can’t repeat it here!)

I’ve seen statistics showing that between 65-80% of all comments are tagged as neutral.  The number is somewhat lower among the microblogs – maybe only 40%.  Nevertheless, it’s obvious that huge biases can be introduced using this approach.

Elsewhere I’ve read that humans aren’t very accurate either.  Using Mechanical Turk (which isn’t something I’d personally recommend), humans agree only 79% of time on how items should be coded.

We should be moving forward in this discipline. Tom Anderson, of Anderson Analytics, agrees.  In a correspondence with him several months back, Tom said that he was using three different software packages and gained confidence based on the intersection their joint findings.  I’m not sure if more data is successfully coded or if this just provides more confidence in what he reports as findings to his clients.  I think this is very responsible, given the accuracy of single-solution approaches today.

Currently, I’m leaning in the direction of relying more on Market Research Online Communities (or Insight Communities), which straddle the space between pure discourse analysis and more structured qualitative or quantitative marketing research.  I view them as a hybrid, where I can get out of the way and follow the discussion.  With only 200-500 members, keeping up and staying accurate is more manageable.  Further, when custom research is needed, the recruit is a snap.

And sometimes we don’t need to build communities.  One may already exist because the product or person built their following online.  In this case, the fan group may be the “go to” community because it’s larger, more representative and more cost effective than creating a new community.  Think Justin Bieber, who was “discovered” by his YouTube videos.  Those subscribing to his stream, particularly in the early days, would have been a great insight-mining resource.

I’ll be keeping a close eye on both these topics as they appeal to both my quant and qual sides.  And if you have any thoughts, I’d love to hear and learn!

Some reading:

The best of the articles I read, BrandSavant:  http://brandsavant.com/the-hidden-bias-of-social-media-sentiment-analysis/

August 2010 Quirks has several articles on social media research

Mashable:  http://mashable.com/2010/04/19/sentiment-analysis/

No wonder marketers are confused about how to use social media effectively!  The juxtaposition of the two article titles made me laugh.  (Actually, the FeverBee.com piece is rather good, offering some creative pointers on how to engage community members.)

Staying with the subject of online communities, this month’s “Quirk’s” piece on the 2009 Globalpark Market Research Software Survey reported that only 17% of companies worldwide manage at least one online community and that 56% have no plans for starting one.  (A bit more detail:  60% of small and medium companies say they have no plans, with 41% of large companies responding the same way).

Bottom Line:  From their actions, most marketers seem aware of the a fine line between staying attuned to customers and invading their privacy.  Sure, an online community is a major investment of both time and money.  But there are other ramifications that marketers are having to consider which may be slowing down this new research method’s adoption.

A project last week reminded me once again how nice it is to include even smaller markets in the mix.  I drove to Bakersfield from Los Angeles, only an hour and a half drive.  (Bakersfield is much smaller than Kansas City, with slightly less than 400,000 in the urban area.)  The client picked the market as it’s a good market for them.  There isn’t a focus group facility in the area and there is no longer a recruiter based in the market.  But I got great respondents (using my LA recruiter) who didn’t have prior research experience, who showed up on time (a good incentive strategy worked!), and who really helped to give us the needed insights.

I had a similar experience many years ago when I had a client who loved the Atlanta market, yet after three projects in a row where we had “dud” groups, we both decided to drive up to Chattanooga, TN.  The facility was in someone’s house.  Again, the respondents were great, candid, and while not as sophisticated, they gave us a perspective that we hadn’t heard before.  (And what a lovely city!)

Online qualitative research does afford us the opportunity to include more of these participants in the mix, but a lot of the smaller markets may still experience some problems with Internet connection speeds.

BTW, I’m not recommending that small markets always be included in market selection – that depends on the client and the target.  However, if you’re going to three or more markets, you might want to consider one of the smaller gems to add perspective to your exploratory undertaking.

What’s different – and noteworthy – is the way technology is being leveraged.  Two companies come to mind, yet I’m sure there are others:

iModerate – Fields a larger quant study and then pulls off certain participants to have a one-on-one follow up with a professional moderator.  As their site says:  Research>iMpact provides a hybrid approach in which online qualitative sessions are incorporated into virtually any online survey, giving you qualitative insight that enhances and clarifies the quantitative data.

Invoke Solutions – Seemingly more quantitative in orientation, Invoke is leveraging their technology to obtain richer and deeper open-ended responses.  They say that it’s like a focus group experience, but I’m going to need to learn more before I’m convinced.  Having said that, I’ve seen enough that I’m intrigued and I want to learn more.  The client-interface portion looks very powerful.

But are either of these “new”?  I started this topic thinking, “No, they’re not really new at all.  Just a new twist on an old approach.”

But the more I think about this, the more I believe these might be new approaches.  The dividing line for me lies in the definition of “qualitative research.”

• Qualitative research explores.
• Great open-ended responses in quantitative research explains.

For now, I’m going to continue to think of these newer approaches in these ways – and make client recommendations accordingly.  And, as always, I’ll continue to seek out more info.  If you’ve got something to share on this subject, please do!

One could argue that everything on the left side of this diagram is anthropologic.  However, I contend that “discourse analysis” is so reliant on words that we must acknowledge techniques that rely equally much or more on physical observation.

The three types of approaches utilized in anthropologic market research which most immediately come to mind are:

Ethnography.  From the Greek “ethos” (the character of a people and their distinctive culture) and writing.  In short, to observe and document.

Semiotics.  The original Greek meant “interpreter of signs.”  Today, in business applications, it involves studying communication processes, signs and symbols.  A British colleague pointed out the two main influencers of this discipline:  Peirce and Saussure.  (For a quick read, see Wiki).  To me, semiotics is about cues and norms.  For example, we tend to think of banks as “institutions” with strong edifices or tall buildings; we want them strong.  “The friendly bank” sounds slightly oxymoronic.  Clearly, it’s more complicated than this – and I hope others will weigh in!

Ergonomics.  For the sake of brevity, this is about how people interact with things, either physically, mentally, or organizationally.  Human factors and usability testing fall into this category.

The second thing to keep in mind about this quadrant, as with “sociologic” research, is that we can study people individually or in groups – a major consideration in study design.

This quadrant has been “trending” for the past decade or more.  With the aid of technology, we’ve been pushing into the lower left of this area.  There are dozens of new innovations and techniques, but the two that have stuck with me are the following:

Wireless cameras for individual ethnography and usability studies.  People can create, with limited direction from the moderator/facilitator, mini-documentaries.  Want to go along with a man when he€™s trying to pick out a gift for his wife?  You can do that.  Want to see and hear how a woman works with a new appliance she just bought?  You can do that too.  I really don’t believe this is a gimmick or fad.  While some might think a “con” of this method is that people will only show or discuss what they want, the clear upside is the natural setting.   My experience:  I’ve never done an in-home study where the bed wasn’t made, but I have done online sessions and seen unmade beds and messy rooms.  So, like any other method, the pros and cons need to be assessed.  Overall, this is a terrific advancement.

CCTV for merchandising.  Several years ago I did interviews with executives in the corporate security business.  One of the men I interviewed told me that his closed-circuit TVs were attracting the interest of major retailers; the software they had developed, for use at airports, looked for things that weren’t moving (e.g., bags with bombs).  What the retailers wanted to know is if the software could be modified to assess what I call “merchandising wear-out.”  That is, a display is put at the front of the store.  Could the camera/software tell 1) if the display had “stopping power”  (to get people to enter the store and/or look at the display) and 2) if the traffic flow patterns indicated it was time to change/modify the display?  A bit creepy, perhaps – but realistically we know there are security cameras in stores.  Net, net:  this is a clear example of when the absence of behavior can represent critical learning.

What are your favorite new methods or tools in this area?