I love when online retailers allow for customer reviews. For me, details from consumers like me are way more important than any product details or photographs a vendor can provide. Very often, my purchase decisions will be swayed based on positive or negative product comments. Customer feedback is useful for customers and to organizations — but historically, there has been a disconnect between collecting customer feedback and actually acting upon it. Social media presents a whole new challenge customer feedback initiatives. Metadata, or unstructured data, is difficult for traditional analytics systems to evaluate or measure. And often, when social media data is analyzed, it’s on a statistical basis — and doesn’t get to the heart of what customers are saying. Leximancer, a young text analytics company, is hoping to remedy that situation with its Social Insight Portal product. Leximancer approaches metadata in a different way, and according to Neil Hartley, CEO of Leximancer, seeks the why behind customer sentiments. Hartley ventures to say that he knows of no other companies doing what Leximancer is with insights from social data. He does note, however, that the possibilities are tremendous. “Not only is there more direct feedback on micro- and social sites, but it’s a leading indicator of customer feedback,” he propones. “By the time feedback reaches [an organization] through the traditional method, it’s way too late.” 

The company describes its product’s purpose on its Web site:

Through a rigorous scientific process, Leximancer drills into textual data: documents, e-mails, call center transcripts, blogs, Web sites, etc., and extracts the main concepts, themes and causal relationships to provide the information needed to make critical decisions.

The Concept Map is the heart of Leximancer. It presents your data in an interactive graphic map that clusters key themes together, lists all the key concepts identified in the data and the statistical co-occurrence of concepts. Leximancer automatically generates a thesaurus for each concept that then enables a powerful, implicit search (i.e. searching the corpus under investigation for the intersection of two thesauri rather than simply two keywords). The pathway analysis capability identifies root causes of specific events and provides provenance via the original text from documents that the pathway references.

Hartley says Leximancer’s methodology is scientific and mathematic as opposed to traditional text analytics methology, which is more grammatical. What’s complex to explain actually boils down to a simple, intuitive, and visual product. “The technology used to analyze what peope are saying has to be massively easy to use,” Hartley says.  Below is the Leximancer Concept Map based on social insight gleaned from Coffee Bean & Tea Leaf reviews on Yelp.com. It’s easy to see what main concepts were addressed in the reviews. Hartley said a more interesting and useful example was done for Dunkin Donuts. When analyzing a review site, the concept map showed “Boston” and “line” in great proximity, demonstrating that people were complaining about long lines in Boston Dunkin Donuts stores. Hartley says all of this is automated and could add great value to a CRM system — or even to a social media platform provider. The company is currently exploring OEM partnerships and basically exploring its options in the vast social media landscape. 

Leximancer Social Insight Graph for Coffee Bean & Tea Leaf