I remember growing up learning the lesson: “If you don’t have anything nice to say, don’t say anything at all.” While paying closer attention to social media, I’ve had to question whether this lesson has any place online. The Internet has created this separation between words and their impact and unintentionally offending people and deliberately maintaining disagreements and arguments (ie. ‘trolls’).
A problem that we’ve encountered is that words are not always used as they are defined. There is a gap between their intended use and how they’re actually used. This is something that in the early stages of the beta testing period, Orenda had trouble distinguishing. However, this has since been overcome with our study of language patterns.
Orenda searches online sources and stores every post mentioning a particular subject. The post is then determined to be relevant or non-relevant to the brand through an automated process. Only the relevant data is organized into specific categories, and then scored. The way Orenda determines the categorization of a post comes down to the words and how they’re used. Depending on how a word is used, Orenda will interpret the message as a particular factor of the client’s reputation such as trust, satisfaction, social responsibility and character.
We’ve basically taught Orenda to define different words and phrases in different social and cultural context, in order to overcome the ambiguity of words on their own. For our clients, the learning process is important, because certain words and phrases can be interpreted as either positive or negative, that vary across industries. For example, the word “freezing” can be negative for computers, but positive for a freezer. At the end of the refining process, our clients see the nuances that define their brand among online audiences.
The key to Orenda’s success is being up to date on language patterns, which requires the ongoing monitoring, testing, and defining of words. The New York Times has also utilized this method in a recent article written by Patrick Healey and Maggie Haberman. The article discusses Republican Candidate Donald Trump’s language patterns, derived from an analysis of approximately 95, 000 words taken from Mr. Trump’s speeches, interviews and other public statements over a recent five month period.
The results of this analysis have found that Mr. Trump’s use of divisive phrases such as “we” versus “them” has been successful in rallying his loyal supporters. However, he is unsuccessful in winning over other potential supporters for the same reasons.
According to this study, we have learned that focusing on one category of reputation limits us to the audience we can reach. Trump’s language has prevented people from focusing on the solution to his mentioned issues, and highlights the negative aspects of other candidates’ platforms as well as the efforts of the Obama administration. This shows the importance of language and the impact words can have on how a brand is judged.
Obviously we can gain insight into brands through the language used, both by the intended message and how the audience processes these words. Even online, its not so much what you say, but how you say it.
Here at Orenda, we’re always looking at language patterns, trying to find the real meaning behind words and the context they’re used in. Thankfully our software alerts us of patterns or trends, making this process efficient. This type of insight will help you see the blind spots in your brand, the effectiveness of your words, what appeals to your supporters and gives you a better appreciation of your critics’ commentary. So if you don’t have anything nice to say, try to choose your words more carefully so that your intended message is effectively communicated.