This week, several mobile carriers rolled out updates to older Android phones. The Android 2.2.x (Froyo) operating system has been out since May of 2010. Release 2.2.2 has been on a rolling release since January 2011 and is just now being pushed out to older phones, like my Samsung Galaxy Continuum, by Verizon Wireless and T-Mobile. For me, it has been a disaster with about 20% to 25% of my apps no longer working. How many apps are "not working" depends on whether you include random crashes or only immediate crashes. I apparently have gotten off lightly. Some customers are reporting problems where the phone app itself dies, turning the smartphone into a dumb brick.
Sooner or later, the vendors involved - Google, handset makers, and mobile carriers - will want to know who customers are blaming for this. Despite the eerie silence from all three, I can't imagine they don't care. The sheer volume of noise on the social media and social networking outlets must have gotten their attention. But what are customers really saying? Who do they blame for the problems they are experiencing? This is where sentiment analysis comes in.
Sentiment analysis, also called opinion mining, is the application of data analytics to social media. Software ingests huge streams of de-indentified data from consumer services such as Twitter, Facebook, LinkedIn and others. Using big data technology, sophisticated textual and linguistic processing run on these giant data sets to try and determine what customers are saying about a subject. In this case, sentiment analysis can tell vendors whether the noise they are hearing is only from a few grumblers like me or is widespread. It may even be able to discern who is getting more blame and how deep the discontent is.
There has been a proliferation of companies that produce sentiment analysis solutions, either standalone or as part of a suite of social meeting management tools. Companies like Visible Technologies, Attensity, Lexalytics, Crimson Hexagon, and SocialMatica produce sophisticated platforms for helping to understand what customers think of a company or product. Sentiment analysis promises to free marketers from being slaves to a few, loud people by understanding what everyone else is saying.
Now, about those crashing apps...