When social platforms appeared circa 2004 (Facebook) and 2006 (Twitter) no one really understood the impact either would make some 10+ years later.
What has become synonymous within social media is the association to sentiment analysis.
Sentiment has been sitting around within academic institutions gathering dust for decades, and it would be fair to say that an academic approach to building sentiment models was more advanced using machine learning & NLP back then than what is commercially on offer today.
Sure, positive, negative, neutral, emotions are all components of measuring sentiment and these are norms within the industry today that social platform providers truly believe is an acceptable way of providing real insight into consumer preferences.
We can talk about combing this with entity extraction, topics and themes but does this really constitute understanding the consumer and the commercial impact to a business or brand without having factual metrics that convert to $$$$$$?
Having seen first hand how many companies claim to derive sales uplift and monetization strategies from sentiment, this has only ever been like finding a needle in a haystack and clutching at straws; whatever way you cut it, this just doesn’t work and still requires considerable manual intervention at the platform level. Time is money… right?
When we talk about the evolvement of social media and social data, notably in the last 10+ years this has remained static and enterprises have finally woken up to the fact that senior execs and procurement professionals want more than a nice UI, likes, sentiment, themes and so on.
The true value of the companies marketing and selling platforms with minimal features is being questioned like never before and sentiment analysis is at the heart of it. So much so that M&A is rife within the social media sector and many companies have fallen by the way through a lack of vision as to where the market is heading.
Quite clearly, sentiment has a place but struggles to derive commercial traction.
What is missing and where is the market heading?
There are a few companies (including ourselves) in the market that are using machine learning and NLP to derive meaning and monetization from social data that truly impact a consumer journey. Whether that be segmenting audiences or understanding a buyer’s journey from A to Z, we share the fundamental same goal – to validate to enterprise clients and big brands that building commerce products in days, across industries and languages is achievable.
Whether you wish to offer data monetization through a platform, API or embedded technology offering the opportunities are exponential if you have a proven commerce product suite for monetizing data at scale and, more importantly, a process, method and strategy for staying ahead of the curve.
IBM Watson, Microsoft Azure, SAP Hana, Marketing, Data & Commerce Clouds are all vying to offer a one stop shop for all our data needs.
The reality is the cost of data and access to data is being squeezed, the democratization and pricing being offered by the leading players is proving difficult for existing platform vendors to stay alive let alone compete with.
Fundamentally, if you do not possess strong academic pedigree and proven machine learning & NLP capabilities with a commerce focus then you will be unable to compete in a market where speed, scale and accuracy will decide who wins and loses.
In simple terms the Text Analytics market today stands at $6B, Marketing Automation is at $10B, but the real opportunity lies within the Commerce market which is $100B and growing faster than any other market.
Sentiment analysis will forever be on offer, but building commerce offerings will change the game.