Sentiment analysis and the challenges of content overload

Last week, Socialbakers came to town with the London stop on their Socialbakers World Tour.

The social media analytics company invited me to the London event to hear from CEO Jan Rezab presenting analysis on brand behaviours on Facebook, including the top brands and their ‘returns on engagement’; and introducing new product features in Socialbakers’ offering, notably BrandLove.

Socialbakers describes BrandLove as “a new research method we have designed that measures the sentiment consumers feel towards brands.” That intrigues me as I believe that only humans, not computer software or algorithms, can accurately measure sentiment: what people feel about something.

But maybe such a view is getting less robust these days when you hear about what IBM is doing with its Security Intelligence with Big Data tools that can analyse people’s emails and assess the behavioural characteristics of the person writing those emails; and defence contractor Raytheon with its eerily-named RIOT product that analyses and predicts people’s behaviours from their activities across the social web.

In any case, I spent some hours in the evening of February 20 in pleasant surroundings – including enjoying a really good dinner – along with 60 or so others who gathered to hear what Jan Rezab had to say, as well as hear insights from the experiences of keynote speakers Dave Parkinson of Nissan and Phil Mitchelson of News International.

One aspect in particular has had me thinking since last week – the challenges of content overload. Look at this slide from Rezab’s deck that compares the number of brand pages the average Facebook user interacts with in some way (‘like,’ for instance) from 2011 to 2013:

What users cope with on Facebook

It illustrates the huge growth in volume of brand content that Facebook users are dealing with today, a factor that you could easily extend across the social web, not only Facebook.

It focuses your attention on getting fans’ attention in a very noisy world, where so much is competing for that attention. It’s not enough to just get attention: you have to offer far more than that to meet people’s evolving expectations (otherwise, you may well end up in the deluge of crap content as envisaged by Doug Kessler).

And of course, you must be able to measure all of that, analyse it and interpret meaning in order to make effective business decisions.

Behavioural data analysis and interpretation is undoubtedly a growing business area, evidenced perhaps by Socialbakers’ growth in this past year including opening more offices, hiring more people and securing more investment funding.

Socialbakers says they count more than a fifth of the Fortune 500 as clients, who use the company’s services for competitive intelligence and social media benchmarking, measurement, statistics, and brand performance analysis.

I was pretty impressed by what I heard in London last week, and the confident outlook I saw from every ‘social baker’ I met. Even about what they say BrandLove can do.

As Dave from Nissan noted (with a smile), “sentiment analysis isn’t for software to do. Yet.”

Related post:

Neville Hobson

Social Strategist, Communicator, Writer, and Podcaster with a curiosity for tech and how people use it. Believer in an Internet for everyone. Early adopter (and leaver) and experimenter with social media. Occasional test pilot of shiny new objects. Avid tea drinker.

Comments are closed.
Close