A few tips on ‘social media analysis’
I find a lot of presentations and articles about how to do any type of social media analysis read like a series of abstractions designed to create the appearance of knowledge. The only details being what you shouldn’t do (after everyone has long agreed you shouldn’t do them).
I rarely come away with any useful tips or advice that I can apply to my work.
An opaque approach is favoured by agencies pushing quasi-scientific methodologies; such as those fantastically creative mathematical formulas to predict future voting patterns based on an unrepresentative sample of the population.
As a relatively new methodology, it can also be because those presenting at conferences rarely do (or have done little of) the research themselves. They only concern themselves with the hypothetical.
As a result, it’s actually easier to make what you’re saying sound clever, rather than simplify it so it can be understood by anyone.
It also helps get your unqualified statements, epigrams and unsourced statistics spread via the echo chamber of Twitter. Particularly when the audience has a collective interest in not questioning those facts.
Jargon serves the purpose of acting as a useful ‘handwave cliche‘ in which listeners or readers rush to fill the gap with their own understanding of what a phrase means. ‘Big data’ is arguably the most in vogue and loaded of terms at present.
As such, when I was asked to write a ‘how to do social media measurement’ guide for B2B Marketing, I wanted, within the parameters of the word count and ’10 point list’, to make it as practical as possible.
I’d argue the two most important points here are to start every project with a clear objective and finish it by interpreting (and presenting) the findings in context.
It’s certainly not a definitive guide to doing social media analysis, but hopefully you’ll find something of use:
Is there anything you’d add to these?