Delivering Insights That Matter Through Social Media - Consumer Insights Generation Part II

My last post ended with talking about data slices that are relevant to the consumer segment you are trying to detail through some social media measurement. 

So lets talk about what you can do with social media data.   For starters we need to admit a couple of things:

• Profile data is flawed and out of date
• Content is how need states are addressed NOT profile

So what do I mean by those two statement.    Well profile data suffers from a time capsule issue.  Who I was when I filled it out is NOT who I am now.    Models like quantcast have proven over time that the information that is proxied to create inference models is also often flawed.   More so it is highly skewed by a census model of collection of data through standard behavioral information.  If we are gathering data on one site and not the other we are corrupting the profiles through an unbalanced census.   Blahblahblah I digress.

So what do we know?   Well someone who reads content surrounding buying a car and researching that car is highly likely to be interested in buying a car.   This is kind of implicit.   We also know that if we leverage the social media data that is out there to examine how people are referencing or liking that content we can bring it to a point of understanding which articles are in fact the bits of content that really resonate with the target and their need state to research a specific car model.   Ah...Now we are onto something.    However NONE of this is available in the standard social media measurement tools.   I have used them all when I was agency side and none of them can get this right.  

So the challenge in this becomes the following:

Find the right audience through basic demographic profile analysis.
Query them to determine the behavioral economic need states that occur during the purchase path
Query them to determine what about the brand/product is important

Now comes the role of social.  Find the speech acts in social media that represent a hand raiser for one of the behavioral economic need states or one of the brand/product value propositions or philosophies and you can begin from there.

Anyone that tells ya otherwise is selling corrupt research I tell you!  And let me tell you there are a ton of tech suppliers selling "insights" that are no more then corrupt insights that can skew decisions based on invalid censuses of the target audience.   Rawr.  Not good.

I will continue in my next post with how to analyze this content.