Not too long ago, I undertook an analysis - using Radian6 - to determine the social media sentiment toward a bunch of Australian celebrities.  An old-fashioned survey (remember the ones where we used to go out and ask people questions face to face, or get them to tick boxes on sheets of paper?) was conducted by an external agency on the same celebrities.  

The findings, presented to a room full of individuals from a melange of industries (including, but not limited to, entertainment, advertising, media, PR and marketing), were met with a mixture of wonder, fear, and dismay.  

Wonder:  wow, it's amazing how much you can learn about someone on the internet!

Fear:  will what people say about me on Twitter  affect others' perceptions of who I am?

Dismay:  why was my score so low?  I'm actually lovely - how can I make people realise that?

It made me stop and think - what would I learn if I did an analysis on myself?  I was intrigued.  No rest would be had until I had satisfied my curiousity.  I logged into Radian6, configured a profile on TweetsMcG (my alter-ego), took the red pill, and dove headlong into the social rabbit-hole.

In no time at all, I was confronted by this: my Conversation Cloud.  The sum of my keywords.


When I beheld this nebulous cloud of ME, I understood all too well the Wonder, Dismay and Fear.  You can infer so much about me from my collection of keywords.

I tweet a lot (tweetsmcg).  Many people either retweet my posts, or directly mention me in theirs (@tweetsmcg).  These people include @mrgareth, @snoutley, @laceysnr and @sullmcintyre. There are several products about which I often converse.  It could be supposed that I even use them:  @Salesforce, @Radian6, @Sonos, @Spotify.  Or perhaps I work for one of them?  I am interested in music (@Sonos, @Spotify); quite possibly watch too much television (@getglue); and sometimes interact with questionable entities (@toosaasy). 

The few hashtags you can see indicate I may have attended Salesforce's #Cloudforce event, as well as the Mumbrella #m360 event.  Search the hashtags on Twitter and you can see what was actually posted!  

My conversation cloud does say a lot about me.  But it does not, and cannot ever, tell you who I am.  For all you know, I'm a strange hypo individual who stays up all night tweeting or trolling, simply to work out the aggro.  I could be a fake persona, flexing my creative muscles online.  In these instances, any data you collect about me can be misleading and/or completely fictitious.  

That may be the case, however, there are lots of positive words in that cloud (awesome, love, great, yay) - I appear to be an optimistic person, rather than a troll.  And too many of you know my online persona is fairly representative of whom I purport to be, to be a fake account.  How do you know this?  Because I post regularly, and engage with you online.  Maybe you should check my Klout score to see how influential I am?  ;)

Last night I read a great post from the Harvard Business Review HBR Blog Network - What Data Can't Tell You About Customers.  One comment particularly struck me:

To innovate for a future in which consumers' desires and habits change as quickly as their mobile devices, businesses must be nimble in delivering emotional connections beyond just functional utility. That requires understanding customers as people — nuanced, dynamic, unpredictable — not just collections of data.

This comment rather perfectly encapsulates the point of my post.  It is so easy to take someone at (inter)face value.  Analysing our target markets using data collected from social media is an amazing and powerful way to try to understand, or get to know, our existing and potential customers.  Engaging with our online audience humanises our brand.  It shows our customers we are interested in who they are and what they have to say.  It provides us with potential opportunities to delight them in some way.  

But I do think we should bear in mind that we are not seeing the whole story in our data clouds.  We are more than the sum of our keywords - it might just delight our customers more if they knew we understood that.