I remember some years ago a University lecturer warning myself and a lecture theatre full of bleary eyed students about the perils of carrying out academic research via search engines and not putting in the leg work of good old fashioned reading. She said we would be “panning for gold in a river of mud*”. And of course she was right.

Recently I’ve found myself muttering this analogy as I’ve been researching and using several social media monitoring/research tools.  There’s no denying they can show you all manner of information, which is great, but ultimately can leave you scratching your head and thinking “so what?”. ‘Sentiment’ is a particular bee in my bonnet. As these tools are keyword driven this metric is prone to being less than accurate. IE, a post containing the terms “poor” and “rubbish” would be shown as negative sentiment, but the post could itself read:

“Previously the service I received from brand X was poor, but not since they’ve moved their call centre back to the UK. Frankly they make their competitors look rubbish.”

Clunky example I know, but I see little point in metrics that are so easily flawed.

Now, I don’t want to sound negative, because there are some great findings to be had – showing the media channels where a clients market and brand are being mentioned are definitely of use.  Channelling this down into a list of sites and resources is even better – but remember the results are driven by keyword content, so there is still scope for error.

So, they do have their place, but currently should be seen as a starting point and are not the be all and end all. To make the most if these tools you need to know what you are looking for and what you want to achieve, but be prepared to be surprised and frustrated by what you find and allow time to refine your results and then refine them some more. Otherwise you’ll simply end up drowning in said river of “mud*”.

*She didn’t actually use the word “mud”, something slightly ruder, but the point is the same.