Twitter influence and the new SEO...

If there are two things modern comms technologies have taught us, it is the power and danger of herd behaviour and that if something can be manipulated, it will be manipulated.

Despite the enormity of the dotcom boom and bust and the even greater enormity of the credit crunch, both of which were driven by technology fueled herd behaviour, we continue to rush headlong into the 'next big thing' - until it comes crashing down around our ears again.

Likewise, the massive growth of Google over the last decade has led to the equally massive growth of an industry whose sole purpose is to manipulate Google results.

On social media, the alarm bells are starting to ring for me and I'm getting a sense of 'unrealness' that I got working in the midst of the web and telecoms industry at the turn of the century and in watching the 'financial services' bubbles growing ever larger in more recent years.

Two experiences in the last week have added to that sense of unrealness. The first is playing about with Klout and the second is looking deeper at emerging areas of 'sentiment' monitoring.  Twitter influence and social media sentiment are on my radar in terms of web strategy but, as always, I'm keen to get beyond the assumption and hype often associated with such things and not waste either my time or my employer's time and money unnecessarily.

So, if automated tools such as Klout are to be believed, I managed to raise my 'influence' score by 8 points in less than a week. How? well essentially by using the same triggers that have always been used in these type of environments, since the early newsgroup days .

1. Controversy

2. Audaciousness (nice word - thanks Simon ;) )

3. Flaming

Maybe it was a little more subtle than the average troll, maybe not, but in essence it was these 3 approaches that gained the '@' and RT responses that Klout is clearly using as a primary measure of influence in its algorithms. Is Twitter a more mature environment impervious to techniques employed for years in other online social environments? - I think not. Do I genuinely feel I've gained any influence during the last week - errr no - but Klout thinks so.

If, however, the consensus is that something like Klout is an accurate representation of influence on Twitter then this raises another key question - how representative is Twitter of offline influence and is there a danger of getting fixated with something which in the bigger, wider world currently has very little influence?

Beyond Twitter, I think I'm seeing the rise of a new SEO - Sentiment Engineering & Optimisation - that, like bees to a honeypot, is already attracting the naive evangelists,  'snake oil' salespeople, confusion marketers and cowboy operators.

Anyone who has spent any real time in web analytics and has had responsibility for reporting results knows how easy it is to misinterpret data and how taking numbers on face value can be a dangerous thing, without trying to understand the deeper background to that data.

If it's easy to misinterpret and misrepresent structured data, I'm thinking it's even easier to misinterpret and misrepresent 'unstructured' information.

There was a tweet from one attendee at the recent Alterian customer day that I felt was fundamental to this " It's hard for a computer to comprehend sarcasm and irony." Like it or loath it, our brave new social media environment is loaded with sarcasm and irony and as this article illustrates, it is a complex form of communication. Has 'natural language processing' advanced to a point where it can interpret a comment correctly that actually has the opposite meaning? If it has then I suggest we are much closer to achieving artificial intelligence than perhaps I thought. Add into this, the 'ugly side' of social media, very well described by Econsultancy here, and there is a whole potent mix of manipulated, distorted and amplified information that potentially has to be dealt with here.

With its immense programming capability, Google has a continual struggle to prevent its search results being manipulated (I'm guessing a lot more than 200 signals now) and judging by the numbers of individuals and organisations jumping onto the current SEO bandwagon there is still clearly money to made in trying to do it. So, realistically, how close are we to even beginning to monitor and understand the infinitely more complex realm of social media sentiment in a useful and accurate way? And do solutions such as Klout indicate that a whole new Sentiment Engineering & Optimisation (SEO) industry is about to explode? Sadly, I think they do :(