One would assume there is some sort of predictive power in Twitter’s firehose, given that Google, Microsoft and a bunch of stealth startups seem willing to pay for special access to the data stream.    But my take, based on Twitter traffic and trends as they stand today, is that insights gleaned from Twitter actually LAG more conventional online media.    Here’s why.

First here’s an outline of my experiment.  I used some of the third-party tools that exist to analyze Twitter data, as well as Twitter’s own search and tagging (hashtags) functions to look for tweets about company reputations.   I did this by searching Twitter on combinations of company names/brands and the reputation tags we use at Vanno (social responsibility, customer service, human rights, value and price, etc.)  for which we had good data on Vanno.     The first question I asked was whether the Twitter posts about company reputation were original, or did they contain a link that referenced an online article.    When relevant tweets contained links, I compared the referenced articles to the ones on Vanno for that company/reputation aspect combination.    Finally, I looked at how the Twitter community as a whole reacted to the topic – e.g.  if and when more tweets on the topic came in, and how many were re-tweeted.   I’d be happy to share the details with anyone who is interested.

Here’s what I found.   Most tweets that were relevant to company reputations contained links to stories from a rather small number of newspapers, blogs and other online sources.   Not surprisingly (at least to me), they were generally the same ones that Vanno users (and just about everyone else one the Web) referenced when discussing that particular topic (company/aspect of reputation combination).   There were, of course, some original company reputation-related tweets, but they were predominately real-time customer service complaints.   As far as I could tell,  the Twitter take on a given company reputation topic consisted mostly of independent references to a relatively small number of  links, and re-tweets that contained those links.    Some topics got more tweets and retweets than others, of course, just like some Digg stories get more votes than others.   But the key point  is that you would have learned just as much about the company reputation topic, and learned it earlier, if you had just scanned the RSS feeds from a couple of dozen sources (or looked on Vanno, of course).   This leads me to conclude that Twitter is in fact a LAGGING indicator when it comes to what the online world thinks of companies and their reputations.

I must say that these results make sense to me conceptually, as I view Twitter as being more about ”mobsourcing” than anything else.    The recent focus on re-tweeting only amplifies the mob-like nature of Twitter.   There must, of course, be instigators lurking somewhere in the Twitter stream for the mobs to get started.    And if you follow the tweet stream backward you’ll find those users and can follow them.  But it turns out that the insights themselves are rarely unique to Twitter.   They are either obvious observations or something that has been published somewhere online already.   

As a final point of possible confirmation, I’d note the following.   Though I drew this conclusion for a pretty narrow subject realm - companies and their reputations - it seems that others are also starting to recognize the lagging nature of  Twitter insights when it comes to broader social and political trends.