Saturday, May 03, 2008

The word on the web: 7 keyword trending tools

Unquestionably one of the most powerful ways in which products and services get promoted, word-of-mouth is not only notoriously difficult to generate; it's also very hard to measure. Pre-digital, finding out how 'talked about' your brand was meant arming yourself (or, more likely, a costly market research agency) with a pencil and clip-board and trying to find a representative sample to quiz, either by phone, mail or face-to-face. The arrival of email made contacting a large number of potential respondents much cheaper, quicker and easier but still relies on self-selecting individuals and only captures claimed, rather than actual, behaviours.

And then Google happened. For the first time, a small but significant slice of the world's interactions were being indexed and made searchable. The first tools to mine this data were somewhat limited in scope; Google Zeitgeist (launched in 2001) presented a small selection of top ten lists and charts of popular search queries, which tantalised the stats geeks amongst us with what could be discerned if open access to the database was granted. We had to wait five years, but in May 2006, Google did exactly that when it took the wraps off Google Trends, which enables users to chart trends for the search terms of their choosing.

Whilst knowing what keywords people are searching for is useful (and an important success measure in its own right), it doesn't necessarily directly correlate to how much your brand is being talked about. Fortunately, a new breed of products is emerging which focus on tracking keyword usage on blogs and in other community spaces. Icerocket's Trend Tool, Trendpedia, Technorati charts and Nielsen's BlogPulse Trend Search all attempt to trend word usage in the blogosphere, whilst the recently launched Facebook Lexicon collates keyword data from people's Facebook Walls and Twist charts keyword recurrence on Twitter.

The potential applications of these keyword trending tools are already myriad and my suspicion is that they are just the tip of the iceberg. Obvious next steps include mashing up the existing data sources to provide both aggregate and comparative trends across the various forums/services (e.g. Facebook users talk more about X than Y, whereas Twitterers talk more about Y then X) and beginning to contextualise the mentions to make more qualitative assessments (e.g. X % of keyword mentions were in a positive context, Y % were negative).

Below is a sample chart for each of the services I've mentioned, illustrating some of the interesting possibilities for this data in different market sectors.

Google Trends

BlogPulse Trend Search

Technorati charts

Technorati Chart

Icerocket Trend Tool


Facebook Lexicon


No comments: