Almost a year ago, I met with the President of TrendIQ, Paul Feldman. For some reason, I keep thinking back to my meetings with him. His company is in an extremely interesting space.

TrendIQ provides customized business intelligence that tracks trends based on publicly available information. Roughly speaking, the company searches the internet on topics of interest, analyzes the search results, and scores the aggregated results across time. The end results are extremely interesting, data rich charts and graphs that answer a wide range of interesting questions such as:

• Are peoples’ opinion of our CEO improving or worsening?
• Do people like or dislike my product or service?
• When people are talking about the product or service I provide, how frequently are they talking about me?

An example will make this clearer:

TrendIQ has tracked the last several presidential campaigns. The graphic above shows a portion of a heat map chart for the 2008 Presidential Campaign. The full results can be found here. The candidates are scored in 3 areas:

Campaign Strength measures . . . how much material is on the Internet that discusses the [person] as a candidate for President in 2008. Campaign Effectiveness answers the question “Are people buying it?” and is a sentiment measurement of the ratio of content that is positive towards the candidate winning divided by the content that discusses the candidate loosing. Finally Campaign Progress answers the question “Is the campaign getting noticed? – and measures the percentage increase in the size of the campaign.

The idea of distilling peoples’ formal and informal communications into aggregated, actionable information is compelling. The ability to analyze the content of unstructured information holds exceptional promise. TrendIQ takes a massive amount of intelligence and shows you its contents and how it changes over time.

I’d love to examine how decision makers use this information to improve their decision making. How does informally provided information relate to how people truly act? What is the right way to interpret and utilize sentiment? How can all the information noise be simplified such that the patterns jump out? The problem is fascinating and technology gets closer and closer to providing valuable insight.

I’d love to hear any of your experiences in this area.

Click here for other applications of TrendIQ’s methods.

Comments