Content Ferret from NautilusOne

Screencast Demo

Businesses want to hear from their customers. Many use the Internet as a way to listen to the customer's voice. That works, but generates a lot of material, and you can't keep up with it.

Communispace faced this problem, and I built the Content Ferret to address their requirements. Communispace runs electronic communities for clients, asking questions, moderating discussions and analyzing the answers. The Content Ferret lets them search and drill down for information.

Perhaps more significantly, the tool lets them identify improtant ideas in the midst of all the chatter. Assuming an idea is a noun / verb pair in a sentence, the Ferret identifies all possible pairs, and then sees if they have been "seconded" by somoene else in the community during this discussion. The analyst can quickly scan through all the pairs, see which ones are intersting, and drill down for the details. Specific quotes can be copied through the clipboard to a report.

As an example, HP's medium-business printer division was able to identify calibration as an important customer issue, and address that concern. The printers were stopping in the midst of large runs and requiring calibration. Here is a screen-cast demonstrating the tool and showing how this issue was surfaced. The screen-cast will run in your browser, and when it is done, use the back button to return here.

Identifying ideas in content is a big subject of research and other companies provide complex, sophisticated tools to attempt analysis - such as Clarabridge and Crimson Hexagon. I used a simpler approach which quickly delivered useful results.