Who matters more? Perez Hilton or T-Mobile customers?

by Eric Melin on October 15, 2009

SiteTypeBreakdownWhen a major crisis develops for a company, you can bet that blogs and social networking sites will be a big part of the conversation that is happening on the Internet.

That is why the results of a six-day sampling from Spark regarding the T-Mobile Sidekick‘s recent data-loss problem were so surprising.

You might have heard about this crisis from a popular Internet personality. Sure, celebrity blogger Perez Hilton is angry that his Sidekick data has disappeared (as he demonstrated continually on Twitter for the last week or two), but where is the united public outcry?

Is it possible that Hilton’s high profile is causing more media coverage of the event than any unified or connected consumer outage?

It must be something about the name Hilton. Remember Paris?

As featured on the Web Buzz segment of Kansas City CBS affiliate KCTV5, the sampling showed that the biggest site-type category to mention the Sidekick from Oct. 9 to Oct. 14, 2009 was the “News” category.

Traditional and web-oriented news organizations made up a bigger slice of the overall conversation surrounding the T-Mobile Sidekick (37%) than both blogs and social media sites did individually.

Another surprising statistic is as follows: Of those news sites (which are usually objective), there was a surprising amount of sentiment attached.

NodeSentimentBreakdownSome of this can be attributed to the comments section of each post (since the node sentiment tool averages the amount of sentiment on a given page), but also there were many negative terms used in the news stories themselves, suggesting that news outlets were covering the dissatisfaction of T-Mobile Sidekick customers after learning that their data had been lost by Microsoft.

  • Negative: 57.78%
  • Neutral: 27.22%
  • Positive: 15%

Additionally, the URLs that mention the Sidekick were not as connected as most Internet trends usually are. This also could be partially due to the overwhelming amount of individual news sites that covered the story.

A common virtualization looks more like the one from last week’s H1N1 vaccine sampling (below):


Now compare that with the screenshot of the T-Mobile Sidekick virtualization (below). As you can see, there are lots of URLs, but a considerable less amount of connection between those URLs. This means that the sites that covered the Sidekick data loss were not nearly as influential as the H1N1 sites. The Sidekick conversations were happening on their own and posts were rarely linking to each other.


A few guidelines to understanding virtualization:

  • Each sphere represents an individual URL.
  • Green spheres have positive sentiment.
  • Red spheres have negative sentiment.
  • Gray spheres have neutral or no sentiment.
  • Lines connecting spheres represent first-degree linkages.

{ 0 comments… add one now }

Leave a Comment

Previous post:

Next post: