Martin Källström

Primelabs presenting @ Podcamp Europe 2/2

Primelabs2007-06-135:45384x288 original source ↗

Presentation by Primelabs CEO Martin Källström @ Podcamp Europe, Stockholm, June 13 2007.

In this recording — 6 extracted moments
click any row to play that moment
factPrimelabs measured blog influence by analyzing patterns of citation and linking between blogs.0:04factPrimelabs analyzed a database of approximately 55 million blog posts to extract data on blog influence across different countries.0:57factPrimelabs used blog analysis to measure brand awareness by tracking keyword mentions during specific time periods.1:25factBlog analysis of the Sony brand during 2007 Q1 showed strong awareness in technology and media spheres but weak presence in lifestyle segments.1:50factPrimelabs positioned their blog map technology as a tool for brands to measure where their name or brand receives exposure across the blogosphere.4:24factPrimelabs proposed using sentiment analysis technology to distinguish positive from negative brand mentions in their blog analysis platform.4:52
Transcript

[00:00:04] Speaker A: If we look at the structure here, blog number 1 has clearly influenced blog number 3 more than blog number 2. This is quite a simplistic example, but— and a textual comparison Comparison

[00:00:32] Speaker A: will introduce a lot of noise. That means that for measuring one blog post or even 100 blog posts, this will not be statistically significant. But if you— what we like to do is to take this kind of measure and apply it on the entire database of all 55

[00:00:57] Speaker A: million blog posts to extract information out of that data on which blogs are most influential in Sweden, in other countries, and in the world. So I'm now going to talk about how we use this blog map to measure brand awareness as well.

[00:01:25] Speaker A: And these are blogs that during the first quarter of this year measured the keyword Sony, which is a brand which all of you are aware of. And this is actually the people are— the blogs who have mentioned Sony but not Sony Ericsson because everyone is talking about mobile phones. So

[00:01:50] Speaker A: we separated out the blogs that are talking only about Sony. And what we can see if we apply the legend is that Sony's got a lot of awareness in the technology and media sphere, a little bit in the fashion sphere, but not so much in the lifestyle sphere as we could have hoped. So maybe that

[00:02:16] Speaker A: could be a venue for Sony to look at how to reach the blogs located in the lifestyle sphere most. Here's another brand. Göran Persson, former Swedish Prime Minister. And I was thinking about Put him up to a

[00:02:42] Speaker A: fight with with Eva Fonsino, which is a fashion queen, and she's been very influential. He's naturally been very influential in media, and and if we compare these two brands, so to speak, and and

[00:03:07] Speaker A: using the the map, we can see that Ebba has lots of recognition in the fashion sphere, also in the media and in the lifestyle sphere. And you can see that the

[00:03:33] Speaker A: blogs in the technology and media sphere that have written about Ebba von Sydow are not the geeky blogs who write almost about themselves, who no one links to. But the blogs writing about Ebba are very interconnected with blogs writing about media. So

[00:03:59] Speaker A: you can see that the structure of this map really conveys a lot of information if you delve deeper into it. And you could compare this with Göran Persson. Which is a very much stronger brand than, but not

[00:04:24] Speaker A: in the fashion sphere. So maybe these two should be overlaid, but basically it's a great tool for comparing where are my brand or my name getting exposure in the blogosphere. And

[00:04:52] Speaker A: the natural extension of this would be to also use— apply sentiment mining on top of this to be able to say that for one keyword, which brand— which blogs are talking— are mentioning our keyword in a positive sense and which ones are mentioning our keyword in a negative sense. Yeah.

[00:05:19] Speaker A: Because that technology exists. So then in the map we could visualize where do my brand or my name have a positive influence, where does it have a negative influence. So I think that's it for me. You're welcome with questions, or you can grab me later for asking my

[00:05:44] Speaker A: list. Thank you.