Bloomberg Anywhere Remote Login Bloomberg Terminal Demo Request


Connecting decision makers to a dynamic network of information, people and ideas, Bloomberg quickly and accurately delivers business and financial information, news and insight around the world.


Financial Products

Enterprise Products


Customer Support

  • Americas

    +1 212 318 2000

  • Europe, Middle East, & Africa

    +44 20 7330 7500

  • Asia Pacific

    +65 6212 1000


Industry Products

Media Services

Follow Us

Bloomberg Customers

Businessweek Archives

How marketers can monitor what 10 million bloggers are up to

? Kurt Andersen needs a three-hour sit-down with Jeff Jarvis |


| The Power of Us ?

June 10, 2005

How marketers can monitor what 10 million bloggers are up to

Stephen Baker

Mining blogs for marketing intelligence is all the rage. But how to do it? Howard Kaushansky of Umbria Communications dropped by this week to tell us how his Boulder, Colo., company does it. Umbria has analysed the language and syntax of thousands of blogs, and with that has devised an expert system that they say can analyse not only the opinions in the blogs, but also chart them by age and gender. So, clients such as Electronic Arts or Sprint can learn how their product launches or marketing campaigns are being received, day by day, by, say, teenaged boys or baby-boomer women. If Umbria can provide customer case studies, we'll blog a few.

Legions of grad students provide provide Umbria its raw material--blogs organized by age and gender. For $20 per hour, they go through blogs and categorize them. Given this raw material, the computer figures out what sets each demographic group apart.

I took a quick tour of the blogs and picked out three examples. My idea is to approximate what the software does, however primitively:

This one is pretty clearly a teenaged girl. An extract:

My dad woke me up early this morning. Sux. Anyway I miss my Alex sooo much and now I can't talk to him during the day b/c he's outta minutes and I dropped my phone in the toilet.

I just had to include that bit about the phone. Cracked me up. Anyway, presumably the software associates the following words with this age group (perhaps among others): sooo, sux, outta, b/c.

Next is a tech-minded young man, maybe a teenager.

I have a chatroom discussion on the discrepencies between episode III and episode IV that I dont wanna be late for. TTYL, remember to POAHF. NRN but you can EMSG. If you think were gonna F2F, GMAB I'll be LSHTTARDML because IMNSHO you suck.

Luckily for the computer, it doesn't have to understand all those acronyms in order to recognize them. There's a motherload in this post.

Here's a post from what I take to be a middle-aged woman:

The problem comes in thinking that somehow I can be on top of things, because there are always at least 20 things I am blowing off at any given moment. Better to start to enjoy the feeling of being at the center of uncertainty, permanently. If you grok the irony there.

I was a little surprised by the use of "grok," which I associate more with men. But that shows how much I know.

One more interesting point about this. What if that acronym-spouting second blogger is not a teenager at all, but a 45-year-old man who likes to write like a kid on blogs? No, the computer doesn't catch it. But it doesn't matter. Because from a marketer's point of view, if he's acting like a teenage, he might as well be one.

UPDATE: The woman I identified as middle-aged says she's younger. My apologies. Maybe Umbria's computer wouldn't have made that mistake. I also misspelled "discrepancy" on a version of this story that briefly escaped draft status yesterday.

04:30 AM


TrackBack URL for this entry:

Listed below are links to weblogs that reference How marketers can monitor what 10 million bloggers are up to:

? So, It's Sort of Like Issuing Formal Statements Next to the Water Cooler from Spanglemonkey

What a totally interesting experience! Having the guy who wrote the businessweek article commenting on my blog! It has me exploding with the implications of it. I love blogging! I'll try to explain what I'm thinking, but you probably get [Read More]

Tracked on June 10, 2005 11:05 AM

? Blog-Metric Snake Oil from Clickable Culture

Blogspotting's Stephen Baker discusses the methods behind Umbria Communications, a firm that mines weblogs, message-boards, chat and "etc" to extract "actionable market research about companies, products, people, and topics of i... [Read More]

Tracked on June 13, 2005 09:00 PM

? All Your Blog Are Belong To Us from reBang weblog

Fascinating entry by blogger Stephen Baker over on BusinessWeek's Blogspotting. The entry title is "How marketers can monitor what 10 million bloggers are up to" and it boils down to using a computer program to analyze blog entries and make what amo... [Read More]

Tracked on June 14, 2005 12:03 AM

? Umbria Communications - Understanding the Blogoshere and beyond from Blog: Shawn Rogers

I was reading Blogspotting this afternoon and ran across a piece by Stephen Baker that he wrote about Umbria Communications. His discussion with Howard Kaushansky President of Umbria is worth the read. I also had the opportunity to meet Howard... [Read More]

Tracked on June 14, 2005 12:16 AM

Now here is another application that low cost Indian BPOs will try and pounce on - getting hordes of educated engineers and othe graduates to surf the Net for recent blogs, and then try and categorize them using 'patented Umbria categorization technologies' to suit American marketers. At 10$ an hour!

Posted by: Ashok Karanth at June 10, 2005 07:55 AM

I am a 68-year-old blogger, mother of a 39-year-old blogger, grandmother of a 4-year-old blogger, writing to inform you that "discrepencies" is misspelled in both your article and in the blog that you mistakenly praise. I am surely not the first reader to point this out, but I may be the first one to offer to take $20 an hour to proofread your future articles. Just drop it into my Paypal account.

Posted by: savtadotty at June 10, 2005 08:15 AM

It's funny, you don't imagine that when you're writing about someone in your blog, that they are actually going to read it. Somehow I imagine my blog as some kind of strange combination of public/private that is directed exactly where I want it to go, but it isn't. Sorry if I offended.

That having been said, it's still sort of a creepy feeling to be an individual analyzed for demographic traits. I suppose my issues extend to marketing, in general. It reduces people to their ability and propensity toward spending. All the other parts of life, the individual experience, the joys and sorrows and important relationships, they're all fodder for marketing possibilities.

So anyway, that's my point, not anything personal with your writing of the article.

Posted by: Jo at June 10, 2005 10:45 AM

Savtadotty, It's going to take a while to live down that spelling mistake on discrepancy. Thanks for your proofreading offer.

As far as Spanglemonkey's comment, I certainly understand. I had this feeling as I was dipping into these Web sites that I was somehow treading on personal property. They're open to the public, but they don't expect sites like this to link to them. I think exploring such sites is worth doing, but maybe sending an email before linking to them would be called for.

I remember my first lesson about this type of issue. I was working my first week at a weekly newspaper in Vermont. In a selectmen's meeting, a woman came to complain that the music at a pub was way too loud. In the summertime, she and her husband had to keep their window open, and the music, she said, gave her husband palpitations. I strutted my stuff in the lead, playing on theme that rock music made his heart race.

After the paper came out, a host of his relatives criticized me, rightfully, for using his medical condition for my catchy lead. It taught me a lesson, but occasional reminders don't hurt.

Posted by: steve baker at June 10, 2005 11:03 AM

Umbria is actually one of a host of companies that do this kind of specialized unstructured data mining. Blogs are only the most recent example of public online conversation monitoring of this kind. You should check out and (full disclosure, we designed Kaava's site). Not sure how BM does their work, but Kaava's approach is interesting in relation to Umbria's. Sounds like the later relies on a large, distributed base of "cataloguers" and then processes results with software. Kaava has small teams of professional moderators - they used to be in the chatroom moderation business - who use extremely powerful software to monitor an incredibly diverse set of conversation. And they can get a lot more besides demographic information out of it.

Posted by: John Franklin at June 10, 2005 11:40 AM

You're a good sport, Steve.

Posted by: Jo at June 10, 2005 12:07 PM

What complicates this is the phenomenon of "buzz agents" who are pretending to like a product, and then post comments about it on blogs, online forums, discussion lists, etc....

...because they are paid to promote.

It is insincere, lying, and misleading for "data mining" tools.

These tools cannot judge authenticity, sincerity, or honesty.

Paid Word of Mouth is Deceptive Advertising: people posing as users, when they're not.

P.S. Good job of jumping into comment threads to respond to specific comments. This is what I advise to business bloggers especially, but almost none pay any attention. They're too busy to interact with blog visitors, except in "summarizing posts", a real cop out.

Posted by: steven streight aka vaspers the grate at June 12, 2005 05:21 AM

i was reading the "Evangeline" interview over on the Alphaville Herald last night. while reading it i came to the same conclusion as the virtual reporter - She was actually a He ->>

Urizenus: Not to sound like a biggot, but isn't that odd stuff for a guy to collect?

Evangeline: you believe I'm a guy?

Urizenus: aren't you? You read like one.

Evangeline: read?

Urizenus: you type like one

Evangeline: how does one type?

Posted by: csven at June 13, 2005 10:51 AM

(cut my final line off during preview edit. whoops.)

the thought i had - and i'll now expand upon - was that this technology would probably find it's way into online forums and virtual worlds (if it hasn't already). these are places where "alts" are common and are used to gain unfair advantage over other people/players. i'm not sure it'll be entirely successful in that regard. as pointed out by Mr. Streight, alt's are like actors; assuming different persona's and adopting the syntax of a different profile for their own purposes. even Evangeline makes a crude attempt at that.

however, there are other reasons for maintaining a good online reputation. just the potential for being outed may, in addition to those other reasons, prove compelling to a larger number of people.

on the internet, it's getting harder and harder to be a dog it would appear!

Posted by: csven at June 13, 2005 11:08 AM

Fascinating. Could be very useful to companies trying to interpret a flood of blogs.. and probably will get more useful as blog numbers continue to grow.

BTW, I wish your RSS feed had full articles.

Posted by: Johnnie Moore at June 16, 2005 03:35 AM

blog comments powered by Disqus