Twitter Data: An Investor's Perspective | Hacker News Reader

Twitter Data: An Investor's Perspective

techcrunch.com

44 pointsjakestein16 comments

Open articleView on HN

This is actually the perspective of someone who would like you to know about his new company called RJMetrics.

Perhaps. But smart people know that this is good analyis: http://www.avc.com/a_vc/2009/10/the-cohort-analysis.html

Actually surprised he didn't get it into the title.

"Twitter Data Analysis: An Investor’s Perspective (From RJMetrics)"

Still, 33 mentions of "RJMetrics" on the page isn't too bad.

Who cares that it's from someone promoting their company? They're giving data that a lot of people are interested in, that hasn't been put elsewhere. The fact that it's better, more thorough data than other places also shows that there just might be a better source out there for gathering data (ie RJMetrics). Everyone wins.

>> "38% of users have never sent a single tweet, and over 75% of users have sent fewer than 10 tweets."

That's pretty damning. Are these accounts setup in the hope people will follow them - honey pot style maybe? For spamming later.

>> Source = TXT: <5%

Since less than 5% are using sms to update twitter surely they can abolish the outdated 140 character limit?

Since less than 5% are using sms to update twitter surely they can abolish the outdated 140 character limit?

Never having used it, I can't actually say, but I had the impression that the 140 character limit was the point of Twitter.

I guess to a certain extent, but for most use cases it just becomes irritating and makes people use url shorteners, txt spk etc.

Not to be a broken record, but I think these accounts are setup to "listen" to the brands, people, celebrities, products, etc that the subscriber is interested in. Most people don't blog, but lots of people read blogs. Twitter isn't really a social network, it's a subscriber broadcast medium.

A few interesting points:

the average time between any two tweets from the same user is exactly 24 hours.

i.e. people use Twitter every day.

The rate of new user acquisition has plateaued at around 8 million per month.

Wow, only 8m new users per month. That's gotta hurt.

We know that Twitter has 50 million registered users

OK, small user base then.

the problem with the user base seems more likely to be that you can ignore almost 75% of all users due to no activity

Not exactly - twitter, unlike facebook, is asymmetric, so you expect a large percentage of users who are just listening. That doesn't mean you can ignore them - they are the audience.

you're right about the asymmetry, but let's not pretend that 75% of 60M are actively listening to anyone...

I would be stunned if that number was higher than 10%

With any of these social platforms there is a small percentage of creators and a large population of consumers. For example, look at Youtube. The same can happen for twitter...a lot of people may just consume the tweets of celebrities or other feeds. Low tweeting is not a big issue. The consumption traffic continues to grow

Low tweeting is a big issue, since it negates some of the Twitter narrative. The whole "pulse of the planet" realtime search thing isn't really valuable if most of the users aren't contributing to the corpus.

Also, if it's just a handful of the users creating the value, then the switching costs are much lower. You'll go to "where the celebrities are" (which may turn out to be Facebook, or something else) and not the harder to move "where are your friends are".

If those numbers are real, that's pretty deflating.

Huge increase in user growth rate this year means most users are new users.

New userbase + long average time between early tweets -> most users have < 10 tweets.

Reply on news.ycombinator.com