Our friend Jeremiah Owyang, digital media influencer and analyst at Forrester posted a blog entry on Retweet and the power of word of mouth. To summarize:

[Information within Microblogging communities like Twitter encourage rapid word of mouth –of both positive and negative content]

In the post he talked about how information spreads quickly with Retweets. Essentially on Twitter if you “RT” or “ReTweet” someone you repeat what they say presumably because you agree or disagree with them OR if you want it to be amplified to your audience of followers on twitter.

Using BuzzGain we tracked usage of the Retweet before and after Nov 23rd. About 16,500 twitter users follow Jeremiah and so his influence in twitter is relatively large.

After Nov 23rd usage of Retweet among the top 50,000 Twitter users has increased by ~7%.

Great you think, Jeremiah well done.

Not exactly. I will chalk this as a not so positive trend. Why?

If you drill down deeper, tweets with negative sentiment (we measure sentiment as positive, negative or undetermined) has 45% more retweets than positive or neutral ones.

This is consistent with the view that bad news travels faster than good. This may also be primarily because of the Mumbai blasts and it could be an anamoly I agree, but you cannot equate a Retweet to someone linking to your blog on a post. How is retweeting different then?

1. Immediacy: Given the real time nature of twitter versus the 3-5 min it takes to blog link to someone, the amount of time to think, process and react to what’s being said is a lot less. If you tweet information that’s not factual or plain incorrect the person retweeting has a lot less time to fact check.

2. Value add: In most blog posts that link back to a post (including ones that just provide a list of links with delicious tags for example) there is a sense of “aggregation” or value add to the post. In most posts with link back there’s a lot more value add in the form of a different opinion or counter example. With Retweets, there’s ZERO value add. Its a plain and simple echo. You can argue there’s value in an echo, but I think its useless for most parts.

3. Reinforces a heard mentality. Retweets are an inaccurate measure of ripple effect. On average less than 0.35% of the top 50,000 twitter users posts are retweets. So it should have a relatively small effect right? No. Unfortunately, even though the next 350,000 followers of the top 50,000 are smart, connected, and influential individuals in their own right, their retweets to posts is close to 0.7%. Think about it. Double the top 50,000. This is a herd mentality pure and simple.

Do I think retweeting is going away. No.

Do I think its a measure of Word of mouth influence. Maybe.

Do I think it complicates the measure of word of mouth online to a point of making it useless. Absoutely.

Image credit: Wikipedia ]]> http://news.buzzgain.com/retweet-on-twitter-the-ripple-effect-of-the-echo-chamber/feed/ 0 http://news.buzzgain.com/what-we-can-tell-you-about-blogworldexpo-without-attending-it/ http://news.buzzgain.com/what-we-can-tell-you-about-blogworldexpo-without-attending-it/#comments Wed, 15 Jul 2009 13:17:08 +0000 Mukund http://news.buzzgain.com/?p=135

Blog World Expo recently concluded their Las Vegas Confab with over 2000 participants (includes bloggers, speakers, exhibitors, event staff, etc). There were 256 speakers and panelists. We did some quick analysis of BWE08 using our solution to give you some interesting metrics and some questions to consider.

1. Approximately 43% of BlogWorldExpo attendees are users of Twitter. This includes people that tweeted about it during the days of event. (we took only folks that had more than 3 tweets and included the offical tag (BWE08), and unoffocial tag BlogWord and Blog Expo). About 6% of BWE attendees took photos and either uploaded them to Flickr, Facebook, SmugMug and Zooomr. If other photographers took photos but did not upload it, we could not track them.

2. The top 3 most active sessionsduring BWE were 1. The keynote (with State of Blogosphere by Richard Jalichandra), 2. The TectSet Party and 3. Evening of the last day – after the event & sessions completed. So, its obvious people made a lot of friends, and said their good-byes and were still trying to keep in touch with the folks they met during the conference.

3. The sentiment of the users got better with the days. Clearly 80%+ of the folks loved the show at the end of day 3, compared to 56% on Day 1. The negative sentiment was the most on day 2 – I am still trying to figure out what session it was that contributed to it.

4. AlthoughTechnorati claims over 70% of bloggers are male, on twitter that’s not the case. Based on our numbers, full 43% of twitter users who attended BWE were women, compared to 57% men. We arenot the only ones to question Technorati’s survey information.

5. Finally photos. Collectively over 2848 photos (as of last night) were shared on Flickr, Facebook, Zooomr and SmugMug. (If you did share elsewhere, we can tell you uploaded photos, but cannot tell it was a part of BWE unless its been tagged appropriately). Flickr got the lion’s share of BWE photos followed by Facebook.

Some questions:

1. What happened on Day 2? Why did sentiment get worse?

2. Is there a better participation of women on Twitter? Or are Technorati’s numbers flawed?

3. I was surprised by the number of Twitter users who were are BWE. I thought it would be higher. So, if you did attend, did you not tweet during BWE?

If you attended BWE, please enlighten us.

This post primarily discusses corporate uses and engagement with Twitter, unlike how most people use it to keep up with friends and discuss “What are you doing”? The # of companies on twitter is growing dramatically from fewer than 5 in 2007 to over 500 in 2008.

Given that twitter has about 1.3 million registered accounts, the question we get asked frequently is:

<Update>: Jason asks where we got the 1.3 Million number vs. 3.1 reported by TwitDir. Two items: 1) We only track accounts that have more than 5 friends and at least 10 tweets. We think these are real users vs. bots. If we did all accounts, we track about 3.27 Million. Bottom line there are a TON of fake accounts on twitter.

2) We also did this initial run in July. We’ll update for September, but we doubt There were 1.5+M accounts added in 2 months.

<End Update>

“So if I get on twitter and make a lot of “friends” will our messages get propagated faster?”.

This is usually by a corporate client, who does not have a basic understanding of twitter or has just heard that twitter helps you get “Buzz”.

Usually we have to go over basics of twitter and community with them before they ask the next question:

“So how can we can get the key influencers to follow us and tweet about?”.

Another discussion ensues and topics such as engaging users and not spamming them comes up. They usually look a little lost at this point, but are able to somewhat relate to email spam so they eventually get it.

The next question comes up within the 30 minutes of discussion:

“So how many people have to follow us for our message to be heard?”.

We’re not sure if you really want to answer this question, but its a worthwhile exercise. There are varying degrees of influence on twitter. We define 3 types of twitter engagement. Each of these types of engagement is an increasing amount of influence obviously:

1. Another twitter user follows you (and based on her tweet history) tweets at about the same time as you do. Else if you are in different timezones, your influence on that person is lesser.

2. A twitter follower replies to your tweets (@replies) with a certain frequency. Or the # of @ replies you get as a twitter user.

3. A twitter follower retweets your tweets with some frequency.

There are 3 things that stand out when you analyze twitter statistics.

1. The top 250 twitter users (with the most number of followers, the most number of friends (people they follow) and the # of tweets) carry an inordinate amount of “influence”.

2. The average number of followers for the top 100 twitters is 9304, with the top at about 93,000 followers and #100 at about 6300 followers.

3. If you have less than 100 followers (with some exceptions), which over 500,000 users on twitter do, your influence rank drops to less than 8 (give time zone considerations). Meaning less than 18 people are really viewing your tweets, which indicates 4 of them are likely to @ reply and 1 is most likely to retweet.

The graph below shows Twitter influence. The top 100 users carry 65% of the influence on twitter and they have a minimum of 6300+ followers to a maximum of ~90,000 followers. Users from 100-5000 have 22% of twitter influence and so on.

Some questions:

1. Corporate clients still believe the best use of twitter is to leverage it as “yet another channel” for their message amplification. What are the best techniques you have used to convince them about the value of engaging their customers on twitter?

2. How are you recommending your clients engage with twitter users?

3. What’s the best example of large companies engaging on twitter?