Will Marlow - Filed under 'How to measure Twitter influence'

January 20, 2010

The Point of the Official Retweet Button

A lot of people who are very knowledgeable and experienced at leveraging Twitter do not understand the point of the "Retweet" button that Twitter launched last year. It doesn't make sense to them for a number reasons, including:

1. It was unnecessary, because retweeting is one of the most popular activities on Twitter and it needed no encouragement to continue.

2. Twitter is famous for its lack of features, so in the absence of a need to change behavior (see point 1) it makes things more confusing as to why Twitter would spend time and energy on this particular new feature.

3. It makes retweeting even more confusing than it already is for new users, because it gives two options that seem to accomplish the same thing in different ways, with some users preferring one way and other users preferring a different way.

The reason for the new feature, however, is all about monetization. One of the cool things about Twitter is that every Tweet is equal. Twitter streams are all about timelines. If a good Tweet is shared with people, who share it with their own friends, the viral spread is organic and rapid. This awesome characteristic, however, makes Twitter very difficult to "search" because one Tweet looks the same as all the other Tweets (with the exception of Tweets that are favorited more often than others). And "search" (that is, "searching for relevant Tweets") is one of the most promising ways for Twitter make money (and for it to make good on the $1 billion valuation it earned recently).

So it shouldn't be surprising that Twitter introduced the retweet button, which seems to be an effort to minimally dilute the experience while adding a layer of potential for it to increase the likelihood that the company will produce significant value. The big question is, are people using the retweet button enough? (For the record, it seems that my followers are retweeting me the old way at a ratio of about 4:1, with four penciled in retweets for every one "push-button" RT using the new feature.)

Will Marlow co-founded AlumniFidelity to help his clients reposition their fundraising to benefit from Web2.0 technology and marketing techniques. He’s working with clients such as UVA, the College of William & Mary, the University of Oklahoma, Bowling Green State University, Randolph Macon College, and he loves nothing better than a thorny marketing challenge. Email him at will@alumnifidelity.com.

Loading mentions Retweet

Filed under // How to measure Twitter influence monetization of social networks ReTweet revenue Twitter

Comments [2]

January 15, 2010

Results from Day 3 of Twitter Experiment (with Chart)

In 24 hours, I have lost another 100 followers. (Remember, on the first night, I lost almost 2,000 followers, which must include all the people who use applications to auto-unfollow people who are no longer following them.) As the chart shows, my follower count stands at 8,893.

Although it is still early to look at the results for RTs and @replies, here's something interesting to chew on: In the last 12 hours, I have been retweeted or @replied approximately 23 times on one post. This is WELL ABOVE average, and a new record for any of my Tweets (my viral Tweets are typically retweeted three or four times).

This surge of retweets can be explained by two things, I suspect. First, as Twitter guru Guy Kawasaki (@guykawasaki) says, the best thing you can do to get retweeted is to Tweet good shit, and the Tweet that was responsible for almost all of the RTs and @replies was about how in just several days, the amount of money donated via text message to Haitian disaster relief has surpassed the total amount of money given to all charities and causes via text message last year. (That is interesting, and highly retweetable.)

But the second explanation is that, as I guessed, Twitter reach is not dependent solely on maximizing the number of followers you have. I had 12,700 followers on Monday. Yesterday, I had 8,940 followers. But after losing over 25% of my followers due to my own massive unfollowing, my Twitter reach (as measured simply by @replies and retweets) rose by a factor of more than 5x.

Prediction for the weekend: I still expect to see a more rapid loss of followers over the weekend, as some people tend to update their Twitter feeds only on Saturdays.

Will Marlow co-founded AlumniFidelity to help his clients reposition their fundraising to benefit from Web2.0 technology and marketing techniques. He’s working with clients such as UVA, the College of William & Mary, the University of Oklahoma, Bowling Green State University, Randolph Macon College, and he loves nothing better than a thorny marketing challenge. Email him at will@alumnifidelity.com.

Loading mentions Retweet

Filed under // Followers How to measure Twitter influence Posterous Twitter Twitter Influence Twitterholic Unfollow Unfollow Experiment

Comments [0]

January 13, 2010

Results from Day 1 - Did Bots or People Disappear?

As this chart from Twitterholic shows, less than 24 hours after unfollowing 12,000 people, I was unfollowed by over 3,500 followers!

My first observations/conclusions are as follows:

1. So far, no one from the list of people I interact with regularly has unfollowed me.

2. There were LOTS of bots and spammers following me, who were only interested in a recipricol follow (which I'm happy to lose).

3. There is a big advantage to following back your followers, in that you can always tell if someone is following you just by looking at their feed. If I don't ever go back to a 1:1 ratio, not being able to know this will take some getting used to.

4. I am going to add the initial people back very quickly, because I miss the DMs that this group sends me (I don't miss DMs from spammers or bots).

and lastly,

5. There are many people who take the "follow back" ethic very seriously and personally. I received more than one email angry email suggesting that I was being very rude. If that's the case, I apologize. I tried to take precautions against this by warning people in the days leading up to this experiment (which served the dual purpose of telling me who is reading my Tweets). In any case, my reason for doing this experiment was not to be rude, but to learn more about the dynamics of following/unfollowing on Twitter, and for the next several days I'll be monitoring what happens.

(PS - Interestingly, although my goal was to make it down to zero before adding people back, Twitter's odd numbering system never let me get down below 92 people on my follower list, despite not having a record that I was following anyone. In addition to that, Tweets from people who I no longer followed continued to show up in my feed until I replaced them with newly re-followed people.)

Will Marlow co-founded AlumniFidelity to help his clients reposition their fundraising to benefit from Web2.0 technology and marketing techniques. He’s working with clients such as UVA, the College of William & Mary, the University of Oklahoma, Bowling Green State University, Randolph Macon College, and he loves nothing better than a thorny marketing challenge. Email him at will@alumnifidelity.com.

Loading mentions Retweet

Filed under // Follower Count How to measure Twitter influence Social Media Social Media Plan Social media strategy Twitter Twitterholic Unfollow

Comments [0]

January 12, 2010

Day One of Twitter Experiment: Unfollowing 12,000+ People

Today I am beginning to unfollow 12,000 people, making my way down to zero. Currently, almost everyone I am following is also following me back. And starting last Friday, I began informing my "followers" of this experiment, and anyone who has been paying attention to my feed has had ample time to hear me announce that I am culling faux-followers, and that I will quickly begin to add people back who I interact with, or who reach out to me, or who I am interested in.

Yesterday I wrote about what I hope to learn from doing this.

On Friday I wrote about why I am doing this.

Today I just want to write briefly about which Twitter tools I am starting with. I am going to use Tweepi, which my friend Russ Dean helpfully turned me on to, and once I get beneath a certain level (say, around 10,000), I will use Twitter Karma, which works for Twitter accounts that are not humongous, but begins to fail when it needs to deal with too much information.

I will report back on whether these two tools are sufficient, or if I stumble upon others, or if I end up needing to unfollow folks by hand!

Will Marlow co-founded AlumniFidelity to help his clients reposition their fundraising to benefit from Web2.0 technology and marketing techniques. He’s working with clients such as UVA, the College of William & Mary, the University of Oklahoma, Bowling Green State University, Randolph Macon College, and he loves nothing better than a thorny marketing challenge. Email him at will@alumnifidelity.com.

Loading mentions Retweet

Filed under // How to measure Twitter influence Social Media Social Media Experiment Twitter Twitter Influence Unfollow Web Analytics Will Marlow

Comments [3]

January 11, 2010

Twitter Experiment Part 2: What I Hope to Learn from Unfollowing 12,000+ People

On Friday I wrote about my plan to do an experiment in which I unfollow 100% of the people I'm following right now, and then I'll re-follow only people who are not spammers or bots, and who I'm actually interested in. I am keeping a list right now of people who I interact with, or know, or who I find interesting, and I'll re-follow that list very quickly after I hit zero. If you want to be added to that list, just @reply me, or send me a DM, or email me at will@alumnifidelity.com.

On Friday I talked about WHY I was doing this. In this post, I want to talk about a few things that I hope to LEARN from doing this.

I look forward to learning:

How many people unfollow me, and thus, how many people on Twitter are only interested in following people who follow them back.
What happens to my "click through" rate.
What happens to my @reply and DM rate.
Whether this generates new word of mouth buzz and leads to new followers.
What this does to the overall visibility of my Twitter presence.

Confession: I'm also hopeful that I will be able to provide more evidence that having a large following is NOT a good measure of influence on Twitter. I believe that there is a misconception going around that a Twitter feed is only valuable if it "reaches" hundreds of thousands of strangers, as evidenced by huge follower lists. Firstly, I think that Twitter can be valuable even if it reaches ONE person who you wouldn't have reached with another mechanism. Secondly, I think that huge follower lists are faux-metrics. Hopefully my experiment can illustrate why the second point can be misleading, and possibly it can help people appreciate their small but committed core of followers, who are the source of true value on Twitter.

Will Marlow co-founded AlumniFidelity to help his clients reposition their fundraising to benefit from Web2.0 technology and marketing techniques. He’s working with clients such as UVA, the College of William & Mary, the University of Oklahoma, Bowling Green State University, Randolph Macon College, and he loves nothing better than a thorny marketing challenge. Email him at will@alumnifidelity.com.

Loading mentions Retweet

Filed under // Follower Count Followers How to measure Twitter influence Social Media Social Media Experiment Social Media Plan Twitter Unfollow Will Marlow

Comments [2]