State of the Edmonton Twittersphere – April 2009

Welcome to the fourth State of the Edmonton Twittersphere, my look at the intersection of Twitter and Edmonton. You can see last month’s stats here.

The source of the data this month remains the same – Twitter Search. If a user has his or her location set to Edmonton, St. Albert, Sherwood Park, Leduc, Nisku, Stony Plain, Fort Saskatchewan, Beaumont, Spruce Grove, or matching lat/long coordinates, they are considered an Edmontonian. If a tweet is “about Edmonton” it contains either the word Edmonton, the #yeg hashtag, or both.

As you’ll see in a couple of the graphs below, Twitter Search was down for just over 3 hours on April 19th (from about 8:48 PM until 11:51 PM MST), so tweets posted during that period of time were not counted.

For April 2009:

# of local users: 4489 (an increase of 31% over March)

To clarify, that means there were 4489 users who posted at least one tweet in April 2009 with their location set to something that makes them an Edmontonian as described above. This number should be treated as a minimum – there are probably many more Edmonton users without their location set.

# of tweets by local users: 184015
# of tweets by local users containing #yeg: 9815 (5.3%)
# of tweets by local users that were replies: 65984 (35.9%)
# of tweets by local users containing links: 31056 (16.9%)
# of tweets by local users that were retweets: 7064 (3.8%)
# of tweets by local users that were twooshes: 8118 (4.4%)

To clarify the # of tweets containing #yeg – that number actually includes all the local hashtags that start with #yeg, so it includes #yegfood, #yegtraffic, etc. This is consistent with the stats for previous months, though it made less of an impact before. I’ll try to break out the stats by tag next month.

Here are the numbers above in graphic form:

Here are the number of local users created per day in April, using the best available data from Twitter:

Here are the top clients used by local users for posting updates (remember that web includes all unidentified API calls too):

Some other interesting stats for the month:

  • The ten most active local users (most tweets first): AndyGroenink, DebraWard, akomuzikera, TrinaMLee, Edmontonsun, fcedmonton, Etown_Jenn, britl, bingofuel, wickedmickey
  • About 52% of all local tweets were posted between 9 AM and 5 PM.
  • Local users posted roughly 4.3 tweets per minute in April, compared with 3.4 per minute in March.
  • The day with the most local tweets posted was April 23rd at 8069. On average, 6134 local tweets were posted each day (compared to 4876 in March).
  • Of the 65984 replies posted by local users this month, 24916 or 38% were to other local users.
  • A total of 808 local users posted 50 times or more this month. In comparison, 796 local users posted just once this month.

I’m not including the top ten users by # of followers this month because I haven’t been able to update it yet, and I suspect it hasn’t changed much anyway. Also, I hope to have a dynamic list available soon.

Instead, I have a new top ten as suggested by the Social Web Meetup gang (thanks to @britl for letting me know). Here are the ten “most replied to” local users for April 2009: britl, bingofuel, chrislabossiere, citizenfish, Sirthinks, JodieGiese, frostedbetty, angelzilla, mastermaq, Out_Inc. This stat obviously skews toward the chatty, but it’s interesting nonetheless.

Tomorrow, I should be posting the stats for Calgary. Stay tuned!

UPDATE: Here are the Calgary stats for April 2009.

State of the Calgary Twittersphere – March 2009

Welcome to the first State of the Calgary Twittersphere, my look at the intersection of Twitter and Calgary. After receiving many requests for this from Calgarians after my State of the Edmonton Twittersphere posts, combined with my own curiosity, I figured it was time to do some stats for Calgary. I only captured half of the data for March 1st, but otherwise I think these numbers are fairly solid.

Using Twitter Search, I collected anything posted by Calgarians, or about Calgary. If a user has his or her location set to Calgary, Airdrie, Okotoks, Cochrane, Strathmore, or matching lat/long coordinates, they are considered a Calgarian. If a tweet is “about Calgary” it contains either the word Calgary, the #yyc hashtag, or both.

For March 2009:

# of local users: 3717

To clarify, that means there were 3717 users who posted at least one tweet in March 2009 with their location set to something that makes them a Calgarian as described above. This number should be treated as a minimum – there are probably many more Calgary users without their location set or that were not captured for some other reason.

# of tweets by local users: 147549
# of tweets by local users containing #yyc: 2936
# of tweets by local users that were replies: 51721
# of tweets by local users containing links: 28902
# of tweets by local users that were retweets: 4463
# of tweets by local users that were twooshes: 5821

This graph shows these numbers visually:

Though perhaps a little inaccurate, here are the best numbers I could get from Twitter for the number of local users created per day during March:

Here are the top clients used by local users for posting updates (remember that web includes all unidentified API calls too):

Some other interesting stats for the month:

  • The ten most active local users (most tweets first): codsta, wikkiwild1, strategicsense, yuki_hime, izzynobre, darylcognito, pigazine, MitchyD, dblacombe, devlind
  • About 54% of all local tweets were posted between 9 AM and 5 PM.
  • Local users posted roughly 3.3 tweets per minute in March.
  • The day with the most local tweets posted was March 31st at 6518. On average, 4760 local tweets were posted each day.
  • Of the 51721 replies posted by local users this month, 14650 or 28% were to other local users.
  • A total of 693 local users posted 50 times or more this month. In comparison, 542 local users posted just once this month.

And finally, the top ten users in Calgary (as of April 11th) by # of followers: douglasi, MarkIsMusing, tessaru, VeerUpdate, nolanmatthias, strategicsense, codsta, dayhomemama, scrawforditm, CalgaryRealtor.

I may put together another post to compare Edmonton and Calgary, but in general I’d say that although Calgary seems to have more users, they don’t seem as connected to one another as Edmontonians are (as evidenced by the # of tagged tweets and replies to other local users). I’m a little surprised that #yeg is so much more active than #yyc, actually.

Thanks to @andrewmcintyre for helping me with these stats (he ranked #13 on the most active users for the month btw). If you have any comments or feedback let me know so that I can improve these statistics in future months.

State of the Edmonton Twittersphere – March 2009

Welcome to the third State of the Edmonton Twittersphere, my look at the intersection of Twitter and Edmonton. You can see last month’s stats here, but note that I recently discovered a few of the numbers were wrong and I have since updated the post. The two key figures, the number of local users and the number of tweets posted by local users, remain unchanged.

The source of the data this month remains the same – Twitter Search. If a user has his or her location set to Edmonton, St. Albert, Sherwood Park, Leduc, Nisku, Stony Plain, Fort Saskatchewan, Beaumont, Spruce Grove, or matching lat/long coordinates, they are considered an Edmontonian. If a tweet is “about Edmonton” it contains either the word Edmonton, the #yeg hashtag, or both.

For March 2009:

# of local users: 3421 (an increase of 160% over February)

To clarify, that means there were 3421 users who posted at least one tweet in March 2009 with their location set to something that makes them an Edmontonian as described above. This number should be treated as a minimum – there are probably many more Edmonton users without their location set.

# of tweets by local users: 151146
# of tweets by local users containing #yeg: 6634 (4.4%)
# of tweets by local users that were replies: 55234 (36.5%)
# of tweets by local users containing links: 24449 (16.2%)
# of tweets by local users that were retweets: 3701 (2.4%)
# of tweets by local users that were twooshes: 5401 (3.6%)

This graph shows the above numbers compared to the correct numbers from last month:

I think this data is a little inaccurate, but it’s the best I could get from Twitter – here are the number of local users created per day during March:

Here are the top clients used by local users for posting updates (remember that web includes all unidentified API calls too):

Some other interesting stats for the month:

  • The ten most active local users (most tweets first): britl, fcedmonton, angelzilla, AndyGroenink, bingofuel, TrinaMLee, akomuzikera, grempel, foomanizer, adampatterson
  • About 55% of all local tweets were posted between 9 AM and 5 PM.
  • Local users posted roughly 3.4 tweets per minute in March, compared with just 1.8 per minute in February.
  • The day with the most local tweets posted was March 20th at 6061. On average, 4876 local tweets were posted each day (compared to 2598 in February).
  • Of the 55234 replies posted by local users this month, 25398 or 46% were to other local users.
  • A total of 668 local users posted 50 times or more this month. In comparison, 550 local users posted just once this month.

And finally, the top ten users in Edmonton (as of April 1st) by # of followers: Pat_Lorna, mastermaq, LesM, redneckmommy, cleversimon, ianbakewell, melanienathan, babyrumps, revtrev, ctvedmonton

As expected, the number of local users grew even faster this month than last. The local media’s recent love affair with Twitter no doubt has a lot to do with that! I suspect growth will taper off a bit in April, but will still be impressive.

A common request for these stats is to see how Edmonton compares with other cities. Within the next week or so, I’ll be posting the same stats for Calgary, so stay tuned!

Don’t forget to check out EdmontonTweetup4, taking place on April 7th. Hope to see you there!

State of the Edmonton Twittersphere – February 2009

Welcome to the second State of the Edmonton Twittersphere, my look at the intersection of Twitter and Edmonton. You can see last month’s stats here.

The source of this data remains the same. Using Twitter Search I collected anything posted by Edmontonians, or about Edmonton. If a user has his or her location set to Edmonton, St. Albert, Sherwood Park, Leduc, Nisku, Stony Plain, or matching lat/long coordinates, they are considered an Edmontonian. If a tweet is “about Edmonton” it contains either the word Edmonton, the #yeg hashtag, or both.

There was confusion about which numbers were local or general in last month’s post, so I’ve focused in on local numbers this time. For February 2009:

# of local users: 1314 (an increase of 54% over January)

To clarify, that means there were 1314 users who posted at least one tweet in February 2009 with their location set to something that makes them an Edmontonian as described above.

# of tweets by local users: 72748
# of tweets by local users containing #yeg: 2489 2231
# of tweets by local users that were replies: 29282 28212
# of tweets by local users containing links: 16922 13318
# of tweets by local users that were retweets: 1577 1468

This graph shows the (incorrect) above numbers compared to the same numbers from last month:

Of the 463 additional local users I identified in February, only 320 were new accounts created during the month. The other 143 users had existing accounts and either didn’t post in January, or only changed their location to something matching a local user in February. This graph shows the new user creation by day:

On the 18th, Twitter added a new property to the results returned by Twitter Search – the source. That means I was able to identify the clients that local users use most when posting updates. The top five are: web, TweetDeck, twitterfeed, TwitterBerry, and twitterrific. Note that web source actually includes all unidentified API calls too.

Some other interesting statistics for the month:

  • The ten most active local users (most tweets first): fcedmonton, angelzilla, britl, mastermaq, justNICKI, adampatterson, Pat_Lorna, foomanizer, AndyGroenink, and JodieGiese.
  • Just over 55% of all local tweets were posted between 9 AM and 5 PM.
  • Local users posted roughly 1.8 tweets per minute in February.
  • The day with the most local tweets posted was February 26th at 3742. On average, 2598 local tweets were posted each day. 
  • Of the 29282 replies posted by local users this month, 13141 or 45% were to other local users.
  • Just 343 local users posted 50 times or more this month. In comparison, 164 local users posted just once this month.

And finally, the top ten users in Edmonton (as of March 1st) by # of followers: Pat_Lorna, mastermaq, LesM, redneckmommy, cleversimon, melanienathan, babyrumps, saralees, jerryreeder, revtrev

I knew the number of local users would increase this month, but I had no idea by how much. With a sudden interest in Twitter in the local media, I expect the growth for March to be even larger.

I hope you enjoyed this unscientific look at the Edmonton Twittersphere! I have done my best to provide accurate numbers, but I can’t give any guarantees. If you have any suggestions or other feedback, please let me know.

UPDATE: Great suggestion from @britl – the total number of twooshes (tweets that are exactly 140 chars) posted by local users in February 2009: 2233

Can Facebook become the new default?

I find Facebook incredibly useful, if not particularly exciting. My usage reflects that – I like to add people on Facebook in order to maintain connections, and I like to keep my profile looking fresh, but I rarely surf Facebook like I used to. Yet there’s no escaping Facebook. The numbers tell the story. Check out these statistics compiled for a recent Fortune article:

  • 175 million members
  • 3 billion total daily minutes of use
  • 850 million photos uploaded each month
  • 15 million who update their status daily
  • 24 million pieces of content shared each month

Very impressive. Also in the sidebar, Fortune looks at the race to 150 million users. That feat took Facebook 5 years, versus 7 years for the iPod, 14 years for the cell phone, 38 years for the television, and 89 years for the telephone. Obviously it’s not an apples-to-apples comparison, but you get the idea (and notice how other technologies such as Google or Windows are left out).

With numbers like that, it’s not hard to listen to Mark Zuckerberg and actually think he’s got a shot at achieving his new goal:

"We think that if you can build one worldwide platform where you can just type in anyone’s name, find the person you’re looking for, and communicate with them," he told a German audience in January, "that’s a really valuable system to be building."

In the article, author Jessi Hempel positions Facebook as the new phone system, but I think the new email system is perhaps a more reasonable comparison. I think the “default” right now when you make a connection is to get an email address. You collect business cards at events and they all have phone numbers and email addresses but how many people actually pick up the phone? Email is the default.

What if Facebook could become the new default? Clearly, that’d be a big deal.

Already I think Facebook is the default platform for events, and most people seem to think it’s the default for photos. Can it become the default for communication in general? As I’ve said before, I think Facebook Connect is a step in that direction.

State of the Edmonton Twittersphere – January 2009

I love Twitter, Edmonton, and pretty graphs. In this post, I combine all three! Welcome to the first State of the Edmonton Twittersphere, my look at the intersection of Twitter and Edmonton.

As some of you know, I’ve been working on a new side project for a couple months now called ShareEdmonton. I’ll post more details on that later, but for now all you need to know is that tracking Twitter usage in Edmonton is a big part of the project. As a result, I’ve gathered lots and lots of data over the last month, and I thought I’d dig into a little.

All of the data I am sharing in this post was collected from Twitter Search. It’s a fairly broad set of updates – anything posted by Edmontonians, or about Edmonton. The data set is incomplete (I’m sure I didn’t capture all tweets, and I improved the algorithm for the second half of the month), and is subject to certain restrictions. For instance, unless someone puts “Edmonton” or something similar (lat/long, Sherwood Park, etc) in the Location field of their profile, I have no way of reliably knowing whether or not they are a local user. That said, I feel that the data set I do have is very representative of Twitter usage in the Edmonton area. Going forward, I’ll try to keep the methodology consistent.

Enough preamble – on to the statistics! For January 2009:

# of users: 1948
# of tweets: 52697

# of local users: 851 (43.69%)
# of tweets by local users: 48091 (91.26%)

# of tweets containing #yeg: 1315 (2.50%)
# of tweets that were replies: 18837 (35.75%)
# of tweets containing links: 11033 (20.94%)
# of tweets that were retweets: 265 (0.50%)

And some other interesting statistics for the month:

  • Ten most active users (most tweets first): fcedmonton, angelzilla, britl, zoocasaedmonton, mastermaq, justNICKI, Pat_Lorna, iNews880, foomanizer, bgrier
  • If you ignore the three “bots”, then alainsaffel, GuitarKat, and fusedlogic also make the list.
  • Those thirteen users account for just over 21% of all tweets in the data set.
  • Nearly 45% of all tweets were posted between 10 AM and 4 PM.
  • The number of tweets posted per day generally increased throughout the month (though this is at least partially due to improvements to my code over the month). Monday, January 26th had the most tweets posted (2592).
  • Local users posted 18042 replies. Of those, 7406 (41.05%) were to other local users.
  • Most aren’t frequent posters: 293 local users posted 31 times or more this month.

And finally, the top ten users in Edmonton (as of January 31st):

In the future, I hope to look at trending topics for the month to get a better idea of just what Edmontonians were talking about. Numbers are one thing, content is quite another.

I’m surprised to see that 851 local users updated this month (all but 110 updated more than once). I would have guessed there were about that many local users in total, not active! So that’s good. It’ll be interesting to see how that number grows over the next few months.

I hope you enjoyed this quick, completely unscientific look at the Edmonton Twittersphere! If you have any suggestions on other things I could include in future posts (or potentially as an update to this one) please let me know.

See you at Edmonton Twestival on February 12th!

Facebook's virtual gifts – money well spent?

In a post at VentureBeat yesterday, Eric Eldon shared some estimates that suggest Facebook’s revenue from virtual gifts this year will be in the range of $28 million to $43 million. That’s a serious amount of coin for nothing more than an image on a web page.

Gifts are priced at $1 each, and the study found that an average of 470,000 are sold each week.

Facebook introduced the gifts feature in February of 2007. A gift is simply an image of something, like a heart, a flower, or hundreds of other options, that when given, shows up on a “gift box” in a user’s profile. If the gift is public, then the recipients’ friends can see it, too. If it’s private, only the recipient and the giver can see it.

I think the key there is “simply an image”. This is definitely one of those things where you can’t help but think “why didn’t I come up with that!”

Clearly, gifts are a good source of income for Facebook. I wonder who buys them though. Why are people so happy to pay $1 for a bunch of pixels on a web page?

Surely that $40 million could have been spent on something better?

Edmontonians flock to transit in 2008?

100 Years of ETSI never thought I’d write a headline like that, but apparently it’s true. According to an article at CBC last week, Edmonton Transit is reporting that ridership rose by 8% in the first six months of 2008 compared with the same period last year. That’s roughly equivalent to an extra 2.5 million fares. ETS called the increase “astounding”, but seemed confused about the cause:

“Buses are packed, the LRT is packed, ridership continues to increase at an astounding rate,” said Ken Koropeski, director of service development for Edmonton Transit.

The increase is almost three times the growth being experienced by other transit systems across the country, Koropeski said, a trend for the city over the past few years.

The CBC article cites three potential reasons: high gas prices, the booming economy and related influx of newcomers to our city, and the U-Pass. I suspect #2 is the biggest cause, but that’s just my gut reaction. I wonder what impact the growth will have on security calls, which have already increased 20% over the last three years.

I also wonder why ETS didn’t share any of this information back in April? You might recall that Statistics Canada released information at the time that said nearly 80% of Edmontonians get to work in vehicle. Granted the StatsCan information came from the 2006 census, but it would have been a good opportunity for ETS to dispel some myths about stagnant (or least slowly growing) ridership.

I hope the growth continues. ETS turns 100 this year, and will celebrate Centennial Week from September 12th to 20th. You can learn more at the ETS website.

In Edmonton, we like to drive

Statistics Canada has released some new data from the last census that shouldn’t shock anyone who lives in Alberta’s capital city. Nearly 80% of us get to work in a vehicle:

The new data from the 2006 census found that 12.7 per cent of workers in the city of Edmonton get to work using public transit, while 79 per cent either drive or travel in a vehicle as a passenger.

Statistics Canada said the reliance on cars seems to increase with the age of the commuter. While those under the age of 25 travelled by vehicle 70.7 per cent of the time, that rate increased to 81.6 per cent for those aged 25 to 34. The rate was even higher for those aged over 35, at 87.2 per cent.

Cheap Gas?

The average Alberta commuter takes a car 84% of the time, so we’re slightly better than the rest of the province.

I guess Bob Boutilier, our city’s Transportation Department GM, wasn’t kidding at the ETS conference a few weeks ago when he said a big challenge is the “pickup truck and two car” mentality of most Albertans. Thanks to the census data, I now have a number to attach to that statement.

Some people like to suggest that we’ll never improve our public transit system until everyone experiences just how bad it is right now. Maybe there’s some truth to that after all. I wouldn’t be surprised to learn that the majority of that 80% have never been on a bus or LRT car.

That needs to change.

World Internet Population: Reading the fine print

Post ImageYesterday comScore released their global Internet traffic rankings for the month of May. Their research shows that 772 million people worldwide were online in May, which is a pretty large number. Still, as ZDNet notes, that’s only 12 percent of the world’s population. Here’s how the press release reads:

There were 772 million people online worldwide in May (defined as those individuals age 15 or older who accessed the Internet from a home or work location in the last 30 days), an increase from 766 million in April, representing a 16 percent penetration of the worldwide population of individuals age 15 or older.

And further down the page, we find the fine print:

** Excludes traffic from public computers such as Internet cafes or access from mobile phones or PDAs.

Seems to me that excluding mobile phones in particular would lead to a much lower number than the true online population. In the developed world, computers dominate access to the Internet, but that’s not the case in the developing world!

A quick search led me to this W3C press release (from September 2006):

According to the World Bank, more than two billion people own a mobile phone and 80% of the world’s population has access to GSM service. With one million new subscribers every day, almost four billion people will have a mobile phone by the end of 2010.

I suspect the vast majority of those phones are web-enabled. If anyone has a link to usage statistics, let me know in the comments!

Read: comScore