Archive for November 5th, 2008

I got a message from Si in London recommending that I check out this upcoming TweetUp in Cambridge.  Turns out it was at Andala Cafe, home of my favorite hummus plate and barely a block from limeduck world headquarters.  Once again, the internet pwnz geography.  The TweetUp, called OpenCoffee, happens Wednesdays at 8:30ish, and seems to be a global phenomenon that has somehow landed right in my back yard.

This week, OpenCoffee was pretty Twitter-centeric, with Bijan Sabet of Spark Capital and Sanjay Vakil of LuckyCal leading a loose discussion of location-based apps, twitter platforms, and using twitter either as a start point or an end point for aggregation of what can only be described as your “stuff.”

The session was well-attended with some 40+ tweeple stretching the limits of Andala’s front room.  You can catch up on some of the chatter via this twitter search.

There was some debate about the utility of hash tags as opposed to unhashed tags, and taxonomy vs folksonomy.  I came down on the side of getting it mostly right and keeping it simple, which seems to be the guiding principle of twitter.  And then I fell off a chair while trying to explain Pants Status to the assembled masses.  My point, such as it was, is that there seems to be apps (if you can call Pants Status an app) that use twitter without any special action on the user’s part.

On a side note, the frothy beverage pictured above is an Andala specialty, fresh apple ginger juice.  I highly recommend it.  iPhone not included, but if you come to OpenCoffee, I promise you’ll be near one.

At Boston Media Makers, there was sporadic discussion of Twitter advertising, mainly meaning people monetizing their Twitter use by selling ads either in their streams or on their profile pages.  It was generally agreed that profile page ads were less invasive and obnoxious than in-stream ads.  The mavens in the group predicted large-scale unfollowing and extensive antisocial media shunning for anybody foolish enough to try in-stream twitvertising.

I’m thinking it would be hard for most twitter streams to get less relevant or more annoying, my own included.  My initial objection to the whole idea of in-stream twitvertising is that it just doesn’t seem that it would be very effective.  But first, let’s run the numbers, using me as the guineaduck as it were.

Magpie is one twitter ad service.  They tweet ads through your account (tagged #magpie) at a set frequency (such as one ad per five tweets) based on a keyword bidding system.  They pay per tweet, not per click or per action.  Magpie says that I could earn up to 69.07 Euros per month.

Using Followcost, I discovered that I’ve been tweeting at an average pace of 5.19 per day, so I’ll just guesstimate that I would serve up one Magpie ad per day.  At this writing I have 252 followers, so 252 x 1 x 30 = 7,560 potential ad impressions per month. That assumes two probably untrue things: (1) that all of my followers read all of my tweets, and (2) that there are no secondary impressions from syndication of my tweets, such as in the sidebar of this very blog, or from anybody who’s not a follower just reading.  Let’s just say that those two effects cancel out.

Those figures together imply about 9.14 Euros, or $11.63 CPM.  Since that’s the payout to me, let’s mark that up 20 or 25% so Magpie can earn some money, and assume they’re selling limetweets at $15 CPM.  Is that a good price for promoting your product or service in the limedrivelstream?  Honestly, I haven’t looked at CPM priced advertising in a long time, preferring CPC or CPA if I can get it.  It sounds cheap, but there are a lot of reasons why it should be cheap.

In the process of poking around for this piece, I checked my twitter power at Twitter Grader, and found some interesting factoids.   I scored in the 94th percentile, but what’s interesting is that my overall rank is 10,546 out of 255,406.  There are only a quarter million twitterers?  I’ve been so deep in this bubble I would have guessed a lot more.  And if I’m in the top 6% of them, there must be a lot of inactive or totally dead accounts.  I’m sure it’s growing fast, but I have to wonder if there’s enough total market for Magpie and their advertisers to make a real go of things.

The foregoing generally assumes that the twitterati will be willing to sell their real estate, that doing so will not in itself massively devalue that real estate (if people unfollow you for putting up ads, your ads become less lucrative…), and that – and this one is where I worry the most – those ads will in fact make any actionable or measurable impression on the marketplace.

For now, I remain skeptical and the limeduck media empire remains commercial-free.