Question: what does it mean if the demographic of two-thirds of your audience is not your target demographic?

d) You need better market research capabilities.

I would find out if the audience of your audience is your target.

For anecdote: the best target demographic for bars is young males, but as any bar knows, the best strategy is targeting their audience: young females.

Not sure if subtext of question is Smith-Mundt. The audience for publications on America.gov, based on visits, is currently about two-thirds foreign to one-third Americans. One of the anomalies of the law, of course, is that Congress would never try to tell Americans they can't look at a Web site, but Congress does say State Dept employees are not allowed to tell Americans the URL of certain Web sites.

I would agree with Christopher. The chances are that the majority will represent your intermediate audience, who will form a clear line of communication or influence.

Joe is also right in that you then need deeper target audience analysis to determine the susceptibilities and vulnerabilities of both the intermediate and target audiences such that you can have influence.

This can be demonstrated with some of the UK Police anti knife crime campaigns that use a wide base of targeting to affect multiple lines of communication with varying degrees of influence.

George, the subtext is Smith-Mundt but you're off on the medium.

Chris, smart analysis, one that can be somewhat easily tested.

Nick, smart comments (as I'd expect).

The data set I'm referring to - which I'll send up in a post probably tomorrow - is America.gov's twitter followers. Twitter's open api allowed me to download America.gov's followers. The result of the analysis is 67-70% of America.gov's followers on Twitter are US-based. Using Twitter's API, myself or using existing analysis tools, we can determine how often there are retweets and if there's a trend for a particular type of tweet.

So, in our "private" conversation here in the comments, what does it mean if only one-third of your actual audience is the audience you're supposed to engage? Could we then put up a VOA transmitter in Minneapolis and boost the wattage so that some of the signal gets out of the country and says that's good enough? (Minneapolis isn't a random pick... more to come on that)

For the record, I think America.gov should engage the American public. It is interesting America.gov doesn't advertise or have any non-English twitter accounts.

By the way: http://mountainrunner.us/projects/locationstats.php

I'd like to know more about the data before drawing conclusions. For example, are you counting unique visitors to all the distinct language pages on America.gov and the subsidiary embassy vernacular sites that it feeds?

Brian, the data is for Twitter followers not website visitors. It represents who is directly listening to America.gov at a point in time. I could pull more data to put that data into motion (ie. spread over time) by updating the subscribers (America.gov's subscribers have grown by 300+ since last week... which is great) and to check on how many of those users retweet (or forward) America.gov messages as well as check how many non-followers retweet. Unfortunately, as much as I'd like to do that, I probably won't due to time. There are also some freely available tools that do most if not all of that analysis none of them do the geographic analysis that I did.

See also: http://bit.ly/vwISW (which I need to post on)