IN THE TRADITION of “People who bought ‘Assmasters’ also bought ‘Assmasters II,'” Twitter has chosen four of my Twitter friends and is presenting them to me as being “Similar to You.” Pray what does this odd-in-this-context phrase, with its “Related Products” vibe, mean? Does it mean if I like myself, I would also like these people? Surely not, for I already know that, as demonstrated by the fact that I follow them. Were they chosen for discussing similar subjects (e.g. design, web design, CSS, semantic markup)? Unlikely, as that would imply Google-like keyword data mining and analysis bordering on artificial intelligence.

Then, what? It can’t mean people whose tweets resemble mine, as the Twitter writing style and frequency of the listed friends is purely their own. People with whom I have followers in common? That seems most likely, but it’s just a guess.

I’m curious to know what Twitter and its new CEO (hi, Dick!) mean by this. What is the marketing purpose of this feature? Am I to view Twitter as an informal “personal brand analysis” service? That could be cool for me and for the four people who are “Similar” to me. But surely most users would be uninterested in such a service, unless, unbeknownst to me, nearly everyone who uses Twitter is a marketer who views it primarily as a channel. And most companies don’t spend money developing long-tail features, of interest only to a tiny fraction of their users.

I love Twitter. I wish I’d invented it, and not primarily because if I’d invented it I’d be taking the Japanese women’s gymnastics team on a round-the-world cruise. I wish I’d invented it because it is something really new on the internet, like the web, and filled with potential, like the web. As a designer, I pay attention to Twitter same as I do Apple, Google, Flickr, and Facebook. The new feature intrigues me precisely because its language feels “off” and its purpose eludes me.

Also of interest, although less so: what data is being used, and how is it being analyzed?

What’s your theory?