Every so often, we welcome a OneRiot staff member to share some words on our blog. We have to be careful with this - because everyone’s always so high on dry-erase markers - but most of the time it works out pretty well. Here’s the latest - a deeper look at Twitter search from OneRiot GM (and acid house evangelist) Tobias Peggs.

It’s about time.

Before I got sucked into the internets, I spent several years as a magazine journalist mainly writing about music (and the occasional supermodel). Now Carmel has convinced me to start writing for the OneRiot blog on a regular basis. My initial thought is to write about what I’m doing – which essentially boils down to helping grow OneRiot, trying to keep up with a triathlon training schedule, and ambiently learning from other cool companies (like Twitter), collectives (like Cypher13) and random cool dudes (like @Jeffrey) that I stalk via social media but who probably don’t know who I am. Let’s see how it goes… and see what you think.

This first post is focused on the first theme. We recently launched a new twist on Twitter Search. We learned a lot from doing that – from so many angles. (I’ve got future posts in the mental pipeline on “doing a launch with and without a PR agency – pros and cons” and “how I was schooled to be authentic, by a social media pro who would hate to be called a social media pro.” But more of that later…). For now, the focus is on what we did and why.

Let’s start by echoing John Battelle, who has acknowledged the growing appetite for search results from the “super fresh” web. The web’s most valuable content increasingly can be found in what our friends and other people are saying, sharing and looking at online. Traditional Search Engines (TSEs) just can’t keep up with this new era of “people preference,” solely because their systems of gathering information don’t really support it. Broadly speaking, TSEs index the web based on variations of a web page’s “juice,” which builds up over time. So, logically, TSE’s will struggle to surface fresh, socially-relevant search results because the new, interesting stuff hasn’t been around long enough to build enough juice to rate highly in those indexes. That’s the hole (and it’s a big one!) that realtime web search engines like OneRiot are now filling.

Given its current – and justified – buzz, Twitter Search tends to be peoples’ first port of call to consume search results from the real time web. If you search for something like “Pontiac” on Twitter (ahem, RIP) you’ll see an explosion of conversation from thousands of people in relation to that query – which is pretty exciting to see. In other words, Twitter Search surfaces the conversations happening right now around that query. Your search results experience is essentially telling you “Your query is a hot term! Lots of people are talking about it!” If you parse through, or join, the conversation, you might find out why.

If you do that same search on OneRiot, the experience is quite different. OneRiot’s Twitter search surfaces the content that people are talking about right now, so your results experience is essentially telling you “Check out this page, or blog, or video! Lots of people are talking about it!” It’s search results as you know it, but the content is drawn from the realtime web (i.e. the results are based on what your friends are say, sharing and looking at online right now).

Now, of course, seeing – and joining – the conversation is a key piece of Twitter’s huge appeal. So you’ll also notice with OneRiot’s Twitter Search that we do show the conversation as well – but we group the tweets that have shared a link to a specific piece of content. (Of course, we de-dup all the shortening services for you, so links from bit.ly, tiny.url, tr.im etc that point to the same piece of content are treated as one by OneRiot).

Here’s where the two search opportunities really differ: From Twitter Search you could join a worldwide fast flowing conversation about Pontiac. From OneRiot’s Twitter search you could join a targeted conversation about a specific piece of content – for example, the NPR story that started the buzz (this is the top search result on OneRiot Twitter Search as I write… of course, because this the realtime web, it’ll probably be different when you look).

Our take on Twitter Search was a good experiment, and you’ll start to see some of the awesome things we’ve learned from it filter into the main OneRiot.com search experience soon. In the mean time, if you have any ideas or feature requests, feel free to @tobiaspeggs and let me know, or log them on our community wiki.

(That wasn’t so bad, was it?)

Tobias - you get a gold star.