Our attention spans are getting shorter. While being constantly interrupted or distracted by new toys, new sites, new things, new messages, new people, and new ideas, we're delivering a stream of consciousness into a void which quickly gobbles it up and waits for more. Scads of sites are helping us find out what is trending in real time, and deliver us the "now". What's hot now. What's peaking now. What's being said now. But the focus on the cutting edge of time, combined with reduced capability to search our history is making finding what happened "when" a lot harder.
Consider for a moment some simple questions that pertain to the social Web, blogging and lifestreaming - and our lack of great tools to discover the answers. I'll start with some real examples.
1. What was the first blog to cover the service backtype.com?
2. When was the first time Denton Gentry left a comment on my blog?
3. When was the first time Robert Scoble mentioned the word Twitter in his blog?
4. Who mentioned FriendFeed on Twitter first? MG Siegler or me?I use these specific questions with real people to help consider questions that may pop up from time to time. Substitute your own curiosities.
There are a few places we've trained ourselves to look for such answers, namely Google Web Search (and its competitors like Bing), Google Blog Search (and IceRocket, my favorite), and Twitter Search. For individual blogs, we have to hope they have integrated blog search tools, such as that from Lijit, or Google's Custom Search Engine (CSE), a subset of results from one domain. But each of these has some weaknesses.
First, Twitter Search. Two years ago, when I was a more grumpy person, I targeted the search engine for being quite broken, only going a few days deep and failing in more specific searches. The archive hasn't been built to go deep in the last two years, and the accelerated volume of posts isn't helping the issue.
Second, Google Blog Search. You can search by date, and even set custom ranges, but the archive of blogs is both incomplete, but overly full of low-quality content. So finding out a question like the first blog to mention BackType requires you to visit their own blog and hope they mentioned it, or to just know they were funded through Y! Combinator, and thus, got started with TechCrunch, back in August of 2008. Peeking at Crunchbase helps too in this regard. Icerocket offers search by date, but I didn't find any results in August of 2008, so that's a miss.
The Web has collectively accepted these sites inability to accurately deliver on archive search, with an expectation that questions such as this are only done by a small few.
For the second question, one I posed to Denton last night out of curiosity, I would have to hope one of two things - that Denton had used Disqus.com with my blog and that I could somehow see all of his comments on my domain through Disqus. But looking at his profile there only shows how many comments he's made on my blog, and that's not clickable. If Disqus has the data, it's not easy to find. In fact, the only way we arrived at the answer is that I'm enough of an information hoarder to have every comment he's ever made tucked away in my email and was able to search locally on my laptop to find his July 2008 comment on a guest post by Jesse Stay. He finally posted on an entry from me in August of that same year.
So what about Mr. Scoble mentioning Twitter? I tapped Google Blog Search, went to Advanced search, entered scobleizer.com as the domain and the keyword of Twitter and chose 2007 as the year. The results nicely omitted his URL as part of the query - so that had no value. I searched his own site, using Lijit, but there is no sort by date. One of the first results included mention of an earthquake on Twitter in 2007, but knowing Robert, there's no way that's the first mention. Just can't find it.
As for who mentioned FriendFeed on Twitter.com first, MG or me? That's something of a trick question. We already know Twitter search won't find it - nor will its realtime search partners Google or Bing. But I can guarantee that MG mentioned it first because I didn't even join Twitter until early 2008, several months after I had been using FriendFeed. By that point, MG was very active on Twitter and had no doubt mentioned FriendFeed several times. But this all requires knowledge outside of the search engines.
Wikipedia and Quora and Crunchbase are all human-powered knowledge bases that are being built to answer questions such as these that capture more than the "right now". That we have access to things like Google and Bing and Icerocket that find answers to so many things is fantastic. There are even archival sites like MemoLane (which is beautifully done) to capture your history, but the when is becoming a footnote in history. Will the big engines let the flow of social media and blogs outrun the ability to capture it?