An unsubstantiated assertion I’ve seen repeatedly over the last months is that Google needs to acquire Twitter because Twitter knows what is happening (or what we’re thinking about) now, while Google can only look backwards. The latest version I’ve seen of this argument is from Jeff Jarvis’s post today, entitled “Why Google should want Twitter: Currency“:
Google isn’t good at currency. It needs content to ferment; it needs links and clicks to collect so PageRank can determine its value.
I grant that PageRank isn’t good at currency. But Google doesn’t need to perform link analysis to know what people are thinking about in real time. Google can simply look at its logs to determine what people are searching for–and, in particular, which search terms and phrases are appearing with statistically significant frequency. And Google’s search volume is much higher (and more representative of the online population) than Twitter’s update and search traffic combined.
To be clear, you and I can’t perform that analysis using the tools Google makes available to the general public. But Google can–and I don’t see any reason, other than the fear of raising public concerns about privacy, that Google can’t exploit this data themselves.
What is different about Twitter is that it *does* make the data available to the general public. Twitter exposes Trends as part of its own offering, but it also enables services like Tweetmeme to perform their own analyses to track the hot stories in near-real time. But Google could do something similar and probably better if it wanted to.
I’ve said this before: Twitter is a community (a social network if you prefer), not a search engine. And, if there’s a good reason for Google to entertain acquiring Twitter, it’s probably that Google has a less than stellar track record when it comes to community. But let’s not delude ourselves to think that Google needs Twitter to know what’s on our minds now. They already know.
The Noisy Channel 
5 replies on “Google Already Knows What You’re Thinking”
Case in point: http://www.google.org/flutrends/
LikeLike
Indeed, I meant to cite that as an example. Though arguably that sort of data isn’t quite as “real time” as what people believe they’re getting from Twitter. But surely it’s an indicator of how valuable it is to be a major search player.
Another point I should have made–you don’t have to be Google. Yahoo and Microsoft may have much smaller market shares, but their traffic volume is probably large enough to do a comparable job of tracking real-time trends.
Unfortunately for them, all of the money is in advertising, where Google certainly wins big on its market share. I imagine they also get higher winning bids from attracting more competing advertisers, but I don’t have any data on that.
LikeLike
To see even more examples:
http://googleresearch.blogspot.com/2009/04/predicting-present-with-google-trends.html
LikeLike
Panos, Jarvis does cite that link–in fact, that’s the starting point for his post. But he insists that:
Twitter is even faster, even more immediate. It collects what we’re doing and talking and thinking of doing right now. I’d love to see Varian et al take its data and put it through their algorithms.
At best, Twitter might have an epsilon advantage when a story breaks on Twitter (though I’m curious how often that actually happens). But, even in those case, I’m certain Google catches up extremely quickly, as people start searching on Google. And, even if the first mention of a story is on Twitter, that doesn’t mean that people will hear about it immediately. By the time the story propagates, Google may have already caught up.
Nonetheless, because Twitter is a public conversation, end users today are likely to see many stories on Twitter first. But that doesn’t mean that Google isn’t already in a position to know about them from analyzing their logs.
LikeLike
OK, did not follow the link.
I fully agree with you: “At best, an epsilon advantage”
LikeLike