The One Big Thing: Real-Time Conversational Context In Search – The Twitter Search Deals

The One Big Thing you need to know about this week is that Google and Bing have cut deals for access to real-time conversation data from Twitter (in the case of Google) and both Twitter and Facebook (in the case of Bing).

Bing already has a beta implementation of Twitter search up and running at bing.com/twitter.

While I’m certain Google will also implement their own version of Twitter search as a separate search category, as Bing has already done, I don’t know what is particularly new about this aside from exposing Tweets to a larger audience. We already had, after all, Twitter Search itself as well as other tools that tracked trending Twitter topics (no alliteration intended) and popular links.

Where it gets really interesting is when you combine Twitter and/or Facebook data with general search results so you get a real-time idea of the conversations around either the topic of a given search or around the links returned on the search results pages.

If you ever needed a compelling reason to get on the social media conversation bandwagon, this is it.

I’m still trying to figure out all the implications, but for communications professionals even the obvious ones are huge.

Reputation management has the potential of getting a lot trickier overnight. Search results for product recalls, for example, could display not just news stories and press releases but also what people are saying now about the product, the company, or even their experience as the victim of a faulty product.

If Google and Bing do decide to integrate results, which, according to this article, appears to be the end-game for Google, a great many more opportunities will be created. Google’s example is real-time, on-the-ground Tweets about weather conditions at a ski resort.

There are plenty of questions which will have to be sorted out. First among them, I think, is the question of relevance.

The search engines have become super smart about matching what you ask for with hyper-relevant content. But the content they are giving you, by and large, are Web pages. Individual Web pages offer many indicators of relevance to search engines: The text on the page, the order in which it appears, the labels given to multimedia content on the pages, links to that page and the text used within those links, domain names and suffixes, tags applied to pages from the author or from a visitor using a bookmarking service like Delicious.

With individual tweets, all you have are at most, 140 characters of text, links within a tweet, replies to that tweet, retweets, favorites of a tweet, bio and geographic information about the author of the tweet, etc. That is a lot less information upon which to determine relevance.

The other issue is the question of authority. Google became what it is today in large part based on its ability to determine which were the most authoritative sources of information for a given search and to push those to the top. How do you measure the authority of a Twitter user?

These questions will be resolved but for now, all you need to know is that you need to pay attention to this.