Recency in Real-Time Search: It's Much Broader Than Just 'Right Now'

Editor's Note: First off, what would you like to see covered at the latest Search Insider Summit in Park City, Utah, Dec. 2-5? We want to make sure it's as relevant and timely as possible. Please take our quick survey to let us know.




Having just come off a great discussion about real-time search on an OMMA Global New York panel last week, I gained a sense that my co-panelists and I are as much in the mode of educating as we are talking about strategy (which is fine given the somewhat recent emergence of the topic in marketing and ad circles).  So today I want to back up and talk a bit about the new R-word in search, "recency," and how an expanded definition of what we think is "real time" will also expand the real-world marketing discussions on the topic as well.



Real time is about what's happening right now  -- and yesterday, last week, last month, tomorrow, and next week.... Real-time search is often generally perceived as being about "what is happening right now," and this is simply not the case in terms of its potential usefulness to searchers and marketers from a freshness perspective.  Within the framework of being relevant, what ultimately sets real-time apart is matching a query in terms of recency, or in other words, content that is sourced by fresh and timely indicators, either technically or socially, or both.  "Recent" could mean "right now," "in the last 2 hours," "today," "yesterday," "last week," "in the last month," etc., as it best suits the query.   And recent data can also tell us about what is happening, or expected to happen in the future as well. The big news here is that search with a real-time spin takes time into more of an account, which is a tremendous shift from only seeing "anytime" search results over the entire history of all digital assets.

As many as 40% of total search queries may be better served by a real-time result. Various reports suggest that the amount of standard search queries used today that could be better met with time-factored results is in the area of 20-40% (see this report from One Riot, which claims the high end of 40% of all search queries being suitable for real-time search).   Not everybody is searching for something going on "right now," though; they just want the freshest and most relevant result as it may relate to a certain time period.

So here are a few example query types to illustrate how real time may better serve results that are not only happening and informed about "right now," but also happening and informed about today, last week, and over the last month:

1)     Last week, last month.  Coupons and deal-related searches are great examples of commercial search with a high volume more suited toward parameters of recent weeks or months, or as long as the deal is valid.  Let's face it -- nobody wants to do a search and find expired coupons or deals at the top of the results, but this still happens too often. Smart search advertisers are already planning for this level of recency using paid search, right now. 

2)     Yesterday, today, and tomorrow.  Closing the gap a little bit, an example of real-world real-time search that is not "right now" but still recent, may be looking for things to do, or finding out what's happening this week or today in a particular locale (or what happened yesterday that wasn't captured at the time).  So if I'm in downtown Dallas for the weekend, a search for "things to do in Dallas tonight" would be better represented with current concerts, events, restaurants with recent reviews, etc., as opposed to the typical list of mainstay things to do in Dallas, such as going to museums or the zoo. As another example, if you are at a two- or three-day conference, the conversation spanning the period may be totally relevant toward what you are looking for, but still not exactly "right this moment."  Conversation and feedback about other current events can also be highly relevant to a particular search.

3)     Right now.  This level of search goes beyond mere searching and finding, more along the lines of capturing the developments in a breaking news story like Michael Jackson's death or a natural disaster or emergency, as it immediately happens in the moment.  Hands down, the best example of this form of real-time search discovery (an entirely different search animal, I've argued) is in Twitter search, though a Facebook stream can also be useful. But it still lacks the openly public and keyword-driven stream that allows everyone to access it.

"Right now" is an area where a major crawler like Google needs help from a social layer. While everyone on Twitter knew what was happening when the Jackson news broke, Google admittedly thought the surge in Michael Jackson queries was an automated attack.  Certainly Hot Trends is an area to continue to watch (added to mainstream Web results on Monday), as Google already has the trending topic data down pat, but still struggles with applying the proper context in a short period of time.   But the area of "context" is where I predict Google will make great gains, through crawlers and better processing of social signals - which may ultimately make the company the hands-down leader in real-time search.

So in addition to relevancy, get used to the new term recency.  And keep in mind that it means a lot more than just "right now." As more realize this, hopefully the general discussion will go beyond the perception that real-time search is just about Twitter alone.



Next story loading loading..