The Future of Real Time Search – Twoogle?
So far my musings have centred on Twitter purely because it is the most utilised example of the real-time search concept in action. There are plenty of other people at the party but they draw heavily on Twitter’s functionality.
Google’s own Eric Schmidt has enthusiastically discussed partnering with Twitter by integrating Twitter feeds into the Google SERPs. However, Schmidt was keen to play down talk of a full-on acquisition.
Other participants in the real-time phenomena such as Scoopler also draw heavily on Twitter data.
The additional space for “popular content” is a great idea, however, in my example above it is not yet 100% functional. Looking for David Cameron, who has recently indulged in some casual swearing on Absolute Radio, we see the expected tweet stream awash with posts including a video. However, when selecting the videos link from the popular content panel we are told there are no videos yet!
OneRiot also makes use of Twitter posts as well as online content that is being shared by the masses.
The good news here is that the video filter is working. However, I would expect the most shared content to be tailored to target the user query in order to make the experience more relevant. Overall OneRiot feels a little clunky and does little to enhance user experience.
For me Twitter is the catalyst to something much larger and more significant for search and the web as a whole.
Companies such as Collecta, Relegence and OneRiot are busy putting flesh on the bones of this debate, however, there are considerable problems with putting a true real-time solution in place.
The Future: Definition and Evolution
If the question a search platform needs to ask itself in the context of the traditional model is,
“What is relevant?”
Then the question a real time platform must ask is,
“What is relevant right now?”
To give context to this then consider the events surrounding Michael Jackson’s death.
In the traditional search environment the most relevant result for “Michael Jackson” would be the Michael Jackson website. However, in recent weeks this query would have been entered by people looking for information on the memorial event and the recent issues regarding his personal physician. Therefore it is important that search technology understands the intent behind each query.
Some may say that this is catered for to an extent in our current search solutions but for true real-time information as-it-happens we need to see big changes in technology.
At a recent search panel attended by Matt Cutts, Danny Sullivan et al there was talk of the sheer heavyweight computational requirements of such a task. Not only would engines need to see everything posted everywhere, they would then need to filter this data to determine:
– Uniqueness of information
o 5% of registered Tweeps generate the majority of content.
o Within this, think about the volume of re-tweets generated from a particular article of interest.
o The solution needs to account for the obvious relevance of a re-tweeted article but also discount the content as being duplicated.
– Relevance of content
o As with the example above regarding Michael Jackson’s passing and the evolving events surrounding it
– Intent of contributor
o Suppose 1000 people all post the same link on Twitter. Although the content and intent of the linked article are the same the intent of each user could be different. Think of Cristiano Ronaldo’s departure from Manchester United. There was a lot of opinion surrounding this and these differing opinions need to be considered.
o Some people here may prefer to make a distinction between real-time and social search (What is happening right now vs. what people think right now). I believe that the two go hand-in-hand for a truly relevant search platform. This links back to my opening on Twitter and brands connecting with users.
o What you say about an event should also be considered alongside the event itself.
One way of addressing this problem would be to apply a timescale to all data. Potentially, a real-time engine could examine all information gathered within a 5 minute timescale as relevant. Although this is still a mammoth undertaking from a technological standpoint we can immediately rule out masses of content which is not “of the now”.
Just how instant can we go?
Danny Sullivan labours the very valid point in his article What is Real Time Search? Definitions and Players that qualifying as real-time means as good as instantaneous availability on the web.
“For me, “real time search” means looking through material… where there’s practically no delay between composition and publishing. You take a picture and seconds later, it’s posted to the world to see. You think of something, immediately tap it out on Twitter, and your tweet is shared almost as soon as you thought of it.
What’s NOT real time publishing? Blogging, for the most part. A post has to be written, which typically will be at least a few paragraphs long. It may involve some research, taking more time. It may involve a ton of research, taking even more time. The mere act of creating and publishing the post will likely take more than a minute, if not several.
Publishing in minutes isn’t the same as publishing in real time? Nope. Not when the time to publish a tweet is seconds. You see something, hear something, want to say something, feel an earthquake happening – you bang it into a simple box and bam, you’ve microblogged.”
This represents a major challenge for real-time providers and the reason why Twitter has provided the backbone for many of the emerging platforms out there.
Wishlist: Commercial Application to PPC
When this hurdle has been navigated we are left with a very exciting evolution of search with massive potential as a business model.
In search we are all too aware of the link between TV spots and campaign volumes. A real-time model could allow businesses to leverage themselves at key conversion opportunities alongside their offline activity.
For instance, if I am running TV spots between the hours of 19:00 and 21:00 for car insurance a real-time model should be able to recognize the proportional uplift in searches for my brand in comparison to the norm and therefore place greater relevancy on my ads by discounting my CPC via it’s QS component.
This shift in functionality will surely excite those who often ask “So what?” when they are told about the link between offline and online. If Google and, indeed, search in general is all about relevancy then it is the next obvious step to start introducing this functionality into the paid search arena.
Latitude is one of the largest and longest-established digital marketing agencies in the UK, with expertise in SEO, PPC, display advertising, social media, conversion analytics and affiliate marketing. Original article is here.
Comments
Leave a Reply
You must be logged in to post a comment.