Instant Search: Google's Biggest Ever Mistake?
Last week Google rolled out Instant Search, a major overhaul of the way pages are retrieved and displayed and the SEO community is still coming to terms with the news. I talked about the facts of this earlier – Google Rolls Out Major Search Update – so I won’t go over the nuts and bolts again. Predictions of the “Death of SEO” have been made in dramatic fashion by some bloggers, though this is categorically not the case. Nothing to do with ranking metrics or algorithms has changed. However, searcher habits are likely to change as a result, perhaps significantly, and search marketers will need to adapt their strategies to reflect this.
Big News in Search Marketing
It needs to be said that retrieved-as-you-type search results being rolled out across Google is massive news, and news I would imagine few anticipated. It’s difficult to think of any other changes to Google search of a similar significance in their history – perhaps the introduction of PPC advertising, and along a similar vein to this update, the addition of Google suggest. I said last year that personalised search was big news, and it is, but Instant will have far more immediate and wide reaching effects.
Not New Technology
While this technology is new to the vast majority of people, it is not a Google innovation – it’s actually something they have copied from Yahoo! – Retrieving and displaying results as queries are typed is not a new idea and was first tried by Yahoo in 2005 at instant.search.yahoo.com (now defunct), and separately as part of their “alltheweb” search engine in 2006.
However functionality was not as fast or efficient in those days and Yahoo! did not roll out the changes on their main search engine, limiting it’s scope of usage greatly.
There’s also a currently live AJAX powered instant search engine that has been around for a year or so which uses results retrieved from Bing, at istartedsomething.com/livesearch.
What Effect Will the Changes Have on Surfing Habits?
Google Suggest was a feature likely to be a time saver for many in terms of saving keystrokes and could sometimes be an aid in finding information. However with Google Suggest, a surfer would have to consciously decide to use the suggested keyphrases in order for them to be searched. The previously linked Bing api search engine displays results for typed words, not for suggested terms, and this is the major difference of Google’s implementation compared to previous ‘instant’ search.
For example, searching google.com for “Cheap Car Insurance”, results will be shown as the query is typed for:
Chase (as in the bank)
Cheap Car Rentals
When the surfer has typed “cheap car i”, “cheap car insurance” is the first auto suggest term, the results for which will be displayed.
A side effect of this is that a searcher who may have continued typing to create a longer tail search term, for example “cheap car insurance deals” or “cheap car insurance Boston” could see results which are of the kind they are looking for anyway and stop typing before going to the longer tail.
Some might offer up the point that since many do not know how to touch type, they would not be looking at the screen while typing. Long term however, I think search techniques and habits are likely to change, with searchers who do not touch type learning to look up at the screen at intervals.
The searches these changes will not effect are those that aren’t carried out from google’s search engine itself – such as those from toolbars and address bars. The firefox default search bar is currently responsible for just over 9% of Google search volume.
How Will This Impact Search Volumes?
1. Anything that is the first suggestion in Google suggest will be likely to see significantly more searches, both as a result of people pausing while typing queries, and because of these terms being suggested, in effect, more aggressively than before.
2. Long tail searches may become rarer, particularly for popular search sectors where many suggestions exist. Many small to medium businesses, particularly affiliates, base their whole operation at the moment by targeting less competitive but still popular variants of searches and taking a portion of traffic that the big player(s) can’t cover. This may no longer be such a viable business model.
How will this Impact PPC?
Something that is clear from spending time with the system is that it will probably focus searcher attention more onto organic results. The PPC results also change in real time, but the only thing that stays common across searches is the organic results, meaning that they are likely to be what the success of a query is judged by by most users, and it stands to reason that most searchers will stop refining their query when they see an organic result they wish to click on.
I would expect PPC impressions to increase, but clicks to decrease. If this is what happens, expect Google to respond with some tweaks to the system.
How will this Impact analytics?
It’s been pointed out that when results are served by AJAX, there will be no way for the keyphrase to be included in analytics. However this is not true as Google have implemented an intermediary page containing referrer information which acts as a redirect to the destination page, after clicking on a search result. So analytics should not be directly affected.
Impressions for searches should not increase dramatically across the board as Google requires a three second pause in typing, or a clicking action to count an impression on a SERP.
Prof it Motivations?
I have always found the statistic that a quarter of Google searches have never been seen before an interesting one that shows the health of information retrieval on the Internet. I can only see these changes leading to a decreasing number of ever more popular keyphrases for most commonly searched topics. People will be less likely to search for products and services with a myriad of different phrases, search volume will more likely be concentrated into 4 or 5 big terms for each case.
This could have a longer term side effect of increasing apparent authority and quality of search results to the layman, as the big players concentrate their link-building resources on these particular keyphrases, squeezing out lower quality and smaller name sites. PPC compeitition for some keyphrases may increase substantially.
I think this is Google’s vision for the future – it’s all about profit – concentrating search volume and PPC competition (which is currently spread out more than they would like) into a smaller number of the most popular and commercially valuable searches. A side effect is presenting sanitised search results – lessening the chance of a searcher unwittingly choosing an unusual variant of a keyphrase which has a malware infected site ranking top, or unhelpful link farms in the results.
Is this a Mistake for Google?
Personally, I dislike this change for ideological reasons. I feel this change may cut out some of the opportunity that Google previously gave small, savvy businesses. It is likely to make SEO in many cases a tougher game with a greater emphasis on the short tail. Creating content to target a variety of keyphrases may be less worthwhile and some sites could find the number of keyphrases by which searchers land on their site decreasing significantly.
I feel there is a movement by Google towards sanitisation of search – evidenced by the recent removal of negative terms from suggest (the reason for which becomes apparent with the Instant update) – for example “Apple sucks” is not a suggestion when searching for “Apple”, though it used to be until recently. In fact, once a searcher types the c in “apple suc”[ks] (and loses “Apple Support” as a possible suggestion), no results are displayed without pressing enter first.
This kind of sanitisation is not noticed by most, but some are likely to object. Google search appears to be headed in general towards a greater concentration of ‘safe’ commercial results from big brands for many queries, with this change and the previous implied site search update. This could create a niche for an ‘alternative’ search engine which is more likely to return a wider, less commercial range of results, akin to those that would have been seen from an equivalent Google search in the past.
Finally, as I touched upon earlier, such changes seem more likely to cause searchers to switch from Google to another search engine than the other way round. If Bing had implemented this idea, it would be much more understandable as they are playing catch up and need flashy features to attract attention. For Google there is surely a greater element of risk.