Friday 8th of October saw the launch of Google Instant in India. This technology is Google’s latest tool to try and deliver the fastest possible search results for its users. Initially, Google Instant was the core search experience for signed in users of Chrome, Firefox, Safari, and IE8, in UK, Spain, Germany, Italy, Russia, France, and the US. But as of Friday, saw Google instant roll out to further countries: India, Australia, New Zealand, Korea, Ukraine, the Netherlands, Poland and the Czech Republic. So, what is Google instant and how will it impact search in India?
How does it work?
Google Instant generates ‘instant’ results in page and on the fly as you type each letter of your search string. It does this by using the already developed tech of search suggestion. Search suggestion essentially generates potential search queries based upon Google trending data, so if you begin by typing “e” you get “ebay”, ebay being the most commonly searched term to fall under “e”.
The masters of the letters:
As you type into the Google homepage you are instantly redirected to a free flowing results page:
It then generates suggestions and search results based upon its trending data. So, if you were to search for ice cream the results would be as follows:
As ‘ice cream recipe’ is more commonly searched for than ‘ice cream’ it takes the lead suggestion and generates the subsequent results. Therefore, the results that are displayed in the page are the same as if you were to search for the top suggestion.
The technology has been developed with AJAX, which is a tech that allows on page updates as a user interacts. If Google detects a poor Internet connection, due to the bandwidth needed to deliver as-you-type search results, Google Instant will automatically be switched off. Users may, in fact, disable Google Instant (default setting is ‘on’) by clicking on the dropdown to the right of the search box within a search engine results page (SERP).
The resultant effect of Google Instant is a hi-octane intravenous injection of information - “internet on fast-forward” (Charlie Brooker – Guardian UK).
Why?
Well, the reason according to Google is simple. Google Instant is designed to deliver better search results, faster. Google’s technical insight behind Instant is that people type slowly, yet read quickly – scanning page results as they type. By delivering as you type results and including predictive search suggestions, Instant helps users find the most relevant results to a query faster. In all, 15 new technologies are used to bring Instant to life and to aid the functionality of the Google platform. According to Google, the typical searcher takes more than 9 seconds to enter a search term before Instant was released. Google claims that Instant saves somewhere between 2-5 seconds per search, therefore, reducing potential search times by anywhere up-to 50 per cent. Globally, if everyone uses Google Instant, it is estimated that it will save more than 3.5 billion seconds a day – that’s 11 hours saved every second!
So, all in all Google Instant from the outside looks like a great addition for the user, but what about the advertisers that use Google? Let’s investigate what the implications are for Search engine marketers and optimisers.
Paid Search implications
Google Instant – apart from giving us another acronym for search, GI – technically impacts paid search advertising in four main ways: keywords, impressions, click through rates (CTR) and conversion rates. So let’s look at each:
Keywords:
When it comes to keyword selection, the knowledge and understanding of your target market is now more important than ever before. Optimisation of the core terms that your customers use to search for you is still absolutely vital. Google uses its trending data to make the keyword predictions, suggesting the keywords that are most often searched for, are recently popular or have been searched for locally. So Google Trends and Insights should continue to be used in order to account for rising trends in regard to topics and keywords. But what will change is that Google Instant will force a shift in user behaviour towards related searches.
Because of the added suggestions, Google is coercing it’s users towards ‘head’ search terms. There will, therefore, be a shift increase in impressions and clicks on the head terms, not just on the primary head term but across a broader spread of related head terms. Let’s look at the terms ‘car insurance’ and ‘car insurance deli’ as examples:
Keyword: ‘car insurance’
(40,500 searches in India last month)
This is the primary head term for the sector, before we would type the term and press and get one set of results, but now Google Instant presents us results with five different related options. The additional five related search terms are getting valuable real-estate that they were not necessarily getting before. It, therefore, makes sense that these five terms will actually pick up more clicks because of this, dispersing the ‘car insurance’ clicks across a broader range of related terms. The impact of this will be a marked increase of impressions, clicks for the related head terms and therefore potentially an increase in CPC’s of these terms as they become more valuable and therefore competitive.
Keyword: ‘car insurance delhi’
(880 searches in India last month)
This search term comes from the other end of the spectrum as it is more specific and longer tailed. Here is where the biggest shift will be. As the user starts typing, they are automatically delivered results that could potentially answer their query. Because of this, huge swaths of longer tailed terms will be presented results that answer their query instantly before they complete their intended search phrase – the head term results. The long tail results could theoretically be hijacked by the head terms. So be prepared for a potential increase of click throughs and impressions for the head terms, and a decrease in the longer tailed search terms. Longer tailed search terms are generally less expensive and better converting (if targeted to a specific page). These searches will become more infrequent and can now potentially be diverted to the head term landing page rather than the intended product landing page:
This will not be the case for all searches just a percentage and that percentage remains to be seen and something I will research in the coming months.
Google are expecting users to find what they were looking for faster and, therefore, click through on terms which may necessarily not of been the users original statement of intent, but do actually answer their intent. Because of this broader/ longer tailed search terms will be funnelled into head terms, reducing click throughs on longer tail terms and increasing click through on the more expensive volume drivers. Now, if a user is intent on looking for a long tail term, they can still find that term, by adding to the query, or by typing it out. Therefore, click volumes are expected to be the same, but where the clicks go could be markedly different. Google’s goal is to make the process of searching quicker, to help users find what they are looking for faster and to not change what they are looking for. If the head query isn''''''''''''''''''''''''''''''''t going to give users what they are looking for, they will continue their search as they have in the past.
So, in summary prepare for:
1. Keyword impressions and click throughs to increase on head terms
2. Traffic of the primary head term to spread to secondary and tertiary head terms
3. Impressions and click throughs to decrease on longer tails terms
Impressions
Do expect to see marked differences in impressions for some of your key terms. This upgrade has changed the way in which Google registers impressions. An impression can be counted for any of the following:
- After a 3+ second pause
- Selecting a prediction from the suggestion box
- Clicking on a refinement to the left hand side
- Pressing ENTER while inputting a keyword into the search field
- Clicking the ‘Search’ button to the left of the input box
- Clicking on a search result in the search results page (SERP)
As Impressions are used to calculate Click through Rate (CTR), it is probable that there will be fluctuations in CTR benchmarks since more ads are being displayed per search. One must understand that all AdWords advertisers will experience similar CTR shifts, so theoretically everyone is normalised as they are all being exposed to the same thing. There should, therefore, be little overall impact on individual campaigns as everyone has the same issues/ challenges. At present, Google have not determined the degree to which an advertiser’s campaign impressions will change, but it is expected that for some advertisers, the number of impressions might increase or decrease.
Click through rate (CTR)
Although the increased impression rate may slightly shift the CTR calculation, Google is not changing the method of determining the relevancy and quality of ads. Google is going to continue measuring ad performance relative to that of other ad’s for the same query, position, price and if the ad was served using Google Instant or not. So, compelling ad copy and great content on site is still an absolute must.
Conversion rates
If the head terms hijack the longer tail terms, potentially we could see a decrease in conversion as the original Ad designed for the longer tail term didn’t get the impression/ click. If this is the case, user paths may be less targeted and therefore poorer converters. An example:
Nike.com landing pages:
1. Nike air max
2. Nike home
One landing page is the homepage and the other is a product specific page. The user comes to Google with the intent of buying ‘Nike air max’ so begins his search[N...i...k...e...] They then get the results for Nike instantly coming up and so they click on that - the user ends up on the Nike homepage. The user although on the right site is not at the most targeted destination, the best destination would have been the ‘Nike air max’ page, but in this case the head term hijacked the product term. Even though the user ended up on the right site they did not end up at the best page, so the conversion rate will fall as the user has to do several searches on the Nike site to find what they are looking for. Also, generally the Cost per click of a head term is substantially larger than that of a product term, as the user clicked the head term the advertiser could end up paying a higher premium at the cost of a poorer user experience, so potentially higher CPAs.
Google claims that GI helps users build highly tailored search queries and the subsequent results. Therefore, the quality of the clicks may improve, which may lead to increased conversion rates. So, my advice is study analytics and monitor conversion rates, the slightest modifications to onsite user paths and content may negate any negative downturns and may actually potentially benefit conversions, as at the end of the day as always it will be the quality and execution of the supportive content onsite that will drive the ROI and CPAs.
SEO implications
First and foremost, the algorithm is the same, so SEOs still need to do exactly as before to gain position – people who say SEO is dead are just plain wrong. What will change is which search terms an SEO should chase. With the advent of GI we may find new search terms delivering volume and quality to our sites. So focus will change, but the principles of SEO won’t - quality usable content rules OK.
The impact to the SEO ‘search-to-click’ relationship will most likely follow that of the paid results detailed above, so keyword volume and referring keyword reports in Analytics should continually be assessed in order to account for potential shifts in user behaviour. One thing that you should look out for is key search terms increasing in volume, as multiple suggestions are now easily accessed and display during a search phase. As such, the traffic for head terms may be more evenly dispersed between the head term, secondary and tertiary terms that are presented in the predictive results. So chasing the primary head term that everyone has been after since time began may be a thing of the past. It could, in fact, be easier to land grab the secondary and tertiary terms and reap the benefits of the dispersed traffic, after all these terms are getting more prime Google real-estate than they did before.
Something I will be keeping a close eye on is the paid versus un-paid click-throughs. On most screen sizes, the implementation of the Google Instant feature actually pushes the natural results below the fold of the page leaving primarily paid results above the fold:
The previous example shows a ratio of 9:3 of paid versus un-paid results. So we should expect to see a further decrease of natural click throughs on the primary key terms. This was already kind of the case with the advent of universal search, but at least there where images and the like above the fold:
It seems like good business sense to make as much of the results page above the fold paid for listings, all the more revenue for Google, but it does increase the importance of getting a top 3 position for your SEO so that you get a reasonable share of the spoils.
Conclusion
As a user experience and tool, I think Google Instant is great. It looks great and really does deliver a supercharged search experience. So my advice for SEMs is to sit tight, don’t panic and carry on as normal - just make sure you monitor your analytics shrewdly and carefully and only act upon concrete information. Look for abnormal movement on your primary revenue drivers, if there has been a shift look for where that shift is. The traffic is still there, you just need to figure out where. The reality is if you carry on optimising your ads distinguishing yourself from your competitors, and fixing your landing pages to be as relevant and usable as possible you will only stand to gain from these updates.
If you SEO, maybe now is the time to shift your focus, look at what it is Google suggests. You never know you may find some easily attainable low hanging fruit!
... And then there were blogs
- Preeti Chaturvedi, Manager, Mktg & Biz Dev.,
GIST Justin Rabindra, VP, Training & Knowledge Mngt, Ogilvy One Worldwide - 2/25/2008
The ‘Nano’ Challenge
- Sourabh Mishra, Chief Strategy Officer,
TBWA\India Group - 1/29/2008
Too much of a good thing? Research Vs "gut feel"
- Abraham M. Alapatt, Client Services Director,
Ogilvy & Mather Advertising, Chennai,& Practice Group Partner, Ogilvy Business Network - 4/27/2005
Creating the right buzz through WOM
- Prof. Sushil Bahl, Faculty, Marketing Area, Nirma Institute of Management,
Nirma University, Ahmedabad - 2/24/2004
Disclaimer:
"The views expressed are personal views of the author and not
necessarily represent the views of the organisation author works for
or of exchange4media.com."