• jeroen

    and should my clients advertise on incomplete keywords like “mortgage r” next to “mortgage rates”?

  • ho_logos

    @jeroen – no, the paid ads target the query whose results are being displayed, not the partial query you’ve typed.

    @johnwellis – long tail effect is the million dollar question, however I do see great potential for more long tail queries, not less, because of instant. 5 star hotel shoppers probably won’t stop after typing las. However, las vegas 5 star hotels on the strip is one of the recommendations, and that is one that the searcher may not have queried until she saw the suggestion.

  • http://seoroi.com Gab Goldenberg

    The questions I have – and perhaps SEL can put them to G – are as follows:

    1) How will those impressions/clicks be reported? If I’m bidding on Las vegas hotels and Las vegas 5 star hotels … which gets the impression/click/conversion?

    2) This might make the short tail [appear artificially] more profitable [by attributing long tail demand to the short tail], as well as more difficult to analyze. Any guidance in that regard?

    3) How does this affect QS?

    I also second Jeroen’s question.

  • http://www.JohnWEllis.com John Ellis

    Jeroen – Thanks for the comments.

    No, I don’t see a need to bid on incomplete words. Google is assuming the completed version and will gray the text of the assumed word.

  • http://www.JohnWEllis.com John Ellis


    That’s an excellent point about artificially inflating the long-tail vs. short-tail numbers. I guess we are all at the wait-and-see stage.

  • dev3ine19

    I really don’t see a huge impact to how this will effect the long tail. In all honesty it seems to just be making it easier for the user to get to search results when they don’t really know what they are looking for. For example they might start with a general idea of getting hotels in Las Vegas but seeing what comes up will eventually lead them to narrowing their search faster. It really comes down to real time editing of your query to get to exactly what you want without having to sift through pages of search results.

  • blueoceanmarketing

    You have to remember that search is intention-based. People go to search on something because they have something in mind already. Just because Instant shows them something, doesn’t mean that they will stop there and click it. In fact, I might speculate that many people will find Instant annoying.


    Is there a way to turn it off?

  • http://www.JohnWEllis.com John Ellis

    @dev3ine19 – Thanks for the comments. I am hearing others say it may not have an impact on long-tail either. I am still not convinced. Maybe users like you and I will keep searching, but I think novice users (those not visiting SearchEngineLand.com) are going to get out as quick as they can. If they see what they are looking for (as in Las Vegas Hotels), why keep looking?

    We will just have to wait and see how this plays out.

    @blueoceanmarketing- Matt McGee just wrote a great article on Google Instant. It includes how to disable it. You can find it here: http://searchengineland.com/google-instant-complete-users-guide-50136

  • dev3ine19

    @John Ellis I see what you are saying, but Matt McGee makes a point at the end of his article, hopefully this will teach people to start searching smarter. And really, to your comment; if people find what they are looking for on the short tail then were doing our jobs properly; getting the customer to where they need to go.

  • jeroen

    here’s the 26 lucky ones, once you type in one key google suggests them:
    best buy
    dictionary (ok this is potential)
    jet blue
    quotes (another chance)

  • http://www.rimmkaufman.com George Michie

    John, good points, but I think what will happen is a positive selection bias.

    The Keyword “Las Vegas” performs poorly for hotel chains because many of the people searching are looking for “show tickets”, maps, the moving Leaving Las Vegas, flights, insurance quotes, veterinarians, whatever. If someone is looking for a hotel in Las Vegas and stops their search after typing Las V because they’ve seen a hotel ad that’s attractive, that’s likely to be a valuable click.

    I think we’re going to see an increase in the value of the traffic because of the selection effect, and I think folks who weren’t looking for a hotel will keep typing until they find what they want. Fewer “bad” clicks on the head terms, more targeted clicks on the tail = win + win.

    All sheer speculation, of course! We shall see.

  • http://convertingux.com vperr

    I think the main issue may be with SEO rather than with PPC, even if I do share the concern about need to bid on broader keywords and relative increase in cost.
    I think this is the best way for google to devalue SEO results…people will not want to scroll and 80% of the results visible on a normal 1024 resolution (still the most used) will be paid results.

    Guys, don’t you see this as the most astute move google could make to put even more money in their pockets??

    I can’t wait to start comparing the organic traffic on our ecommerce website for those keywords where we were ranking from position 2-3 down, but I am not expecting anything good.

  • http://www.rimmkaufman.com George Michie

    vperr, I think you may be right. The paid search piece kind of takes care of itself through smart bid management. If more of the quality traffic goes to the head keywords, the value of the traffic will increase as will the bids, if more flows through to the tail the bids will shift appropriately.

    In SEO it may mean if you’re not in the top 5 forget it. May be true in paid search as well.

    Interesting stuff!

  • http://www.glynndevins.com SeniorLivingSEM

    John – thanks for bringing to light the impact Google Instant might have on paid search. Couldn’t it also negatively affect quality scores? If an advertiser qualifies for more impressions but doesn’t get clicked, will Google continue to ding the account?

    Personally – I’m not crazy about Instant, but maybe that’s just because it’s not what I’m used to.

  • http://www.twitter.com/GregBogdan Greg Bogdan

    I think that the impact of this will be muted by the fact that:

    1) Most people are not be logged into a Google account when searching, I am guessing < 20%, despite the popularity of Gmail and other tools that require a login and often result in a persistent login.

    2) Many people have a Google search bar at the top of their browser and won't see the instant results unless they do a second refinement search, if they are logged in.


  • http://www.JohnWEllis.com John Ellis

    @jeroen – Thanks for the list, but that’s not the same list I have. I get a whole new set of keywords. For example, for “w” I show Walmart and for “j” I have Jet’s Pizza. The question becomes how do we get to be chosen for that letter?

    @SeniorLivingSEM – Thanks for the comments. Quality score may be an issue as well. I think the bottom line is we just have to wait and see.

  • http://www.alanmitchell.com.au alanmitchell

    Will Google Instant make people lazy? Will searching become a 100% keyboard affair of key strokes and down arrows, with the mouse only being used when the user has decided exactly where to click?

    If so, above the fold results will be massively biased, with scrolling on Google becoming a redundant activity.

    Surely this will raise the CTR of above the fold ads, while massively reducing the CTR of ads which appear below the fold? And won’t this devalue ads in the lower positions (7, 8, 9 and 10), forcing lower-ranked advertisers to increase bids if they are to maintain click volume?

    It’s not just the long-tail which is threatened, but also lower-ranked ads. It’s the smaller advertisers who will feel this more. Larger brands who can afford the top positions will simply welcome the increase in traffic.

  • http://www.indulgemedia.com paulwould

    I’m not too sure that Instant will have an effect on the long tail, if you are only bidding on an exact, or phrase match for that matter, then your ad should only show up for that query. Many will continue to bid on these long tail, exact keywords and only those who can afford it will bid on the broad match, which was the case before Instant was launched.

    Users will continue to drill down to the search that they want, if looking for a hotel in Las Vegas, many users will realise that the first hotel result for Las V… will not always be the best, people like hunting for the best deals and hidden gems, making the long tail as relevant as ever.

    As for SEO, I am concerned that results below the fold will be affected, making the need for solid SEO effort ultra important, will be interesting to see what happens! Very interesting article and comments as well by the way!

  • http://sanderton sanderton

    If you’re not a touch typist (and no many people are) you’ll never see the intermediate search pages as you’ll be looking at the keyboard, like me now!

  • http://www.crealytics.de tchrist

    Hi John,

    interesting question. I do think that the focus will shift towards the suggested search terms by google. Long tail keywords will lose importance because people will stop typing after the first characters of the word to see if there’s already a search phrase matching what they are looking for.

    The implication of that will be, that the CPC for the suggested terms will go up, because more advertisers will bid on those “predefined” search terms. Which means more profit for Google and less profit for the advertisers.

    Long tail won’t be dead, but it will lose importance. So the window (Google Search) through which we see the world (internet) is becoming narrower with Google Instant. Some other thoughts: http://www.crealytics.de/blog/2010/09/09/implications-google-instant-sem-tchrist

  • jon-

    A someone who has a heavily optimised “long tail” site with 100’s of P1 positions for long tail keywords (think whole product names rather than brands) I’ve been worrying about this overnight.

    It’s now nearly noon local time and things are not looking good. I’d estimate traffic is down around 50% from looking at my basic “real time” stats. 50% is huge.

    I’ll dive into analytics tonight and see what I find. I’m not optimistic, and logic immediately told me this update would be seriously bad for those relying on the long tail.

    Nice one google.

  • nmvalente

    Hi John,

    One of the things that is concerning me is the tendency to compete for the same keywords that Google chooses to be revealed or listed. Won’t this increase CPC dramatically for those keywords, given added competition?

    Won’t you also agree that Google is deliberately reducing the visibility on Organic results by:

    1) Using the drop down suggestion tab to push away organic results.
    2) Making users quickly go through the first results (paid search area) to access if they should continue their query or not.

  • http://www.JohnWEllis.com John Ellis

    @alanmitchell – We are on the same page. There may be a threat to lower-bidding and position as well. Of course, that will force us to spend more money to compete. Thus, eliminating the small-business budget and allowing the bigger budgets to win. … but like everything else, we just speculating now.

    @paulworld – I am not convinced that many “will continue to drill down to the search that they want”. Maybe you, me, or other people in our industry will. However, I am not sure if the average user (my Mom for example) will drill down.

    @tchrist – “more profit for Google and less profit for the advertisers.” Good point.

    @nmvalente – It does look like a de-emphasis on organic rankings. There is a whole set of problems on the SEO side of things that we didn’t even get into here. Excellent points, thanks.

  • OSD

    Does anybody know what the percentage of Google Toolbar users are vs. overall Google search users? I’m guessing it’s low and typically reserved for savvy Internet users.

    Seems any traffic you depend on from the Toolbar, mobile web, Google apps, etc. will be safe from Instant … for now.

  • http://WWW.ELLERTONWHITNEY.COM ellertonwhitney

    Google knows how to line their own pockets, which this should surely do. Not good for savvy marketers as this should logically decrease the long tail. Google is becoming less of a search engine and more of an sheep herder. And as fewer and fewer people submit actual searches, it will only get worse.

  • RobertManly


    I agree with your assertion that this will help teach users to search smarter overall but I think the long tail traffic aspect is the bigger issue. I strongly disagree with the idea that this will decrease long tail traffic overall as this suggestion tool teaches people to be more specific in their queries in order to see the results they want. What Instant is really doing is drastically reducing the amount of time that users will spend scanning a SERP for ads, which puts the onus on the advertiser to write compelling search ads to that speak to all levels of a given funnel, from head to long tail.

  • ChrisGale

    Robert, Totally agree with your assessment…

    John, this means you’re going to have to add back in those broader keywords and just make sure you’re not bidding to highly for them, since they DO have lower conversion rates. I expect that the longer tail more specific keywords are still going to drive conversions and people will be searching for them, they’ll just have a lot of them suggested to them.

    Interesting thought… I wonder if there is any aspect to the potential value to Google of a paid search click to determine which search queries, and therefore ads, to display. Do you think there is a revenue optimization play on Google’s side?

  • http://www.JohnWEllis.com John Ellis

    @Ellertonwhitney thanks for the comments. I guess time will only tell. Ultimately, they are in business to make money, but to do that they also have to make their advertisers happy. I expect them to adapt and adjust as data comes in.

    @RobertManly users like us, those who visit SearchEngineLand.com, are trained to extend our query to get a more precise result. However, I don’t think we can expect the average user to do the same thing. If they see a paid ad that answers their question, why would they continue?
    That means as an advertiser, I need to be there … or else.

    @ChrisGale – Yep, agree. Adding back broad is a must for sure. Of course that means we will need to spend more money. Like you said, surely those will have a lower conversion rate. So,we spend more money to get less. That was the whole reason we paused broad to begin with.

    That being said I think, and hope, that it will all work itself out. Patience may be need.

  • http://mst3.wordpress.com Doc2626

    Great article, John, and also some good points brought up in the comments. To me, although Google has always stressed their “the user comes first” philosophy, I think it’s fairly obvious that revenue takes a front seat. I don’t see how this can fail to drive us to put more money in their coffers, by bidding the more competitive phrases.
    I’ve also wondered how the new data they’ll gather will eventually work its way into their semantic web plans. Chicken… or egg?

  • http://www.thisweekinrelevance.com davesch7

    Interesting article — however, I think given the ability for users not to cycle through so many queries so much faster, the long tail will actually thrive — here’s my reasoning:

    By allowing users to filter/alter their search results so quickly, organic and paid search listings will have to stand out that much more in order to generate a user response. A search session which in the past generated only 1 or 2 unique SERPs before the traffic left Google, may now generate 5 to 10 or more unique SERPs in the same amount of time. If marketers could not make their unique offers stand out previously (and many could not, delivering static and often irrelevant copy that did not match user intent), then they will really struggle to deliver relevance in this new environment. Additionally, users will more easily be able to refine their queries with additional attributes and qualifiers, taking them deeper into the long tail. Marketers who have rested their laurels on targeting only the highest volume queries in their category might be surprised to see their traffic drop until they adopt more comprehensive long tail strategies. As with the SEO case referenced above, marketers who can quickly adapt to the changing behaviors borne of these changes will most certainly benefit from them.

  • http://www.JohnWEllis.com John Ellis

    @Doc2626 – thanks for the comments. Google mentioned several times during the announcement that they put the user first. I agree that should definitely be the case. But from an advertiser viewpoint, I am afraid it means more spend … which of course I will do.

    @davesch7 – Excellent points. Thanks for the comments.
    This is where I am not sold yet: “users will more easily be able to refine their queries with additional attributes and qualifiers, taking them deeper into the long tail.” – we are assuming the user doesn’t find what they want. But in the Las Vegas example, they see hotel ads. So they are finding what they want. That means from an advertiser view point I need to be there for “Las Vegas” if I want to compete.

    But that being said – I have to think it will all work itself out eventually, at least I hope.

  • http://www.alanmitchell.com.au alanmitchell

    Hi John,

    Thought I’d explore the implications of Google Instant on searcher behaviour further, so wrote an article exploring how Google Instant could make us lazy:


    I think whether or not the long tail is dead will ultimately come down to the range of search suggestions Google offer in their Autocomplete drop-down list. I think people will come to accept Google’s search suggestions as finite, and the number of creative / unique / ad hoc searches would steadily decline (already there is some evidence of this as I point out in my article).

    If that is the case, I don’t think the long-tail is necessarily dead, but it is certainly limited to some extent by the comprehensiveness of Google’s search suggestions.

    On the other hand, if people take a completely different approach to Google Instant altogether, and instead increase their demands for even more specific, detailed results, then the long-tail could in fact grow.

    Great discussion you’ve got going on here – some really good comments.

  • http://www.erocket.co.uk erocket

    I can’t help thinking this will definitely impact the long tail, the only unknown being by how much. My own thinking is that Google Instant will make head terms more important, and reduce the long tail.

    For example, say a user is looking for novelty gifts for men. They start typing ‘novelty gifts for men’, and by the time they get to novelt (sic) the search drop down shows an option for the head term novelty gifts.

    The impact on the novelty gifts long tail then depends upon how many people click on that initial, broader listing, with the intention of exploring the SERPs and possibly refining their search once they visit a website ranked for novelty gifts.

    If that happens often enough, the novelty gifts long tail (for men, for women, for girls, for boys, etc etc) shrinks and Google start to report higher popularity for the ‘novelty gifts’ head term.

    If head terms become more prevalent in this way, the long tail will surely reduce (in volume if not in the actual breadth of terms). The top SERPs for head terms will then become even more important and competitive (especially with the first organic listings getting pushed further down the page), and ranking for head terms will become critical to some businesses currently sustained by their long tail traffic.

    People have commented that, as users type, they’ll see alternative options displayed and this might educate people about the long tail and the benefits of refining a search beyond the head terms. But these options presumably come from historical search data, so these alternatives could shrink in time? Also, again using my novelty gifts example, having typed in ‘novelty’ the only related term displayed is ‘novelty gifts’ – in other words you have to start typing long tail to see long tail.

    Just my tuppence worth.

  • comradity


    First of all, thanks for raising the Long Tail issue. In my opinion, the ability of the cream to rise to the top via search has been deteriorating rapidly. So the impact of Google Instant Search is a continuation of a trend, or worsening of the trend.

    I don’t understand why the media, especially a publication like this, isn’t asking why the dominant leader in Search is pursuing faster search instead of better search. I mean is search advertising generating 100% click through rates :)? If not how is faster search going to improve that?

    Katherine Warman Kern

  • Adam Dost


    While not an expert in SEO or PPC, I immediately thought the same thing as your hotel search analogy. IMO, this will lead advertisers to be more specific when bidding on the broad and short keywords. If my hotel advertisement is designed to capture the person looking for 5 star hotels on the strip, but my ad campaign is designed for the keyword “Las Vegas”, I should capture the lead before others who are trying to reach the same person but not bidding on that term. The key difference is trying to accomplish the same thing that Google is, determining what the client is thinking.

    So yes, this can have a huge affect on long-term keywords as advertisers and SEO starts to target and optimize towards the broad terms with targeted PPC to capture the client before they can even finish typing the long-term. CPC should also be minimized with proper ad wording as the ad can still target the long-term keyword.

    My initial thoughts,


  • http://www.JohnWEllis.com John Ellis

    @alanmitchell – Thanks for reading. Google Suggest is going to play a big part in this, but just not sure what yet. I guess time will tell.

    @erocket – yes, right on track. The user will stop typing once they see results. I know I would. After all, that’s the point to speed up the process. That would kill the marketing and the demand for long-tail

    @comradity – Google has focused on faster results. The may not always say it, but it’s always been a top priority, from the very beginning. But you raise some good points. At what point are we losing quality to gain speed?

    @Adam Dost – Thanks for the comments. Yes, at minimal advertisers are going to have to rethink strategy. Re-thinking strategy is never a bad thing. It’s always worth taking a second look at marketing plans and techniques.

  • PPCSwede

    Very interesting article. I do think, however, that the impression count will and has increased since launch last Wed, mainly because of the three second rule.

    Consider making a search on Google using Instant, you start off your search query and interesting (or irrelevant) results appear. You pause and take a look at them, 3 seconds is a very short time, but then realize that you’re not finding what you’re looking for. So you complete your full search query and have now doubled the impression count you would’ve created with the old system. I have noticed an increase in impressions and a clear reduction in CTR for all campaigns. There is however no clear indication that the long tail keywords are losing any momentum.

    The interesting question is of course if this will lead to an increase in CPCs (Google’s hidden reason I’m sure). We’ve been repeatedly told it won’t since it’s status quo for everyone but we’ll have to wait and see for the results of that.

    I have switched off my Google Instant for now since the function adds little value to my search behavior. I just don’t like the idea of Google, through their probability algorithms, try to determine what I’m looking for. Maybe we should just let Google determine what we’d like to search for as well?

    Great discussion guys.

  • PPCSwede

    Also a small note to consider. Paid Search is center stage with this move. Google Instant ensures that paid listings get more space on the SERPs while natural search falls below the fold (do a search and you’ll find that only the first two listings appear while searching, using Google Instant).

    This will logically lead to more paid search clicks for Google and perhaps, for some, enhance user friendliness. However, in my humble opinion, this does more to increase Google’s bottom line than improve search user experience.


  • ngowdy

    My big worry with Google Instant is that in decreasing the length of search queries it will negate the value of my negative keywords.

    In my campaigns, there are a large number of harmful keywords that expose themselves in the long tail. If searchers aren’t compelled to continue typing, they wont reach these words that disqualify them in my eyes. As a result, I’m going to be getting a flood of impressions on root terms that I otherwise don’t want.

    I’m also already seeing advertisers that *DO* target these keywords start bidding on the root terms from which they normally shy away. I foresee this causing a ton of confusion by forcing a wide array of advertisers to shove into the same root terms.

    Take for example, [800 numbers]. The results, historically, are flooded with 800 number phone service providers. [800 number reverse lookup] is an entirely different concept for which I don’t want my ads to show. But some pioneering reverse lookup providers are starting to be more aggressive in getting into the [800 numbers] root in order to catch the users attention and short-circuit the query before their would-be competitors show up.

    I’m not saying this is sustainable, just that it’s happening. It has me thinking I should try bidding on keywords like [800 flowers].

  • seolsearchkaty

    Are there any stats around about how many people search from their browser v’s google/IE/firefox etc. toolbars? I imagine google will inevitably find a way to add Instant Search functionality to these, but for now, searches via the toolbar don’t bring up Instant results. I wonder what % of people search from the google homepage and therefore how many people this will affect. Also, does this affect any mobile or ipad google aps?

  • http://www.JohnWEllis.com John Ellis

    @PPCSwede thanks for your comments. I am not ready to say impression count has increased on my ads yet. It’s still too early to tell and too many other factors at play. I’d like to give it at least a month before I give that a definitive answer. I do agree that paid search gets exposure, but in a sense, that’s part of the problem. That exposure will force us to bid on those broad terms to compete.

    @ngowdy I hadn’t gave much thought to the negative keyword list. Thanks for bringing that up and thanks for the comments. It seems to be changing the game either way.

    @seolsearchkaty Katy – thanks for stopping by. I have the Google toolbar myself, but surprisingly I don’t use it as much as you would expect. However, that could be just me. Google did mention mobile in their announcement last week, they expect that to roll out over the next few months. To what extent? I am not sure.