Sorry Andy, You’re Wrong: Google ‘Search Within A Site’ Is A Big Deal

A while back Google released an update meaning that users would, after carrying out a navigational search for a site, be offered the option to search within that site – on Google. I thought it was a bad idea at the time & I still do.

However Andy Beal, who knows as much about this industry as anyone (& certainly more than me) thinks that The New York Time is over-reacting with its article questioning the new move. I don’t.

I wrote about this move when Gooogle’s search within a site feature first came out and was surprised that it didn’t raise more hackles at the time so I’m glad that it’s now getting the attention it deserves a, to me, it stinks of Google overstepping a few boundaries. I’ll quickly summarise the points I raised in my original post before going on to those the New York Times raises, and why Andy seems, to me at least, to have got this one wrong.

Google is, obviously, presenting this as something aimed at helping users (apparently lots of people search for the name of a site, and then carry out another search to find the particular content they’re after); what I wonder though is whether this has the potential to hurt publishers.

You see, when users come to a site and search for information on that site, it can do two things;

  1. Firstly it can generate revenue; if a site’s commercial model is based on CPM (where advertisers pay based on how many times the ad is ’seen’) then this would mean that ads are less likely to be served on search results pages within the site itself. Instead the search is done on Google which, of course, serves its own AdWords.
  2. It could starve publishers of valuable user data. One of the best sources of keyword research data is studying the words & phrases that users enter into internal search boxes. Now this data will reside with Google and, whilst it will be possible to find it if they do end up clicking through to a page (by checking the log files or analytics) it’s unlikely to be as comprehensive.

I still think that publishers & other site owners have a right to feel aggrieved about any move, no mater how well intentioned, that has the above implications. However the issues the Times raises, which I missed, is in many ways much worse; it actually verges on suggesting that the whole move has rather dubious intentions.

Take, for instance, a situation last week, when users of Google searched The Washington Post and were given a secondary search box. Those who typed “jobs” into that second box saw related results for The Post’s employment pages, but the results were bordered by ads for competing employment sites like CareerBuilder or Monster.com.

So even though users began the process by stating their intention to reach The Post, Google’s ads steered at least some of them to competitors.

Now Andy argues that this is no big deal.

If I search for YOUR COMPANY and then decide to shop elsewhere–after using a site search for YOUR SITE and seeing an AdWords ad–how strong was my relationship with you in the first place?

But the thing that this misses is the fact that this move seems to mean that Google now has a way of showing ads against terms which would otherwise be blocked because they are protected brands. Personally I think this is just sophistry on Google’s part, pure & simple. If I search for a protected brand and then go on to use the site search function, that doesn’t mean that the terms suddenly stopped being protected. In many ways its the exact opposite as Google is, to all intents & purposes, putting itself forward as the brand’s own search function & there is no way that I would allow a competitor to advertise on my own site.

And as for the issue of whether a brand should be strong enough to deal with competiting ads, well… it’s just not relevant. If we accept that many people now use Google as a default browser, even going so far as to type in URLs, I think we can accept that if someone types in Washington Post, they want to be taken to the Washington Post. And, therefore, that if someone uses this new site search, they probably expect to only be shown content from that site. It’s not a matter of brand strength or reputation, it’s simply about usability & user expectations on the one hand, and Google taking the piss on the other.

After all, if I’m using Firefox and type Washington Post into the address bar I’m taken, via a Google search, to the Washington Post site. Christ, I only have to type face into the address bar for Google to assume that I want to go to Facebook. So why, when I actually use Google, should all these hurdles be placed in my way?

I don’t want this post to turn into a rant, so I won’t even get on to the fact that this could deprive retailers of the ability to highlight special offers & the like, which would normally be shown against internal search pages; or the fact that, for this to work, all the pages of a site have to be indexed, so it will penalise sites which have not been optimised (I know that there’s barely any excuse for that anyway, but that’s not the point); or the fact that, shock horror, sometimes Google’s results aren’t the best way of finding things – after all, internal search systems often allow a user to filter results by date, price, category & lots of other things other than a secret algorithm.

I love Google – I really do; it makes my online life better is so many ways. And I guess that’s why I get so disappointed when they screw up like this. The only thing I hope is that users never take to this function (a result that wouldn’t surprise me) and it withers on the vine.

Pirate cupcakes by slushpup on Flickr

Share

16 comments

  1. rishil

    I was quite surprised that many people didnt catch on to the sneaky way google found of cannibalising brand protected searches – which is one of the resons I knocked out that post that was referrenced in the NY Times – it further surprises me that google is kind of hinting a warning – if you take it off – you wont get it back, in the same article.

    In fact I am glad that the NY Times picked up on it – we wouldnt have had the response from google other wise – and its indicative of how important it is to NOT have SiS seeing that Amazon was quick to react and request getting it taken off.

    I am also still uncertain as to the user experience as well – the example of Virgins SiS results left much to be desired too

  2. Ciaran

    Rishil – your post was refererenced in the NY Times article? I missed that (saw Ann’s) – good work!

    Great minds eh…

  3. rishil

    lol – the one they referenced was actually a guest blog by yours truly on Ann’s blog… so you didnt miss it – just missed the small byline saying that I wrote it… lol.

  4. Will Critchlow

    On the indexing problem, it isn’t just poorly-optimised sites that suffer. You may deliberately noindex some product pages because they are close to being copies of others (e.g. old iPod variations) in order to rank better for the short-tail queries, but you’d certainly still want them to be returned in an internal search.

  5. rishil

    “but you’d certainly still want them to be returned in an internal search.”

    Thats an excellent observation Will – I hadnt thought of how SiS will affect PR sculpting techniques.

  6. Malte Landwehr

    I often prefer a search via Google with site:example over the default internal searches of many websites (mediawiki, wordpress and phpBB always surprise me with their inability to find content relevant to my search query).

    I like this feature and even though it might hurt some websites by drawing the users attention to AdWords ads it is (from a user perspective) a very helpfull feature!

  7. Ciaran

    Thanks for the comment Malte – but I’m going to have to disagree with you.

    “I like this feature and even though it might hurt some websites by drawing the users attention to AdWords ads it is (from a user perspective) a very helpfull feature!”

    I think what you mean is that it is, from your perspective, a useful feature. I feel quite strongly that most normal users (and anyone who found this on Sphinn is, by their very nature, not a normal user) will find it a disappointing feature. If you’re expecting to search a retail site, where is the option to filter by price; if it’s a publishing (newspaper site) where’s the filter by section or author; etc…

    For anyone used to using the site: search, this may not be a big deal. But I’d bet that 99% of the online audience have no idea that the site: search even exists.

  8. Dr. Pete

    This seems to somewhat contradict Google’s push to remove search results from search results. That move was good for Google, sites, and users, helping to insure that search results land on deeper pages (such as product pages). Pulling the search box across the Google threshold, though, completely violates expectations. People expect Google to be an inclusive search, not a site-specific search. I don’t see how this helps anyone but Google.

  9. Catfish SEO

    I agree with Dr. Pete. It seems contradictory that Google would tell Web masters to block their internal search from robots because those results don’t represent valid content, only to make them available through their own interface where they can monetize the traffic.

  10. Dave Foreman - Interactive Limited

    Here here,

    Google is an out of control monopoly. Better buy the stock now at 450 down from 700. If you cannot beat them Join them.
    At some point publishers will revolt with a lawsuit and Google will get smacked. But while we argue about some egregious infringements on property rights. Google silently chips away at the fringes by slowly controlling all the tools, data and optimization methods.

  11. Federal Watch

    Although the new feature of Google have advantages..we couldn’t deny the fact that it also have its own disadvantages…I think it’s better that NY Times reacted on Google’s new feature early on…so that Google can improve its new feature..

  12. Pingback: Day Two Recap - SMX Sydney SEO SEM Conference 2008 < Rambling Thoughts Blog - Neerav Bhatt
  13. Gremio

    Wow. I noticed you responded in the comments, so obviously you knew this.

    The site: search feature has existed as long as I can remember. Years upon years. And if the user is too unintelligent to click “search help” (which is the first thing I did when I got heavy into searches around 11 or so) or the current “advanced search” then they have absolutely no business existing, much less using any kind of indexing utility.

    1) Those large corporate companies are not loosing money from AdWords.

    What you’re basically saying is people are too lazy to refine their search on google to find a competing company. If they want to try something different from the major label, that’s exactly what they’re going to do. Whether AdWords exists or not really has little effect on this. It simply gives them another route for finding one.

    2) All those companies have AdWords of their own I’m sure…that show up multiple times when searching for products from other companies and the like. It’s not just a 1-way street here…

    3) Why wasn’t this made an issue 10 years ago?

    4) In my opinion this is just another over-hyped bulletin to seed interest.

  14. Ciaran

    Gremio – thanks for the comment. Couple of thoughts in response:

    if the user is too unintelligent to click “search help” ..then they have absolutely no business existing, much less using any kind of indexing utility.

    This is a joke, right? People aren’t too unintelligent, they simply can’t be bothered – and why should they be? For most people the web is not a way of life, it’s just a tool (and who really ever reads the instructions?)

    Those large corporate companies are not loosing (sic) money from AdWords.

    I’ll just have to disagree with you here (as do a lot of major brands here in the UK).

    All those companies have AdWords of their own I’m sure…that show up multiple times when searching for products from other companies and the like. It’s not just a 1-way street here…

    That’s part of the point; at the time of me originally writing this, bidding on other brands was not allowed on Google. Amazingly, just after introducing this function, they changed their policy. As mentioned above, the lawsuits are now flowing and we’ll see how that ends.

    Why wasn’t this made an issue 10 years ago?

    Err, what? Because Google only just added the ‘search within a site’ box in the last few months. I fear that you may have missed the point of my entire post.

    In my opinion this is just another over-hyped bulletin to seed interest.

    And I entirely respect your opinion. Even if it is wrong.

Post a comment

You may use the following HTML:
<a href="" title="" rel=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>