Google Launches SSL Encrypted Searching

The following, as with anything else found on this blog (my personal blog) are my own thoughts and do not reflect the views of my employer or anyone else but myself.

As I am sure anyone in the digital marketing space will have heard by now – Google announced in the past couple of days that they would be launching SSL encrypted searching as the default for logged-in searchers on Google.com. There are many implications of this change but I will leave the technical analysis to this article which I believe does the best job of summarising what it means and what the impacts are. The long and short of it for we marketers is as follows: natural search traffic will appear in analytics without keyword referral data and will appear simply as “not provided”. This was done as a response to some malicious tracking and identification from unsecure Google searches and was done on Google’s end as a change to the search engine rather than as a change to Google Analytics.

Single digit numbers?

Being in the UK we are at present seeing a fairly small number of referrals changing for any of the sites that I have looked at, but I do expect to see that change over time. I think the “less than 10%” or “single digit” figure provided by Matt Cutts to Search Engine Land (in the above link) grossly underestimates the number of true users there are as a percentage because that quote was made in reference to “of searches” I believe, rather than “of searchers.”

As evidenced by folks like Martin MacDonald with his query manipulation discussion in a handful of presentations he has done over the past year, there is a WHOLE lot of search activity (searches) that are not done by humans at all. I would suggest the true number of searchers that are logged in is at least double what we’re being quoted in terms of a percentage of human searches. At this point we are looking at a much more sizeable chunk of the market which will have a large impact when you actually want to look at conversions and true USER behaviour – but we’ll have to wait and see what the true percentage of valuable search data this impacts I suppose.

This will almost certainly change the way we report (though in the UK probably won’t be for a few more weeks until it really hits hard) and brand vs. non-brand will be one of the first things to go in my opinion – as mentioned by Hugo Guzman in the following post.

Checks and balances?

For me though, the slight inconvenience to us as marketers is not my main source of contention it is how blindly Google have defended the move (even some of their best known data lovers). I know that there were some dangerous loopholes and so forth (see additional resources below) but I also know that organic search marketers have now lost our only way to verify or test on our own Google’s assertions “PPC adds incremental traffic x% of the time” or anything to do with branded versus non-branded testing. All of that would now be statistically insignificant and there is no check on that power or independent verification possible anymore. This will only add further speculation and mistrust in Google’s figures about search behaviour as there is no genuine way to verify any longer.

Further to the point, the amount of hypocrisy associated with applying this to organic results and not paid results seems like playing with fire to me and I would expect a strong response from a lot of watchdog agencies on this front. If this was truly undertaken as a means of protecting privacy this surely needs to be applied to all data – which would also be a considerable blow to Google’s position in the provision of web analytics (which coincidentally they have recently started charging large sums of money for premium accounts).

Unforgivable?

Lastly, it makes me a bit upset the degree to which this has been defended. There would be plenty of ways to pass referral data on in a more secure manner without some of the previous exposed risks with tracking and identification – if it was in the interest of privacy I could not see this being rolled out only to logged-in users it would be to everyone and paid search results would not be excluded.

The potential implications of this are massive to competitor analytics platforms as well as organic search marketers – something that is probably not a big concern for Google – however, the biggest single issue with this announcement for me is the fact that it seems to go directly against Google’s stated preference for webmasters to create great content that is of value to our users. If we can’t know what keywords provide the greatest satisfaction for our users and those keywords that are most likely to lead a user to make a purchase how on earth are we to target our pages and improve our content accordingly. Unfortunately, the logic that follows, is that this data is only readily available if you’re willing to pay for advertising.

This is of course a complicated and complex issue and at the moment I think there have been a lot of overreactions – however, I cannot help but stress my concern about the direction in which Google appear to be heading with decisions like this and cannot see how this particular “fix” is considered the best way to deal with the original issue that triggered the problem.

This feels to me more like a warning shot across the bow of SEO practitioners, affiliate marketers, and loads of other webmasters and though it may not be that far reaching yet, this could have huge implications.

The most important and most difficult part:

SEO would not and will not die as a result of this. What really sucks is this hurts the genuine SEO agencies that take their job seriously, that actually report on their work and create clever reporting and optimise in a manner that provides value to their users and leads to “satisfied searchers” (i.e. ones that convert). Taking this away doesn’t make SEO go away, it makes it harder for the good guys in the industry to continue to provide a professional service.

The fact that we can’t track the results of our work as easily won’t make webspam go away or even be that much less effective – the results will still come from whatever methods are working for a given search engine at a given time and the revenues will still stream into the pockets of those that optimise best, but most won’t know how to communicate what is working.

It looks like Google has moved first, it will be interesting to see what this growing battle might amount to.

Edit: What they’re not talking about (!)

One thing that no one seems to be talking about and was pointed out to me by one of my colleagues at my day job, Rob Hammond, is the fact that whilst Google are no longer sharing the referring keyword they are still sending the site that was visited. If the concern was truly with privacy and the fact that people might see your search history, surely it is no better that they can see the website visited even if not the query. We will leave the graphic details aside but as Rob pointed out, if the concern is that your family, employer, loved ones, etc. might tie a query back to you as an individual for the term “redacted” surely if they could see the fact that you subsequently viewed the website “redacted.com” this would not be any better.

Lastly, as Rob also raised, in addition to the fact that this doesn’t really close the loophole that was allegedly the initial concern, it leaves another potential loophole wide open to be exploited with the data that is passed.

Additional Resources:

https://www.eff.org/deeplinks/2011/10/google-encrypts-more-searches

http://searchnewscentral.com/20111019195/Latest/dear-google-this-is-war.html

http://econsultancy.com/uk/blog/8163-google-ssl-encryption-for-search-queries-the-experts-view

http://googleblog.blogspot.com/2011/10/making-search-more-secure.html

http://www.google.com/support/websearch/bin/answer.py?hl=en&answer=173733

http://twitter.com/#!/avinash/status/127153191824539648

Comments

  1. Think the privacy issue comes from sites being able to reconstruct profiles of people (IPs) based upon the search phrases from server logs, as opposed to users of a computer looking up browser history (after all there are also plenty of initiatives to keep that private). Of course I can’t imagine there are many sites big enough to build up any sort of useful profile.

    At a guess the SERPs and Adwords already pass data, whereas the SERPs and GA don’t, hence we have one and not the other – or at least some half-ass attempt using GWT as the intermediate.

    Reply
  2. Hi Chris,

    As I’ve understood it there is/was a problem with both. The “show me your cookie and I will tell you who you are” concern is addressed here: http://planete.inrialpes.fr/~ccastel/PAPERS/historio.pdf which then also could lead to identification of user click history as well as contacts.

    Additionally, the other concern I am aware of was the concern about the Google suggestions about individuals and the potential disclosure of private user information VERY easily: http://planete.inrialpes.fr/~ccastel/PAPERS/historio.pdf

    Now, something that I didn’t mention in my post but I actually find quite interesting/ironic in all of this is as follows: both of those papers (also cited by the EFF) suggest that the TRUE problem comes as a result of Google’s attempts at personalisation rather than having anything to do with whether or not the keyword referral is passed. Admittedly, that it is passed, is part of the problem but only because of the way it is being done.

    Now, again this is a massive oversimplification but in all seriousness the TRUE reason for these insecurities, as I’ve understood them, have more to do with Google’s unyielding desire to display personalised results rather than whether or not keyword referral is passed… what do I deduce from this? The aim is (at least in part) to move away from links, rankings and other stuff SEOs understand and can replicate – which throws us other problems. As a result, SEOs (and loads of other innocent bystanders) will be negatively impacted. Who am I to say if that was intentional or a somewhat fortunate side effect, but more food for thought!

    Thanks for your comment,
    Sam

    Reply
  3. It always amuses me when SEOs talk about ‘organic’ vs. ‘paid’ search results. If a gambling company pays an SEO agency for a high ranking for a search on “Help with gambling problems” (and I’ve worked in the gaming industry – I know this happens) does that count as a paid link or an organic one?

    Fight your corner and defend your business model, sure – no-one’s going to criticise you for that – but don’t pretend that you’re fighting the noble fight for quality, free exchange of good ideas and the user experience. Anyone who’s searched for ‘Cheap airfares’ and clicked through to the appallingly lazy and unusable farecompare will respond with hysterical laughter.

    Reply
    • As someone who actually does turn down work regularly for ethical reasons (i.e. gambling, aggressive targeting of people trying to find a solution to a problem, reputation management to hide things that I believe are in the public interest, etc.) I don’t really appreciate that you paint with such a broad brush.

      You can continue to believe whatever you like about “what an SEO does” and there are plenty of people that do things to cement these beliefs so it’s perfectly understandable. But please don’t try to discount my honest concern with Google’s monopolistic behaviour simply because of the industry in which I work and which ultimately is a part of a much larger marketing set in which I’m involved.

      Of course I have my own interests as well – namely data – which is why I’ve offered nothing more than a point of view here.

      There’s nothing noble about my argument but I do personally think that this was handled very poorly and rewarding advertisers and effectively crippling everyone else (for the record not every website that ranks or receives and values data from Google does so because of manipulative behaviour). For me, this goes very much against Google’s crusade for the user and given how careful they tend to be about not favouring advertisers for fear of legal recourse or potential lack of trust in their results this to me doesn’t seem in line with their stated goals and objectives.

      I value and respect your opinion and I don’t pretend that I’m proud of a lot of things that individuals within our industry do, but I would ask that you try to suspend your disbelief as we are all capable and free thinking individuals who can approach a “similar” role with a very different approach.

      Reply

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>

Terms & Conditions