Practical Ecommerce

Google’s Secure Search Squeezes SEO Planning and Reporting

Google’s secure SSL search protects users’ search results and the keywords they searched on. Unfortunately, it also poses a growing threat to data-driven search engine optimization. Firefox recently joined Google Chrome — and Google.com, for logged-in users — in defaulting to secure search. This has the side effect of increasing the number of “Not Available” or “Not Provided” search keywords in web analytics reports.

In the past, SSL search was too slow and cumbersome to use as a default. Web analytics programs could easily pick out most of the keywords that referred traffic to the site from Google.com. Today, with SSL search the default on Google.com for logged in users as well as the default on Chrome and Firefox browsers, a growing number of Google.com referral strings are coming into web analytics with no keyword information associated. Consequently, if a site optimizes a page for a certain keyword phrase, its ability to measure how many organic searches were referred from Google via that keyword phrase is diminished.

Why Secure Search?

Why is Google accommodating secure search? According to a Google blog post, it’s to address privacy concerns by users.

The blog post says, in part:

As search becomes an increasingly customized experience, we recognize the growing importance of protecting the personalized search results we deliver. As a result, we’re enhancing our default search experience for signed-in users. Over the next few weeks, many of you will find yourselves redirected to https://www.google.com (note the extra “s”) when you’re signed in to your Google Account. This change encrypts your search queries and Google’s results page. This is especially important when you’re using an unsecured Internet connection, such as a WiFi hotspot in an Internet cafe.

The Size of the SEO Data Loss

In 2011, Google reported that it handled 143.5 billion searches a month. Google represents 66.4 percent of the search engine market share, according to comScore’s February 2012 report. The Chrome and Firefox browsers together represent 39.8 percent of the browser market share, according to Net Market Share’s February 2012 reporting. Consequently, 26.4 percent — 66.4 x 39.8 — of all searches last month were conducted on Google in Chrome or Firefox. If this trend holds, 26.4 percent of all searches going forward will be stripped of their keyword data. And this doesn’t include the searches conducted by users logged in to Google properties and searching Google from any web browser.

Zoom Enlarge This Image

U.S. market share of search engines (left) and browsers (right), according to comScore. In Feb. 2012, Google controlled 66.4 percent of the search engine market.  Google Chrome and Firefox accounted for a combined 39.8 percent of the browser market.

U.S. market share of search engines (left) and browsers (right), according to comScore. In Feb. 2012, Google controlled 66.4 percent of the search engine market. Google Chrome and Firefox accounted for a combined 39.8 percent of the browser market.

In Oct. 2011, Google’s Matt Cutts stated that fewer than 10 percent of searches would be affected by SSL search, as reported by Search Engine Land. However, 26.4 percent — plus whatever percentage of searches from logged-in Google users not using Chrome and Firefox — is much higher. Data-driven decision making for SEO is at serious risk.

Keyword Research at Risk Too

If the keyword strings are stripped from the referral strings in SSL search, is the data in the Google Keyword Tool also affected? We’ve frequently recommended that ecommerce merchants use this tool, most recently at “Optimizing a Page for Search Engines, Part 1: Keyword Research.” Despite the risk SSL search poses to this tool, we still can’t recommend a better free keyword tool.

Seasoned SEO professionals always suggest taking this Keyword Tool data with a grain of salt. But if 26.4 percent of the keyword data turns up missing from the keyword tool, what’s the use of using it at all? Keyword research is the foundation of SEO planning, the one window available into the exact words real searchers use when they search for the products merchants sell. If 26.4 percent of that data is gone, an SEO manager’s ability to plan content creation and optimization is diminished. Some long-tail phrases that drive very small amounts of traffic but in total drive a lot of traffic may disappear all together.

The impact to Google Keyword Tool and any other keyword tools that use Google’s data is speculation at this point. It’s possible that the keyword information could make its way into the tool while still being striped off for analytics purposes. But that would violate the spirit of SSL search. If it’s not entirely secure then it can’t be called secure. So I have to believe that this will affect all of Google’s products, including Google Analytics, Google Keyword Tool, AdWords, and others, as well as non-Google web analytics and keyword tools that use Google’s keyword data.

What to Do Now

Firstly, watch the percentage of “Not Available” or “Not Provided” keywords in the search marketing keyword referral reports. Trend it back to at least July 2011, because Google released its latest iteration of SSL search in October 2011 — the Google blog post describing it is here. If the percentage of “Not Available” or “Not Provided” keywords in the search marketing keyword referral reports is not growing for your site, you may be among the lucky few whose audience either skews away from Google products or uses Internet Explorer, Safari or Opera.

If the percentage of “Not Available” keywords does grow for your site, turn to the “Search Queries” report in Google Webmaster Tools. I recently addressed this report, in “Top 5 Reasons to Use Google Webmaster Tools.” The Search Queries report shows the organic impressions, click-through and average rankings for keywords that drove traffic to the site. This report, to be sure, is keyword-based. But does the fact that Google refers webmasters there as a reliable data source indicate that it may not be affected by SSL search? I have doubts, but the ranking data at least should be unaffected in my opinion.

Normally I’m not a fan of using organic rankings as a metric of SEO success because they vary by person due to personalization. Some ranking tools are based on a user’s personal browser and are subject to the same personalization biases, and rankings themselves can change hour to hour. However, if Google Webmaster Tools gives me a true average ranking position for the top 1,000 keywords that drove traffic to a site, and if I at the same time I can’t trust the completeness of my web analytics data, a combination of average ranking and web analytics data seems like the best way forward in reporting SEO performance.

Summing Up

This secure search story is far from over. As more browsers and engines jump on the SSL search bandwagon, the current sources of data SEO professionals rely on may become dangerously depleted. That doesn’t mean SEO is dead. SEO changes as often as the engines and the Internet marketing industry change. It will find ways to work around the issues and get the data to plan and optimize sites for maximum organic search visibility.

Tags:    

Jill Kocher
Jill Kocher
Bio  |  RSS Feed


Get the Practical Ecommerce RSS feed

Comments ( 6 )

  1. seoconcepts March 30, 2012 Reply

    Outstanding article. It will be interesting to see its impact over time.

    Bernie
    http://WWW.seoconcepts.com

  2. brustici April 3, 2012 Reply

    Google’s Keyword Tool used for AdWords is not effected by secure search, those numbers are still being reported. There goes the rub for me as in a very real sense you are required to pay for data on keyword performance. Google is creating some very mixed messages here. It will be hard to measure the performance of organic terms, thus make a case at times to decide if bidding on these terms has ROI.

  3. Bowerick Wowbagger April 3, 2012 Reply

    Google could very easily pass data in terms of raw numbers of searches to its keyword tool, without jeopardizing the privacy of their individual SSL users. This same thing does not hold true for analytics information. When someone comes to your site and you gather data from the analytics, you get much more personalized information (such as IP address, geography, browser statistics etc). To give you the information about what search term they used to get there would entirely violate the spirit of having an SSL search, as you could potentially track their IP and connect the identity of a person to certain search terms. As I said, passing raw numbers to the keyword statistics that they carry does not pass any identity information that could connect an individual user to an individual search term and therefore does not compromise their security.

    With all that said, as an SEO professional, I am still all for this. Privacy and security are paramount. Taking away (at maximum) 36% of data from a GIGANTIC pool of data leaves a site owner with… a GIGANTIC pool of data to still work with.

    If the data that was taken away was entirely random (it is not) then you could still count on that other 64% to provide a very good view of what your customers are searching for and where they are coming in from. The only issue with this whole thing is that the 36% (at maximum) of data that is being subtracted is probably a little bit skewed: a site is probably going to lose data on the more tech-savvy consumers that visit their site, since chrome / firefox and logged in google users are going to tend to be more tech savvy than others.

    However, as I said, you will still have a gigantic pool of data to work with. I am not worried. I will take increased privacy over keyword data.

  4. Strilets Irina April 4, 2012 Reply

    great! I was wondering why the data in keywords tool is different from what I see in Analytics,
    e.g. if check monthly traffic for a keyword for which your website rank in the 1st place in GA and compare it to keyword tool data in Adwords, you can see, sometimes, very big gap …. now it’s clear :)

  5. David Pavlicko April 17, 2012 Reply

    Brustici is dead on the money with this – you can still get the keyword data, IF you use Adwords. It’s obvious that this is a money grab to pacify their investors and has zero to do with privacy.

    I’m quite surprised at Bowerick’s comment – as an SEO, he’s all for this, because it’s more secure? Please. If someone wants to search anonymously, every browser today offers that option already. That has nothing to do with why Google is pushing this, IMO.

    Certainly, you can still obtain decent insight from the available data, but masking the keyword queries creates other problems for someone who’s trying to improve the usability of their site for their users. Say a popular page on your site has a high bounce rate, and the primary medium visitors reach that page is via organic search. What are they looking for that you’re not providing? Good luck figuring that out.

    You may say that the Google Webmaster Tools provide search query data that you can use…I’ve found it to be wholly inadequate, and often totally wrong. Want to find out which organic keywords lead to transactions? If you sell technology related products, there’s a big chance that ~30% of secure searches will be more like ~60% +.

  6. grover002 March 5, 2013 Reply

    This implies also that the statistics I get through Clicky.com about visits from my adwords campaigns, do not tell me which keyword the ad was found for. So Google may show my ad for incorrect keywords and I will never be able to tell, I will be asked to pay for the CPC. I would at least want to know what I pay for, wouldn’t you?

Email Newsletter Signup

Sign up to receive EcommerceNotes,
our acclaimed email newsletter.

And receive a free copy of our ebook
50 Great Ecommerce Ideas