Firefox: Improving privacy and eating analytics data for breakfast

Posted by Mal Darwen on 10 May, 2012
View comments Search News
Read about Firefox's move to pass Google searches via https - meaning SEOs will see even more (not provided) keyword data in their analytics report

Search query data from Google Analytics has been invaluable for SEOs for years. When Google introduced hidden search queries in October 2011, there was uproar. Since then, there have been a couple of developments. Which you're probably not going to like.

The search query data that analytics provides has been invaluable to SEOs for a number of years, and a fair fuss was kicked up when Google introduced hidden search queries in October 2011. Since then, there have been a couple of developments ...

Mozilla's Firefox has become one of the more popular browsers on the planet in recent years with around 20% of the market share (depending on who you ask). Like many browsers (certainly the popular ones), Firefox offers a search bar already baked in, which along with the huge developer community around it developing fantastic tools  and add-ons makes it a very flexible browser to use.

The news from Mozilla's privacy blog earlier this week that searches made in Firefox's search bar or in the 'Awesome Bar' will now be hidden from analytics' users is good for web privacy campaigners, but not so great for those who rely on keyword data coming into Google Analytics, or, for that matter, any other analytics tool.

It's well worth remembering that Google have funded much of Mozilla's development of Firefox, and there are murmurs that Mozilla may lose that funding, so there is some speculation that this may be the spur for their shift to Google's way of thinking on hiding search queries from website owners.

We've already seen the curse of "(not provided)" in our keyword data since Google started hiding searches made by signed-in Google users (to varying degrees):

keyword not provided by Google Analytics

For some companies the impact has been far greater than for others: up to a third for some sites. Wordtracker's own (not provided) data is rising and doesn't show any sign of slowing down.

progression of 'not provided' keywords

This figure looks set to rise dramatically in the near future, as Firefox's update means that they'll be sending all search queries to Google via https (which means along a secured route), which in turn means that any visits to your site from a Google search from Firefox's search bar will be reported as (not provided) rather than giving the search query to you in your analytics.

What does this mean for SEOs?

What this means for SEOs struggling to find the best keywords is that they'll have to start to find a new way to establish their best target keywords - in the absence of search query reporting in Google Analytics, this means more A/B testing, deeper keyword research in the beginning, and hacking analytics data even more than at present. Avinash Kaushik published a method for analyzing the (not provided) data which is handy for those comfortable with customizing dashboards (there are some good dashboard resources around)

There is keyword data available in Webmaster Tools, but the general feeling is that this data is loose and inaccurate, so perhaps it's time for SEOs to find other ways to make sure they're targeting useful keywords and not stabbing in the dark. Of course, keyword research is going to be as important as it ever was, and keyword-heavy offline campaigns for larger organizations (in the UK think of the 'search for "Army Jobs"' TV adverts the military have been running for some years, or Orange's 'search online for "I Am"' campaigns) may start to play a larger role in corporate site promotion.

Why are Google encouraging this?

There are bound to be theorists writing about Google's strategy with this - are they trying to increase AdWords revenue (PPC search terms are still reported in Analytics)? Are they trying to clean up their privacy mojo in the wake of the Street View cars debacle? Are they just collecting as much data as possible for their own purposes? Are they planning to get people to move to a paid version of Google Analytics?

There's a saying which goes "If the product is free, you are the product". We're seeing increasing evidence of this with things like Facebook sharing our data with developers and marketeers when we 'like' their pages or use their apps. Analytics data has always been provided for free, and while we still have access to huge amounts of (reasonably accurate) data, I don't think we should be too surprised when this free information ends up being curtailed in one way or another, stinging though it may.

Google clearly has interests in a 'better internet' for users, and is concerning itself (publicly at least) with fashionable (but nonetheless very real) privacy concerns, but its real motives are going to remain hidden from us, at least for the foreseeable future.

Why do you think Firefox has taken this route? How do you feel about the way search data is treated in analytics? Share your thoughts, let us know your view. Leave a comment: join the discussion.