Say Goodbye to Paid Search Terms from Google

Google’s official announcement confirmed the end of sharing paid search query data with web analytics tools such as Adobe Analytics. After Google removed organic search query data last September for security reasons, most digital marketers knew the other shoe—the blocking of paid search terms—would drop at some point. To be honest, Google looked hypocritical when it only stripped search query data for organic searches and not for paid searches. However, the popular search engine probably moved faster than many of us expected (or hoped) when it announced the removal of paid search terms this month, especially as many digital marketers were still struggling to come to terms with the disappearance of their organic search keywords.

To clarify what this means, Google will strip the search query information from the “q=” (q=search+query) parameter in its referrer string. The search query terms are used by digital analytics tools to populate their search keyword reports (yes, even Google Analytics). For example, if someone clicked on a paid search ad for the search term “analytics hero”, previously the referrer string would look like this (I’ve simplified it):

http://www.google.com/url?q=analytics+hero

Now, the referrer string will look like this:
http://www.google.com/url?q=

Google’s Paul Feng stated that advertisers will “continue to have access to useful data to optimize and improve their campaigns and landing pages,” but now you can only get the search query data from Google through either its AdWords search terms report or the Google Webmaster Tools Search Queries report. If you want to tie your paid search terms to a conversion metric, your only option now is to implement Google’s AdWords Conversion Tracking pixel (with a fraction of the rich insights you previously received from your digital analytics tool). Hmmm. Are you seeing a pattern emerging here?

It’s beginning to feel like you’re playing basketball against another team that calls all the fouls and timeouts, controls the scoreboard, and always forces you to play skins (I always hated playing shirtless in gym class).

What does it mean for my Adobe Analytics reports?

Regardless of whether you believe Google’s decision was really about providing more secure search to its users or a savvy business move to expand its AdWords business, it will impact your Adobe Analytics data in the following ways:

With the elimination of paid search terms from Google, it’s going to be more challenging to perform keyword expansion and refinements for your AdWords campaigns (such as identifying cost-saving negative keywords) without relying on Google. You will still have search query data from other non-Google search engines to offset the loss of Google’s search query data. Based on data from Adobe’s Digital Index last year, in the US Google generated 80% of the search traffic followed by Bing (9.5%) and Yahoo (7%). Regrettably, other search engines have shown a tendency to follow Google’s lead rather than defy the search giant and march to a different beat. Most likely all search query data will eventually dry up as other search engines replicate what Google is doing with its search encryption.

Due to Google’s dominance in the search engine space, losing its data represents a troubling blind spot for digital marketers. If you’re worried about your Google search traffic specifically, going forward it will mean more guesswork and greater dependence on Google for its search query information. I’ve seen some people state this change is a non-issue as Google will still provide its search queries report in AdWords. Obviously, they are ignoring the rich insights that digital analytics tools give advertisers if they want to better understand segment attributes, pathing behaviors, micro conversions, etc. Cameron Cowan, Product Manager for Adobe’s Media Optimizer solution, noted that search marketers will lose valuable data from their digital analytics tools such as geography, language, time-of-day, device types, operating systems, products browsed, and a host of other important dimensions that could previously be used in campaign structure and targeting decisions.

How is this different from losing the organic search terms last September?

Fortunately, in the case of paid search, you’re not left completely in the dark like you were with organic search terms. With organic search, you have no other means for capturing or measuring the actual search query data in Adobe Analytics. If the search queries aren’t passed to your site by the referring search engine (Google, Yahoo, etc.), you can’t measure the search terms against your engagement and conversion metrics—end of story. In contrast, with paid search, you at least have another method for measuring your paid search efforts—campaign tracking. If you use Adobe Analytics to track your paid search campaigns, you’ll still be able to understand how they perform for your digital KPIs.

Now most companies are not measuring their AdWords campaigns down to the keyword level in Adobe Analytics. To get insights at the more granular keyword level, you would need a unique tracking code for each exact match phrase, which isn’t feasible or recommended when companies are bidding on hundreds of thousands or even millions of keywords. While I don’t feel the sting is as painful as losing organic search terms last September, it still means marketing organizations can’t be as precise or efficient with their paid search spend as they were in the past.

Note: If you’re an Adobe Media Optimizer (AMO) customer, you will still be able to access reports for keywords you’re bidding on within AdWords. A targeted keyword associated with a particular paid search ad is different than the actual search query term that an individual types into Google to trigger the AdWord ad. For example, someone might search for “dog shampoo” (search query term) and click on my paid search ad for “shampoo” (targeted keyword). Only the search query terms reports (Search Keywords – Paid/All reports) will be affected by “Keyword Unavailable”, not the targeted keyword reports from your AMO integration with Analytics.

A final eulogy for the Search Keywords Reports

As I reflect on the recent changes to search keywords tracking, I feel as though we’ve lost an effective tool in our digital analytics toolbox. I’ll use a car analogy to help explain how I feel. Back when I was a poor MBA student in business school, I purchased a used 1994 Honda Accord. It ended up being a great car and lasted many years beyond my MBA graduation. However, one year it started having more and more serious issues. I made the difficult decision to let it go. To this day, I still have fond memories of that trusty sedan and the mileage it gave my family.

For the past several years as digital analysts, we have been driving an equally reliable vehicle that has given us lots of analytics mileage. Indeed, search keyword data has been a mainstay of web analytics reporting and analysis since our industry’s inception. It gave us deeper insights into the intent of our visitors so we could figure out what they were looking for when they arrived on our websites. In most cases the keyword data was highly actionable as we could take quick action by adjusting our bidding strategy, ad content, landing pages, and navigation options.

In 2011, we experienced our first hiccup when Google announced it would hide search queries from individuals who were signed into Google. Even though Google assured us the impact would be limited, many of us started to hear a subtle but irregular ticking sound in the search keyword engine. It also gave us a scare as we suddenly realized how dependent we were on search engines for their search query data. When Google stripped the organic search terms last year, it represented the first major malfunction in our search keyword reports as the dreaded Keywords Unavailable became more pervasive.

Although Google’s most recent move doesn’t necessarily mean the end of analyzing search keyword data, it does mark the end of an era. Whereas previously you might have relied on keyword data for different segments or analyses on a regular basis, it will now be relegated to occasional usage—more based out of hope than actual utility (especially when more search engines simply replicate what Google is doing). With a growing abundance of digital data to analyze, it’s a shame to see something useful like keyword data removed from our analyst toolbox.

No eulogy is complete without a humorous story or memory. When I was an MBA intern at Lands’ End in 2003, I remember encountering some interesting paid search phrases as I was analyzing AdWords campaigns for its B2B clothing division. For instance, “school uniforms” had a decent conversion rate but in contrast “sexy school uniform” didn’t convert very well. I found it fascinating how just one additional word could completely change the intent or purpose of the visitor.

Over the years, I would estimate that keyword-related analyses (like my simple example) across our Adobe Analytics customer base have led to millions of dollars in cost savings and incremental revenue. On behalf of all digital analysts and data-driven marketers, I’d like to salute search keywords for your many years of service. Although you’ll be sorely missed, I’m confident we in the digital marketing community will find creative and innovative ways to overcome your loss. In loving memory, Brent Dykes.

Feel free to share your favorite keyword memories in the comments section below.