Not Provided Comes To Paid Search: What Will The Impact Be?
If you’re reading this, you probably already know that Google is making changes that will bring the much-maligned “(not provided)” to paid search. The AdWords world has been abuzz about the change for the past few days.
The gist of the development is that when people using secure search click on AdWords ads, the user’s search query in the referrer string — the actual words that people entered into the search box — won’t be passed to analytics packages and third-party software.
This is an annoyance, but in the great scheme of things, it’s not a major problem. I’ll explain more, but first a little background.
At SMX West in March, Google search chief Amit Singhal directly acknowledged the company recognized the contradiction in providing advertisers with data that non-paying SEOs weren’t receiving. He said the lack of parity was a concern that Google understood and would address. That got some marketers and publishers excited that user query data would soon come back to web analytics packages.
Not so. Instead, we now know that Google’s plans for achieving parity involve blinding paid search folks to query data.
Google announced today that, going forward, an increasing amount of secure search would pass referrer strings to advertisers that omit the user’s search query, so this will appear as “(not provided)” in analytics reports. Over the next month, the percentage of search queries omitted for paid search traffic will reach organic levels, representing nearly all user queries.
Google gave us a courtesy heads up that this announcement was coming, which we appreciate. Having had a chance to consider the situation, my thinking now is pretty much the same as it was back when Singal spoke and the speculation began.
An Annoyance But Not Major
From our perspective, the largest headache will be a change to how we currently analyze search query performance. By pulling the user query from the referrer string, we’ve been able to determine which user queries led to conversions and which didn’t. As a result, we could isolate strong- and poor-performing tokens and add them as additional keywords or negatives, respectively.
If the account has conversion tracking in AdWords, we can still view this type of data through a search terms report in the AdWords interface. We prefer to use referrer data, since Google’s conversion data does not always align well with the advertiser’s own view of conversions, but the capability is not completely lost.
And if an account doesn’t have AdWords conversion tracking, we can still view these reports in aggregate, mining for keywords to add, or identifying tokens that clearly indicate that the user is looking for something other than what our client offers.
Those who use search query data for other purpose may need to change their approach. Advertisers using dynamic search string insertion on their website or mobile landing pages may need to begin using keyword insertion instead. And this change may pose a larger problem for companies that gather search query information and use it to retarget search users on other sites.
Most in paid search, though, will still be able to use the referrer string to identify the keyword served, and certain information about the user who clicked the ad. And certain reports, like Google’s click performance report (available via the API), can show useful information for each click as well.
What Does This Mean For Personalized Search?
As with organic, the argument is that this is about user security. Google says people will type names and addresses into a search bar, or other bits of information that may make them personally identifiable. I’d argue that those people’s names and addresses are also probably in the phone book, or publicly available anyway.
Indeed, there is an interesting tension developing here: we’ve evolved over the last few years from keyword targeting to a much more personalized approach to targeting that understands device, location, proximity, demographics, context, and past user behavior. Yet, Google is trying to keep this particular piece of information (what the user just typed or spoke) private. Seems like an odd place to draw that line, and it’s still not clear to me what privacy boogeyman has them concerned.
Ultimately, we only want user information so we can serve the right ad to the right person in the right context at the right time. Understood that there are bad actors out there with bad intent, but…
As we read it, this is a drag but not a major problem. How do others see it?
Opinions expressed in this article are those of the guest author and not necessarily Search Engine Land. Staff authors are listed here.