How to Find (Not Provided) Keywords

Any webmaster that monitors Google Analytics, especially the data hawks, to develop online marketing plans knows the (not provided) keyword segment of Google can be a source of major frustration. In November 2011, Google began protecting the search data of users logged into a Google account (i.e. Gmail, Analytics, AdWords, Google+, etc.). With a growing number of Google products and users, more keyword data than ever is falling into the (not provided) segment—leaving webmasters and business owners with less and less data to improve site performance. In fact, a new study by Optify has found that 39 percent of search data from Google falls within the (not provided) segment, up from roughly 15 percent a year ago. For some websites, it can even be considerably higher!

Take a look at this example:

At Blue Corona, data is at the core everything we do, so naturally Google’s policy change presented us with a challenge. How could we adjust our methods to account for Google’s change to incrypted keyword data for logged in users?

Initially, with a relatively small portion of the organic search data being hidden (not provided), the data we captured was generally sufficient, but now with roughly 1 out of every 2.5 keyword searches hidden that game has changed. We understand how frustrating this can be for business owners and search marketers alike, so we would like to help you mine as much data from your website visitors as possible despite this roadblock.

There are several ways to gather and interpret the data in a way that can still be beneficial and measurable.

The two methods we prefer are:

Visits Segmented by Landing Page

Segmenting your (not provided) traffic by landing page is a quick way to give you a ball park figure for how many of your (not provided) visitors are branded versus non-branded organic traffic. Determining branded versus non-branded organic traffic is vital to any successful SEO campaign as it is a measure of how many people are finding your website that are not already familiar with your brand!

To do this, filter your traffic for (not provided) by applying a filter:

Or by simply clicking on the (not provided) keyword. The result will look like this:

From there, click on Secondary dimension>Traffic Sources>Landing Page. This will provide you with a list of pages that drove traffic to your website, sorted by volume of traffic.

In Google Analytics “/” is your homepage (depending on how your web guys have set things up, your homepage might also show as /index or /index.html). Virtually all branded traffic will land on your website’s homepage. In the example above, there are 2,352 visits to the homepage, or 24.5 percent of all (not provided) traffic.

Although virtually all branded traffic will land on your site’s homepage, not all of your homepage traffic is from branded terms.

If your website has been optimized properly, your homepage will likely rank for several high-level keywords related to your industry. To get a better idea of just how many of these visits are branded versus non-branded, you must filter out the (not provided) data.

Using the data available at this point, filter out the non-branded keywords so that only branded keywords remain. For example, I would want to filter out any search phrases for “Blue Corona,” employee names, or similar if I were perform this on Blue Corona’s analytics. Once you have done this, divide this number by total visits (excluding the not provided segment) to determine the percentage of branded visits.

In the example above, roughly 22.5 percent of the total site traffic was branded traffic, allowing us to determine with relative accuracy that 9,064 of the 9,594 visits are likely non-branded keywords. Simple enough, right?

In addition to helping you determine how much of your organic traffic comes from branded versus non-branded search terms, sorting by landing page can also help give you a rough idea of what broad keyword phrases are driving the most traffic to your site. For example, if you are an internet marketing company like we are, and you see that your blog post, “What is a Good Website Conversion Rate” has driven 100 visits to your website in the last week, you can make the assumption that a broad-match of the keyword phrase “what is a good website conversion rate” (or similar) is driving traffic to your site.

Overcoming (Not Provided) With Webmasters Tools

Although the process above is a relatively accurate—and quick—way to get a general idea of the type of traffic coming to your site and the keywords that are driving traffic to your site, you can obtain much more reliable data if you have a Webmasters Tools account set up for your site. If you have a Webmasters Tools account in place, simply navigate to the Traffic tab and click “Search Queries.” You will be provided with the following information (and more):

This chart will help you determine the keywords for which you are appearing the most and give you a general idea of how many visits each keyword is sending to your site. The data is not exact, but from our calculations is generally a better gauge than the Landing Page method described above.

Something like 80 percent of all businesses use Google Analytics to track their website activity.  Based on our experience, fewer than 5 percent of people using Google Analytics a) have it set up correctly and b) translate the data into improved business performance. If you’re a business owner and you’d like someone to help you turn analytics into action, contact Blue Corona today.

Let us handle your website analytics & tracking so you can focus on doing what you do best – running your business!

Want to learn more?  Let’s talk! 800.696.4690