Working with Internal Site Search Data in Google Analytics

Updated on August 13, 2019, to reflect more current information.

As marketers, we can put so much effort into getting our sites into Google’s organic search results that we often neglect the search data from people who have already reached our sites. Internal site search data can show you what your target audience wants to read, what they have trouble finding, and, most importantly, which search phrases have the most conversion intent.

It requires a brief bit of set-up and some understanding of what to look for in the data, but it can be massively valuable in improving your site for existing users and helping you reach new users.

Getting Site Search Data Flowing

To set up this data on your site, go to your internal search function…

…and search for literally anything:

Screenshot of Portent blog page with search bar at top

Once you get to the search results page, look for what you searched in the URL parameters. In our case, it shows up in the addsearch parameter. Copy this parameter for later.

Screenshot of Portent blog highlighting search URL parameters.

Open up your Google Analytics Admin (1) and go to View Settings (2) to access the site search settings.

Screenshot of where to find View Settings in Google Analytics Admin.

Flip the Site Search Tracking toggle to “ON” and plug in the query parameter you copied under it.

Screenshot showing where to enter the query parameter you copied in Google Analytics.

Unfortunately, this setting isn’t retroactive, but it will start collecting site search data in the Behavior > Site Search section of Google Analytics from here on out.

Screenshot of Portent blog page highlighting the search icon.

What the Site Search Reports Look Like

Once the data starts flowing, there are several ways Google’s canned site search reports allow you to slice the data for better insights.

Behavior > Site Search > Usage

The first is simply segmenting those who perform a search vs. those who don’t and allowing you to compare a range of engagement and conversion metrics under each of those lenses.

This can give you a general feel for how effective your internal search mechanism are at helping people find what they need and complete objectives on the site.

In our case, site searchers convert at over 6x higher than those who don’t search!

Screenshot of Google Analytics site search usage report results for Portent blog.

Behavior > Site Search > Search Terms

The next piece is the nitty-gritty: what are people actually searching for? Although the Top 10 may indicate some broader trends, I recommend getting all search terms out into Excel and grouping them into categories of similar searches. For us, SEO-related topics (“seo” and “serp”) and searches around our tools (“title generator” and “content idea generator”) would be worthwhile groups.

One thing to keep in mind: you’ll get a much larger range of long-tail searches with only one unique search than you will short-tail searchers with lots of unique searches. So it’s important to bundle performance across a range of similar searches in order to make the data more meaningful to act on.

Screenshot of Google Analytics site search search terms report results for Portent blog.

Behavior > Site Search > Search Pages

The last report in the section gives you an idea of where people are on their site journey when they’re getting lost. A high percentage of Search Refinements and low Time after Search can be indicators that the page they started searching on isn’t effectively showing folks where to go next or answering their questions in terms of navigation.

Screenshot of Google Analytics site search search pages report results for Portent blog.

What to Do with the Data

This data, like any marketing data, can be massively overwhelming when you start collecting it over long stretches of time. But focusing on a few important things can help you make the most of it.

Content Ideas

Pretty obvious, right? Visitors searching for content that doesn’t exist on your site is a one-stop-shop for ideas your content team can work on creating. We had a lot of people searching for “shoes” in our data, probably because of this old post we did on Shopify SEO that included a ton of examples around shoes.

Maybe they want to come back to that post specifically? Or maybe we just have a lot of apparel brand marketers looking for all-around marketing advice now? In any case, it will help us create better assumptions about our target audience and what kinds of posts get them reading.

Reorganize Site Navigation

Another avenue for exploration is re-working your nav based on what gets searched for most often on the homepage.

Our site search data showed a high concentration of folks looking for specific tools we built in the past. That link is nested under our “Resources” navigation menu, but perhaps we’re making people interested in our tools work too hard to find it there?

Screenshot of Portent blog page with "resources" in the top navigation highlighted.

Retargeting Audiences

But the most powerful way you can use this data is to generate audiences based on specific site search phrases who haven’t converted yet and reach back out to them with very targeted offers in Google Ads.

Screenshot showing how to segment users in Google Analytics.

With this segmentation and audience-building method, you don’t have to hope somebody will come back to the site after you’ve optimized for site search experiences. You can actively pursue them and bring them back to specific pages that they might not have seen on their last visit.

Screenshot showing how to build an audience to re-engage those users in Google Analytics.

While focusing on organic search results is important to help grow the number of new visitors to your site, don’t forget about those individuals who have already discovered your product or brand. Hop into site search data today and see what you can learn about the visitors you have and learn what they might tell you about the visitors you want.

The post Working with Internal Site Search Data in Google Analytics appeared first on Portent.

This is a companion discussion topic for the original entry at