When monitoring performance and identifying improvement opportunities in Google organic search traffic, it’s critical to understand the queries that are bringing search traffic to the site. Queries fall into three broad categories:
- Branded queries – visitors are searching for the brand directly. These searches can be further categorized: brand name or domain name searches, searches for product names, local branded searches (if your company includes physical locations or service areas) and so on.
- Non-branded queries – visitors aren’t searching for your brand specifically, but are looking for what your site offers. These searches can be further categorized into topics and intents. For example, if you sell products, you might want to organize these queries into not only topics based on product category, but also purchase vs. research intent.
- Irrelevant queries – visitors are searching for unrelated topics. This traffic is often not valuable for the site (as it doesn’t lead to conversions or additional page views). Increases in this type of traffic can misleadingly look like SEO success and content improvements to the site can sometimes lead to a decline in this type of traffic (which can misleadingly look like an SEO problem). So it can be important to understand how much Google organic search traffic falls into this category.
The Value of Tracking Branded Queries
Future posts will dive into non-branded and irrelevant queries, but first, let’s look at branded queries. Branded queries are useful to track for a variety of reasons. For instance:
- Branded Queries and Direct Traffic – Searches for the brand name or domain name are similar to direct traffic. Many visitors use Google search rather than type a URL in a browser address bar. It’s useful to understand what percentage of your site’s Google organic search traffic is from these types of searches.
If most of the organic search traffic is from searches for your brand name or domain name, your site may be missing substantial SEO opportunities. Increases in organic search traffic may be due to non-SEO factors, like increases in offline advertising. Tracking branded and non-branded organic search traffic separately enables you to better measure the impact of your SEO efforts.
- Local Branded Queries – if your brand includes physical locations or service areas, it’s important to track the site’s SEO performance for local branded queries. These include explicit local searches like [wild birds unlimited seattle] and implicit local searches (for which Google’s algorithms infer local intent even though the query doesn’t include it). Often, implicit local searches have overlap with direct branded queries, as is the case with the query [wild birds unlimited]. Google may show location-specific results for these queries based on the searcher’s location.
- Product-specific Queries – It can be useful to know if customers are searching for specific products and which products they’re searching for.
- Support and Contact Queries – It’s just as important to keep existing customers happy as to attract new customers. Tracking support queries provides insight on what issues your customers are having. Also, you want to be sure that your site is what ranks for these queries so customers get the best support possible.
Tracking Traffic From Branded Queries
Segmenting Google organic traffic into branded and non-branded categories can help you better understand a site’s SEO performance.
- If most organic search traffic is from branded queries – the site likely isn’t fully optimized for organic search visibility. For many industries and verticals, many audiences are not seeking out a particular brand or site directly when searching.
- If most organic search traffic is from non-branded queries – the organization may not be building up brand recognition or engagement. Ideally, some audiences will seek out your brand directly and even more ideally, some audiences will seek out your brand directly in conjunction with more general queries.
You can see what brands are commonly searched for in conjunction with some general queries from Google’s suggested searches. For example, for the query [hotel review], Google suggests [hotel reviews yelp].
For the query [portland hotel review], Google suggests [portland hotel reviews tripadvisor]. The more searchers add your brand to a general query, the more likely Google will suggest a search with your brand (leading to additional visibility and traffic).
In addition, rankings and click through rates generally vary greatly between branded and non-branded queries. For example, click through rates are generally much higher for branded queries since the searcher is specifically seeking out your site. When monitoring click through rates or using them for calculating market opportunity, separating average click through rates for branded and non-branded queries provides much more accurate data.
Tracking Performance of Branded Queries
The first step in tracking the performance of branded queries is determining the traffic and percentage of traffic from these queries. You can then:
- assess whether the percentage seems right or is too high or low – the “right” percentage varies based on industry, and other factors such as whether your marketing strategy includes Super Bowl commercials and whether the brand has a strong offline presence, such as physical stores). Many online-only brands may get 3%-5% of their organic search traffic from branded queries and many “offline” brands may get closer to 60% of their organic search traffic from branded queries.
If the percentage is too high, then opportunity likely exists in increasing query visibility for non-branded queries. If the percentage is too low, then you may want to consider ways to improve brand awareness and engagement (such as social campaigns).
- track brand awareness campaigns – for instance, do offline advertising campaigns lead to additional branded search traffic? does paid search spend on branded queries increase or decrease organic search traffic for those queries?
- evaluate average rankings for branded queries – in general, your site should rank well for all branded queries. (Exceptions include cases that your brand includes multiple sites or if your brand name overlaps with unrelated intent.) If it doesn’t, you can identify the queries with poor rankings and determine what might be causing the problem. In many cases, the issue is something easily fixable, such as a vague title tag like “Home” on the home page, but could also be something more serious such as a Google penalty.
- evaluate average click through rates for branded queries – As with average rankings, average click through rates for branded queries should be fairly high except in cases that multiple branded sites rank or the query could have unrelated intent. If click through rates for high ranking branded queries are low, look for issues such as non-descriptive title tags or meta descriptions or the wrong page ranking.
Note that for this evaluation, it’s most useful to look at the click through rate for the query, not the click through rate of a particular page. Since multiple pages may rank for branded queries (both with sitelinks and with multiple listings), any given URL may have a low click through rate.
Also note that for local branded queries (particularly those for which the intent is information about a physical location), the click through rate may be low because the searcher may click on a listing in the Google Maps onebox or may not click at all (if all of the needed information displays in the brand knowledge graph or local onebox panel).
Using Google Search Console to Track Branded Queries
One way to track all of this is via Google Search Console. The pros of this approach include that it’s free any easy. The cons include that:
- if you’re using the Google Search Console user interface, you can only view and export up to 999 queries; you can only filter on keyword matches (and not more complex regular expressions), so if your brand includes multiple variations, you may need to do several filtered searches and then combine the data; and you can’t set up saved search filters. In addition, you’ll need to calculate the percentages yourself and you can view only up to 16 months of data (although 16 months is enough for most purposes).
- If you’re using the Google Search Console API, you can get additional data, but it takes some initial set up work to get the API connection in place (and if you really want a robust system, you would need to set up additional code, such as error handling); and you need a separate system of some kind to store, process, and analyze the data.
To track branded queries with Google Search Console:
- Access the Performance report for the domain.
- Choose the search type (likely “Web”) and date range. For general traffic and percentages, you can leave the device type filter set to “all”.
- Click New > Query.
- Filter by “Queries containing” and type the shortest version of your brand name as the keyword. For instance, if your brand is “Kimpton Hotels”, type “kimpton”. If your brand is “Wild Birds Unlimited”, type “Wild Birds Unlimited”, since “wild birds” is likely to include many non-branded searches.
The resulting summary data will sum clicks and impressions and average click through rate and ranking for all matching queries, but will display only up to 999 of them.
Review the list of queries and make sure that your filter didn’t inadvertently include a lot of queries that aren’t actually for your brand. For example, when I filter the data for this site using the “keylime”, I see the following list of queries:
Queries related to [keylime athletic wear] are not for this brand. But since Google Search Console filters don’t allow for optional spaces (which would enable me to include either “keylime toolbox” or “keylimetoolbox” or for exclusions (which would enable me to filter for queries that include “keylime” but exclude “athletic”), I’ll need to generate reports using multiple filters and then aggregate the results.
When looking at average ranking and average click through rate trends, it’s useful to filter by device (since these may vary between mobile and desktop). For example, below is the last 16 months of average ranking on desktop devices for queries that contain “keylime toolbox”:
Using Keylime Toolbox to Track Branded Queries
Keylime Toolbox connects to the Google Search Console API to retrieve as many queries as possible, stores that data over time, and includes configuration for a “branded” definition. This definition is a permanent configuration of the account (that you can adjust at any time) that uses regular expressions, so you can include as many variations as needed for tracking your brand.
The Light Edition ($49/month per reporting group) includes segmented metrics with the ability to export filtered queries for any date range and device to Excel.
Traffic Dashboard – Branded vs. Non-Branded Distribution
When Google Analytics is connected to your account, Keylime Toolbox can estimate the total percentage of branded vs. non-branded Google organic traffic based on the total Google organic traffic number and the percentage of query data that the Google Search Console provides.
Query Details Report – Branded Segment
SEO Metrics – Branded vs. Non-Branded (Year over Year Comparison)
The Complete Edition ($115/month per reporting group) includes trended reports (including number of queries that match the segment, traffic, impressions, average rankings for the queries in the segment, and average click through rate for the queries in the segment) and the break down of average click through rates for the queries in the segment for each ranking position.
SEO Trends – Branded vs. Non-Branded Year Trend (Traffic)
SEO Trends – Branded vs. Non-Branded Year Trend (Average Ranking, Average Click Through Rate, Impressions, and Number of Queries)
SEO Trends – Comparing SEO Metrics for Branded Queries Year over Year
In the example below, branded traffic has increased due to an increase in click through rates for those queries (average ranking appears to have declined, but the actual ranking decrease is only .2, so rankings are actually basically flat):
Click Through Rate Distribution: Branded vs. Non-Branded Queries:
Below you can see aggregate data about click through rates for each ranking position for branded vs. non-branded queries. As we found in our earlier study, average click through rates by ranking position vary widely by industry, query type, and site, so the best way to estimate what the click through rate would for your site for a new set of queries is to use data from your site’s existing SEO performance.
As noted above, this data is much more useful when segmented by branded vs. non-branded queries, since click through rates for branded queries should generally be higher.
With this Keylime Toolbox report, you can filter by device type (mobile, desktop, tablet), can see the click through rate trends for each query segment/ranking position, and can drill further into the number of impressions and clicks behind each calculation, as well as see the specific queries (and associated click through rates).
Setting Up the Branded Query Segment
In addition to the standard “branded” segment part of every account configuration, with the Complete Edition you can set up additional segments. For tracking branded queries, for instance, it may be useful to set up a segment that tracks traffic from queries for products or local service areas.
(The Complete Edition includes five configurable segments for all trended reports; The Enterprise Edition includes 20. For either edition, additional segments are available. In addition to configurable trended report segments, unlimited custom segments are available to use as filters to export query data. More on that below.)
Keylime Toolbox query segments are set up with regular expressions, which enable flexibility in what queries are included. Our user guide includes details on setting up these regular expressions, and the account configuration tool includes testing functionality. You can also always email us at firstname.lastname@example.org (whether you’re using a trial or regular subscription) with a description of the segments you’d like to set up, and we’ll help you configure them.
To configure the “branded” query segment for your Keylime Toolbox reporting group:
- Choose Settings in the upper right, then choose Reporting Groups from the left menu.
- Click Segment Patterns beside the reporting group you’d like to configure.
- Click the Edit icon to the right of the Branded segment.
- Use the Pattern field to include regular expressions that match branded queries.
- Use the Exclusion field to exclude regular expressions.
- Click Test to see what queries that bring traffic to the site are included, excluded, and not a match for the definition you’ve set up.
- Adjust if necessary.
- Clicks Save. Keylime Toolbox re-processes all historical data overnight.
Below are examples of regular expression patterns that might be useful to configuring your site’s branded query segment.
If your brand name includes multiple words or could be searched with singular or plural, and the brand name doesn’t have an alternate meaning, you might be able to include just a single word or word stem for your branded configuration. For instance, a definition of “pepsi” would include the query [pepsi] but would also include queries like [pepsi flavors] and [pepsi.com]
Word Stems with OR
If your brand could be referred to in multiple ways, you can include any variations with or statement using the pipe character (|). For example, Coca-Cola could set up a definition like this:
That configuration means to include queries that match “coke” or that match “coca”.
Word Stem + Exclusion
If your brand name could be used for unrelated searches, you can exclude them with the Exclusion field. For instance, with the “keylime” example above that includes searches for “keylime athletics”, the configuration could be:
If your brand name includes multiple words, it may be useful to include a regular expression for an optional space between the them. This method is useful if a single word of the phrase would include non-branded queries. For example, Wild Birds Unlimited couldn’t set up a branded segment that simply included “wild or “bird”. But the branded segment should include any query variation for “Wild Birds Unlimited” and “wildbirdsunlimited.com”. To configure that, use a space followed by a question mark (?). The question mark means that the character just before it (in this case a space) is optional, so the regular expression will filter queries that have a space and those that don’t.
wild ?birds ?unlimited
For this brand, searchers might include singular or plural, so the branded configuration should match both. To do that, simply add a question mark (?) after the “s”, like this:
wild ?birds? ?unlimited
That will match any phrase variation with or without spaces and with or without the s at the end of “bird”. Queries this pattern would match include:
wild birds unlimited
wild bird unlimited
Using Multiple Regular Expressions Together
Your regular expression can be as complicated as you need it to be. For instance, for the Wild Birds Unlimited example, the domain name being used is wbu.com (wildbirdsunlimited.com redirects there). That domain should be included in the configuration. The complete configuration would include both the brand phrase with optional spaces and optional plural as well as an OR operator, like this:
wild ?birds? ?unlimited|wbu
This configuration can be useful for misspellings, for instance, even if your brand is only referred to one way.
Lots of other configurations are available. Below is a quick reference of common regular expressions:
|Regex||What It Does||Example||Matches|
|string of characters||Matches any queries that contain those characters||cake||cake
|\b||Sets a boundary either at the beginning or end of a word (or both)||cake\b|
|cake but not cakes
pie recipes but not cute puppies
|?||Matches whether the character that precedes the ? exists or not.||crab ?cake|
cake ?-? ?balls
|crabcake and crab cake
cakeballs, cake balls, and cake-balls
|||Matches on any one character in brackets||theat[re]|
|theater and theatre
any one digit 0-9
any one letter a-z
|.||Matches any character. Generally only used in conjunction with exclusion patterns (match everything except xyz)||.||everything|
|\.||Matches a period||keylimetoolbox\.com||keylimetoolbox.com|
|^||Matches only queries that start with the following characters||^cake||cake recipes but not pancakes|
|$||Matches only queries that end with the preceding characters||recipe$|
|recipe and cake recipe (but not recipes)
recipe and recipes (but not recipe list)
Reviewing and Finalizing the Branded Query Segment
Once you’ve set up the filter, click the Test button to see what types of queries are included and excluded. Adjust as needed. Then click Save. Keylime Toolbox will reprocess all historical data overnight and will include a label of “we’re updating your segments” until that’s complete.
Exporting Query Data Using a Regular Expression Filter
The segment configuration for trended reports described above is ideal for tracking SEO performance for branded queries over time. But you also might want to parse the data in different ways for ad hoc analysis. For example, you might want a report that details queries for a particular product.
To do that, click the Tools menu in the left navigation, then Filter and Download Queries. You can set up a custom filter by clicking the New Custom Filter button. Custom filters work the same way as trended report query segments (you can use regular expressions, include and exclude, and test them). You can save these custom filters for later use as well.
Once you set up a custom filter, just choose the date range and device filters, then click Filter and Download Queries. The system will generate an Excel report and will email you once it’s ready. These reports generally take only a few minutes to generate but may take a bit longer for large amounts of data or long date ranges.
Tracking SEO Performance of Non-Branded Queries
As you expand content areas on a site or optimize existing pages, it’s useful to monitor increased traffic from non-branded queries. This is particularly the case since traffic changes from branded queries are often influenced by non-SEO activities (such as offline advertising), which can change overall branded search volume.
When a visitor uses Google instead of the browser address bar to type in a site’s domain name, the resulting traffic isn’t truly a result of SEO activities. But when a visitor comes to the site after searching for [apple pie recipe] or [spring dress], the resulting traffic likely is.
Often, it’s useful to further segment non-branded traffic into topic areas, but that’s detail for a later post.
Setting Up a Trial of Keylime Toolbox
If you aren’t already using Keylime Toolbox, you can set up a free trial. Our set up wizard walks you through adding Keylime Toolbox as a user to your property (or properties) in Google Search Console and Google Analytics (if applicable). The system imports and processes the query data and everything is available generally within 48 hours.