Every day, billions of searches (also called search queries) take place online. These searches often include a variety of words and phrases made up of keywords that play a role in determining the search results provided by search engines. In late 2015, Verisign sought to understand whether or not having a keyword-rich domain name that shows up in search results provides any benefits over domain names appearing in search results that do not include keywords.
To help answer this question, Verisign researchers analyzed millions of search queries licensed from the cross-platform measurement company comScore and discovered that:
Internet search users are almost twice as likely to click on a domain name where the second level includes at least one of the keywords in their search query, compared to a domain name that does not include any of the keywords in their search query.
For example, if a user searched for: “How do I create a personal brand online?” he or she would be more inclined to click on a domain name such as www.PersonalSpaceOnline.com than a domain name that does not contain any of the keywords included in the search.
This finding illustrates the importance of thinking like your target audience and asking: What might someone type into a browser or search bar to find my product or service? Incorporating those same keywords into your domain names may make the difference in ensuring people find and click through to your website.
To help determine if there is a benefit to registering domain names that include keywords over those that do not, Verisign studied 11 million search query results from approximately 1.55 million search query sessions in the month of September 2015 from comScore’s U.S. consumer panel.  These search query results included the following fields:
- search_id: Unique ID for each search event
- machine_id: Unique ID for each machine
- search_link: URL link shown on the Search Engine Results Page (SERP)
- rank: Rank of search link on the SERP
- search_phrase: Search phrase entered by user
- click_flag: Flag for whether link was clicked
The data analysis was conducted by Verisign data analysts. The analysts focused on the second-level domain names (SLDs) (the characters and keywords that come before the dot in a domain name) returned for all of the search queries conducted by comScore’s U.S. consumer panel in September 2015. The analysts only focused on SLDs because the number of search query results returned for top-level domains (TLDs) that could be considered keywords was not significantly measurable.
Of the top 50 TLDs that appeared in the search query results, 38 of them appeared in only 1 percent or less of the search query sessions. The 12 TLDs that appeared most frequently were:
- .com (98 percent)
- .org (47 percent)
- .net (18 percent)
- .edu (9 percent)
- .uk (7 percent)
- .gov (7 percent)
- .us (3 percent)
- .info (2 percent)
- .au (2 percent)
- .ca (2 percent)
- .tv (2 percent)
- .me (2 percent)
Additional filtering was conducted by the Verisign data analysts to remove data that was incomplete, unusable or would negatively impact the accuracy of the analysis. The filtering process included data that:
- Included personally identifiable information (PII): According to comScore, to protect the privacy of its panel, comScore anonymizes potentially sensitive information in search data (including in the search link and search phrases) by – among other things – replacing keywords and certain consecutive numbers that may represent personally identifiable information (e.g., proper names, telephone numbers, mailing address, etc.) with generic text. Because this anonymization precludes accuracy in studying the relationship between the search phrase and search link, this anonymized data was filtered out during Verisign analysis.
- Did not appear in the top 20 search results: In this study, 99.9 percent of the click-through data occurred in the top 20 search results. Therefore, Verisign’s analysis was limited to results with a rank of 20 or less.
- Appeared to be a navigational query: Many of the search queries appeared to be intended to navigate directly to a specific, known domain name, and did not appear to be the result of general searches. In other words, a specific domain name was searched for instead of the user entering the domain name in the browser directly. Examples of these include searches for highly popular internet brands and social media sites. Because these search queries naturally include the domain name in some form, they were filtered out to prevent the risk of skewing the results.
- Included search queries with no clicks: In other words, a panel participant searched for a topic, but did not click on any of the results returned. Since the behavior being observed was propensity to click on a search result, this data was removed.
Once the data filtering was complete, analysts identified the overlap between the search terms and the domain names appearing in the top 20 search results by:
- Separating the search terms with spaces and identifying the unique set of keywords in each search. For example, “HOTELS ON CATALINA ISLAND” would have the terms “HOTELS,” “ON,” “CATALINA” and “ISLAND.”
- Counting how many of the search term keywords were included in the resulting domain names (no keywords = “zero match”).
- Counting the number of search results that were clicked on (click_flag=1) for the search results with “zero match” and dividing it by the total number of search results with “zero match” to compute the click-through rate for these search results. Analysts repeated the same process for “one match” search results and “two or more match” search results to derive click-through rate for each.
The domain names with keywords that overlapped with the search terms had a click-through rate almost two times higher than domain names without overlap. In other words, Internet search users are almost twice as likely to click on a domain name that includes at least one of the keywords in their search query, compared to a domain name that does not include any of the keywords in their search query.
The value of keyword-rich domain names has been debated when it comes to topics like desirability, memorability and search engine ranking. Based on the results of Verisign’s analysis of comScore data, it appears that keyword-rich domain names can play an integral role when it comes to user click-through rate on search engine results. Given that “traffic to a website” is frequently cited as a top priority for website owners, they should seriously consider registering keyword-rich domain names that align with their website or business focus to take advantage of a search user’s higher propensity to click on domain names that include one or more of their search keywords.
While there are many variables that go into search rankings, like content quality, cross-linking, advertising budgets, faster website speeds, etc., having a portfolio of descriptive, keyword-rich domain names may make the difference in being found online. Verisign’s analysis of the comScore data illustrates that registering keyword-rich domain names may be a smart strategy, giving businesses a leg up when it comes to getting prospective customers to click on their links.
Search today to see available keyword-rich domain names. You can use tools like DomainScope for domain name ideas, to research the registration history of a domain name and compare available DNS traffic information for multiple domain names.
- According to comScore, their measurement of the U.S. online universe is derived in part from an opt-in consumer panel that passively and anonymously measures the internet behavior of its approximately one million participants. comScore uses a broad array of panel recruitment techniques to attract a diverse group of participants who agree to download a proprietary software meter to measure a variety of online behaviors, including pages they visit, videos they view, searches they conduct, purchases they make and ads to which they are exposed. According to comScore, anonymized information is statistically weighted and projected by comScore according to a number of variables to represent the behavior of the total U.S. internet population.