04th December 2017

The debate about how URLs should be structured is something that has existed for a number of years. Over this time the questions have still remained the same:

  • Are targeted keywords to be included in the URL?
  • What is the best way to place keywords?
  • Where do subfolders fit in?

As there are so many rules of thought on this subject, we thought it would be a good idea to put the different theories to the test.

Leapfrog Blog22

The Theory

We decided to test the keywords ‘Levi Jeans’ and focused this around men’s clothing only – this became the subfolder. We wanted to find out firstly, what impact did the structure of the URL have, and secondly, could a URL become over optimised?

We used the same site but structured the website addresses differently. These were:

www.exampleclothing.com.au/mens/jeans/levi/

and

www.exampleclothing.com.au/mens/jeans/levi-jeans/

We tested which Google preferred, subdirectories (keyword1/keyword2/) or hyphen-connected keywords, effectively a keyword phrase (/keyword1-keyword2).

Our Test

In order for our test to be accurate as possible, we needed to choose a few “test” keyword phrases that were not displaying in Google. This ensured that no other factors were affecting our result. These keywords were referred to as “keyword1” and “keyword2”.

Next, we needed to set up four tests sites for each keyword phrase to test all possible variables of the phrase:

.com.au/keyword1/keyword2/
.com.au/keyword1-keyword2/
.com.au/keyword2/keyword1/
.com.au/keyword2-keyword1/

Once again, in order to maintain consistency, all the sites needed to contain the original “lorem ipsum” content to ensure that no other information affected the search.

We then Fetched the sites as Google so as to get them indexed and allowed Google time to understand the sites. This gave ample time for Google’s bots and spiders to do their work, once again to ensure ‘fair’ results. We then monitored the keyword’s ranking in Google for each particular site.

Results

Our results found that keyword phrases that are hyphenated rather than divide by subfolders were preferred by Google. Perhaps less surprisingly, Google ranks the first keyword of the keyword phrase higher when placed earlier in the URL.

During the initial testing, we found that perhaps some of our experiments were flawed as the middle sub-directory didn’t have any content and was only there for the purpose of the experiment. We decided that this may have some influence on our results. We decide to keep the number of subfolders constant for a second test but only used two keyword sets. The URL structures were as follows:

.com.au/random/keyword1-keyword2/
.com.au/keyword1/keyword2/
.com.au/random/keyword2-keyword1/
.com.au/keyword2/keyword1/

For the benefit of the test, the /random/ subfolder was again named after a keyword that produced no results in Google. The keyword also had no connection to either keyword1 or keyword2.

leapfrog blog image73

Results of Test 2

The first thing that we noticed was that Google had problems indexing our sites. In order to conduct the test properly, we decided that we needed to allow the test sites to remain live for several months. The results fluctuated significantly during this period although more consistent results appeared towards the end of the test and we were able to compare these to the initial results.

Conclusion

Regardless of subfolders and their structure, the best URL is structure is /keyword1-keyword2/. Therefore, to increase your likelihood of getting a higher ranking on Google you should aim to adopt this form.

More Information

At Leapfrog Media, we can help you to ensure that your website reaches its maximum potential. We offer a range of packages to help you achieve this. If you would like more information about our SEO services, please contact us for more details.