• http://www.pagerank-seo.com Robert Visser

    Thx for including the statement on how Bing values underscores vs. hyphens in URLs. It balances Matt Cutts’ statement in the Google Webmasters Central video.

    Part of this is to understand how the Google algorithm interprets punctuation, in this instance underscores and hyphens. From what Cutts disclosed in the video, Google sees a hyphen as a space whereas an underscore is discarded to create a compound noun.

    Not covered and beyond the immediate scope of the video is how the algorithm interprets punctuation in dynamic URLs including the ‘+’ and ‘%20’ characters.

    What’s not mentioned is, of course, how people conduct queries. To use the examples in the video it’s highly unlikely anyone would actually query a long tail keyword phrase, ‘war of 1812’, as a compound noun, i.e. warof1812. So the value the Google algorithm attributes to having keywords in the domain would be diminished if the domain was http://www.war_of_1812.com rather than http://www.war-of-1812.com.

    To be fair the ‘war of 1812’ example is itself over simplified; ‘of’ is a stop word would not be valued unless the query was conducted with the long tail keyword phrase inside quotation marks, i.e. “war of 1812”. Historically, quotes signify the query is to return the exact long tail keyword phrase in the order it was entered. While I believe this method does retain some value, recent changes have influenced how the Google algorithm sees search terms. For instance, the algorithm can now recognize some common occurrences of how a search term is queried. If one queries ‘realtime’ the results appear to be unchanged from a query for ‘real time’, while a query for “real time” (with quotation marks) will render different SERPs.

    Another variable — assuming one is logged-in — is how the algorithm interprets user intent & attributes value based on personalized search.

    When choosing a URL, naming folders, or naming files, the greatest value will be realized by implementing a naming convention with long tail keyword phrases which reiterates the manner in which those long tail phrases are queried. The best reference, of course, is the keyword logs in one’s web analytics app. To use an ecommerce example, assuming one has created advanced segments, the long tail keyword phrases with the greatest values are those queries from which a gaol was realized. If a [completed] purchase originate from a query, it’s the keywords which brought the buyer to the funnel.

  • http://www.seo-suedwest.de/english-blog.html Christian

    Dashes can take quite different effects in Google: On the one hand side they are a useful way to seperate multiple terms in order to get them indexed properly. On the other hand too many dashes – especially used in URLs – seem to be suspicious to Google and can result in a worse ranking. So one has to be careful and think about it before using.

  • http://www.seocatalysts.com/ I.M.

    Dashes can useful to describe about the page in short..and that is a really effective way to promote your pages with proper naming convention

  • MarkWarner

    I am pushing the following comment: (but not spamming it)

    Could I suggest that you add an option in webmaster tools \split on underscore\. By default this is off. So if you are one of the idiot \underscore\ webmasters like me you can opt in.

    This should then mean that the project to split is really a lot simpler to implement in respect of checking for webwide impact etc.

    You’ll be producing better quality results!

    Smart idea eh?



  • Razibul Hassan

    For an existing website with a good ranking the fact can be ignored for now. But if you are going to create a new website then you should definitely put the matter in consideration. Simply place underscore for closely connected search term phrases that you can treat as a single phrase or key word . But for the a long URL with different keywords placed together the dash(-) is the appropriate one.

    Thanks for the great post.