Serving Content Based On IP Address & Cookies

Continuing the interview between Eric Enge and John Mueller, Eric addresses the issue of serving content based on IP address to differentiate between users speaking different languages and content dealing with country specific issues.

A good example of the latter would be serving a listing of products for users in Auckland and a different set of products for users in Christchurch on the basis that consumer preferences are quite different.

John has come across a similar problem in Switzerland where there are 4 official languages. He says the probability of a user getting served the wrong language is quite high. John also says that legal rules add to the complications. For example, casino sites cannot be shown to American users whereas it is okay to do so to users in France and Germany.

In this extreme situation, even if  Googlebot tries to index a casino site, it should be treated the same way as a  user from US by showing the bot the same content that a user from US would see. It is not wise to block the site per se to all users from the US. Instead a general page can be shown stating legal restrictions to users and the bots coming from an American location.

In a situation where a site is in different languages, rather than trust geotargeting by IP address, a general page can be shown where both users and bots see the same content. From there, links to different language sites can help users select the language of their choice and also help the bots spider and index the various language sites. This is a safe yet elegant solution though it creates an extra click for both parties to reach the language specific content. It is wise to have different URLs for the different language sites.

If a webmaster is hung up on using geotargeting by IP address and serving appropriate content, John gives the example of amazon.com site. If a user from Germany lands on the amazon.com site, a suitable banner could be displayed saying something like - Want to visit Amazon Germany? It is closer to you and offers free shipping.

Eric queries if the above mechanism is a variation of having 3 types of a site with different domains as in site.com, site.co.nz and site.com.au to be indexed in the local versions of the search engines. John affirms it.

John adds that there are a couple of downsides to serving region specific language content based on IP address. Firstly, the IP location and language detection is often incorrect. John says that even at Google, things go awry when when they feel a user is from Germany and serve German content when in reality she is from France. So, it can go wrong at some point in the whole process.

A second version of the above problem arises when users in the wrong location can still reach your website.  Google can crawl and index the site without any problems though John advises to have elements on the page that has links to various language sites so that the user can get to the language of her choice even if the IP targeting fails. Also, barring any legal issues, the Googlebot can index the French and German language pages and display them in the local search results in Google France and Google Germany respectively.

Eric throws up a question on serving varying content based on cookies according to user preferences.  John yet again emphasises the fact that this technique should not be used deceptively. The same content should be served both to an unregistered user and the Googlebot visiting the site.

For existing users who have set preferences, it is fine to show them extra content on logging into the site along with a user enhanced experience based on their preferences. If the logged in users were to see different content or very little content, that would be a huge problem.

It has been a good discussion of various techniques that can lead to cloaking of some kind or the other either wittingly or unwittingly. It is a good idea to know these issues so that webmasters can avoid techniques that is frowned upon by the search engines.

Here at Netconcepts, we write squeaky clean code and run it through W3C Validator to ensure that the search engine bots find it easy to index the pages. Our very skilled and experienced Auckland web development and Auckland web design team ensure that the site is user friendly and easy to index for the search engine bots.

This entry was posted on Friday, May 29th, 2009 at 2:17 pm and is filed under Blogging, Natural Search, Ravi Venkatesan. You can follow any responses to this entry through the RSS 2.0 feed. You can leave a response, or trackback from your own site.