While you can localize your website in a number of different ways, each method significantly changes how your site is crawled by popular search engines like Google and Bing. With all of the different methods out there, how do you choose one that’s convenient, but SEO-friendly? Explore the pros and cons to three common website localization methods below.

Traditional approach

The most standard approach to website localization is serving the completely localized webpage from your servers. This requires that you maintain a one-to-one list for each term in your site/application and then look up each corresponding translation prior to serving the page to users.

Pros:

• Most frameworks and languages have integrated support for this method
• All content is under your control and only your servers (not outside servers) are sending content
• Many translation management services provide tools for organizing this content, like Pootle and Gengo partners Phrase, Crowdin and Transifex
• SEO works the same as it does for your source language content

Cons:

• Can be difficult to extract content out into localization files if this hasn’t been done from the start
• Exporting the content and importing into a translation management system (TMS) can be time-consuming
• Code deployments can be slowed down by the need to incorporate a localization workflow prior to releases

Proxy server approach

Another common localization technique is the proxy server approach, where all localized content is served from a layered service provider (LSP) or TMS when a request comes to your site. This typically means that you’ll set up domain name system (DNS) entries for different languages and let the other service deal with each request. The service will then monitor the source content for updates, request new translations and serve new pages as needed.

Pros:

• Easy for anyone with DNS access to set up
• All translated content is stored in one place
• Updates happen with very little if any involvement

Cons:

• Site content may be out of sync with source language at times
• Dependent on hosting services’ uptime
• Since the page is being served by a different domain, some search engines will penalize the content in rankings (note: this can be mitigated by using a proxy of your own to request the translated page and then serve it, but this can be more complicated)
• Can increase response times quite a bit for all translated content
• Can get a little clunky for applications requiring session management or containing a large amount of dynamic content

JavaScript-based approaches

More recently, services like LocalizeJS, Wovn.io and Transifex Live began taking a different approach. By first indexing the site, they extract all translatable content into their TMS. When the site is loaded, a single line of JavaScript will request the translations asynchronously and replace source content found in the DOM. Traditionally, this technique meant that search engines wouldn’t see any of the translated content, but a recent change to the way Google indexes webpages can make this a feasible solution.

Pros:

• Very easy to implement on almost any site
• All translated content is stored in one place
• Updates happen with very little if any involvement
• Pages are served from a single domain

Cons:

• Only Google executes JavaScript before indexing. Yahoo Search, Baidu, Microsoft Bing and other popular search engines will still not index the translations
• Adds additional load time when site is rendered
• Dependent on hosting services’ uptime

Of these three methods, the traditional technique of managing flat files and serving localized content from the same server continue to offer the most reliable SEO benefits. With recent indexing changes that Google has introduced, too, there’s new hope for a less complicated solution in the future.

Interested in translating your site, but not sure where to start?

Learn More   or Contact us