How to Translate a Tilda Blog into Multiple Languages
In this article: why a Tilda blog is harder to translate than regular pages, how streams work and why they are needed, implementation options, and how a proxy solves the problem without duplicating the project.
The client wants a blog in Russian and English. You've already enabled multilingualism on other pages — and then it turns out that the blog is a separate story.
Why a Blog on Tilda Is a Separate Task
Regular Tilda pages are blocks of text. The proxy intercepts the HTTP request and replaces the content — Tilda knows nothing about it.
This won't work with a blog. A blog on Tilda works through collections: each article is an entry in a collection, and the collection is linked to a project. When a user opens /blog, Tilda dynamically serves entries from the collection of the current project.
The English version of the blog is not a translation of pages. It is a separate set of articles in a separate collection.
Three options that don't work
Simply translate pages. The proxy will translate the interface: the "Blog" title, buttons, meta tags. But the articles — titles, announcements, content — will be taken from the same Russian collection. The reader lands on a page with an English interface and Russian articles.
Create a /en/blog/ folder. Tilda allows you to create folders, but all pages of one project use the same language settings. A separate language blog within one project will not work.
Duplicate the project. Technically, it works. But any template change will have to be done twice. Add a third language — three times. With five languages, it becomes unmanageable.
How it works through a proxy
Multify solves the problem at the routing level. The user opens mysite.com/en/blog — the proxy recognizes the request and takes entries from the English collection. Tilda works as usual, just with its own stream connected for each language.
The language switcher on the article page leads to the translation of the same article, not to the blog's main page. This works through slug mapping: if the slugs match or explicit mapping is set, the transition is correct. Search engines also see explicit interlinking via hreflang at the level of each individual article.
Real articles are needed for each language version. Multify can automatically machine translate the collection, but it's worth checking before publishing — especially headlines and announcements, as they have a stronger impact on CTR in search.
Also needed: translated slugs if you want meaningful URLs in each language, and meta tags for each article in the corresponding language. Multify automatically picks all of this up from the collection.
Can I translate an existing Tilda blog without creating a new collection?
No. Blog articles are stored in a collection linked to the project. Each language version requires a separate collection with translated articles.
How does a search engine know that Russian and English articles are versions of the same material?
Through hreflang at the level of each individual article. The proxy automatically generates hreflang provided that the slugs match or an explicit mapping between versions is specified.
Is machine translation of blog articles acceptable for SEO?
Machine translation is acceptable as a draft. Google indexes high-quality, edited text better. Titles and announcements affect CTR — they should be edited first.
What happens to comments on articles in a multilingual blog?
Comments are tied to a specific page. If you use Tilda's built-in comments, they will be separate for each language version. Third-party services (Disqus) can be configured for a shared stream.
Launching a Multilingual Blog on Tilda
We will set up a blog with folders, automatic hreflang, and correct interlinking between language versions of articles.