What does it actually take to run a site in three languages?
It always arrives the same way. We are opening in Brazil and Spain, can we just add the languages? Yes. Webflow does this natively now and the build is genuinely not the hard part. The hard part is that you have just tripled the size of every content decision you will make for the next three years, and that never shows up in a quote.
The honest version
Adding a locale is cheap. Webflow sells localization as a per locale add-on, roughly ten dollars a month per locale at the entry tier and around three times that at the tier which unlocks localized URLs, with the exact figures depending on whether you pay annually or monthly. Take those from Webflow's pricing page rather than from this post. Platform pricing moves, and that number is the least durable thing on this page.
It is also the smallest. Everything after it is people. Every landing page you launch is now three landing pages. Every pricing change is three changes. Every blog post has a translation request stapled to it. The platform fee is the cheapest line in a multilingual site and it is the only one anyone budgets for.
Three things worth knowing before you decide
Most of what you will read about Webflow's localization is a tour of the panel. Three details actually change how you plan the project.
- Hreflang tags and the localized sitemap are generated for you. This is the part teams get wrong on other platforms, and getting it wrong means search engines serve the wrong language to the wrong market for months. Webflow handling it removes a real category of risk, so do not spend budget solving it.
- Locales publish independently. English can ship while Portuguese is still in draft. A late translation never blocks a launch, which changes how you sequence a rollout.
- Localized URL slugs sit on the Advanced tier. If a product URL stays in English on your French site, you are competing for French search with an English address. If organic search in the second language is part of the business case, the cheaper tier is not the cheaper option.
What Webflow will not do for you
It will not tell you that your Spanish pricing page is three revisions behind the English one. There is no drift alert, no staleness flag, nothing that surfaces the gap. The product assumes somebody is watching.
And machine translation gives you a draft, not a page a buyer trusts. Auto translated meta titles and descriptions get indexed and served to real people, and the version they read is the one nobody on your team ever looked at.
What this looked like on K2BTools
K2BTools sells developer tooling for GeneXus, and its market is actually split across Spanish, English and Portuguese. Not aspirationally split. Actually split, with buyers in each.
The decision that made the site survivable was structural rather than linguistic. Build it component first. When a feature card is a component, changing it is one change that lands in all three locales, and the only per locale work left is the words inside it. When it is not a component, the same change is three edits, and eventually somebody does two of them and the third sits wrong for a year.
That is the whole trade. You spend more at build time defining components than a single language site would need, and you get it back every time the content changes, which on a marketing site is constantly.
Three questions before you add a second language
Answer these before anyone quotes you. None of them is a technical question and they decide more than any technical answer I could give.
- Who owns the second locale, by name? If the answer is marketing, it has no owner.
- How often does the content change? Weekly content across three locales is a role, not a setting.
- Is the market real or aspirational? A locale with no sales motion behind it is a monthly bill attached to a page that slowly stops being true.
Where multilingual sites actually rot
Not in translation quality. In the gap that opens after launch. English gets a new pricing page in March. Portuguese still shows the old one in September. Someone in Sao Paulo lands on the stale version precisely because hreflang worked and sent them to their own language.
That is a process problem, so it needs a process answer. One named owner per locale, and a rule that the primary locale does not publish a structural change until the others are queued. Neither of those is something you buy.
When the answer is no
If one market is most of your revenue and nobody will own the second locale, do not localize the site. Localize one page, the one that market actually converts on, and leave the rest in English. A single maintained page in someone's language beats a whole site that stopped being true a year ago.
And if the plan is machine translation with no native reviewer, English only is the better product. A buyer reading obviously machine translated copy concludes you did not take their market seriously, which is worse than concluding you have not got there yet.
So what does it take
Webflow makes the mechanics close to boring, which is what you want from infrastructure. What it cannot do is make the difficulty disappear. It moves the difficulty back to where it always was: having something to say in each language, and someone whose job it is to keep saying it.