Should Your Portuguese Business Website Speak English Too?

Portugal received over 30 million foreign visitors in recent years, Lisboa and Cascais have some of the highest expat densities in Europe, and Portuguese exporters sell into markets where nobody reads Portuguese. So the answer is obvious, right? Every site should have an English version.

No. For a good share of Portuguese small businesses, an English version is money spent on pages nobody visits — and a badly translated one is worse than none at all. Here is how to decide, and how to do it properly if the answer is yes.

When a second language actually pays

A bilingual site earns its cost when a meaningful slice of your buyers — not your traffic, your buyers — cannot comfortably read Portuguese. In Portugal, that usually means one of these situations:

Situation	Why English pays	Typical examples
Tourism zones	Customers research before arriving	Restaurants, tours, surf schools, rent-a-car in Algarve, Lisboa, Porto, Madeira
Alojamento local / hotels	Direct bookings bypass platform commissions of 15–18%	Guesthouses, AL apartments, rural tourism
Expat-dense areas	Residents who live here but search in English	Clinics, gyms, lawyers, accountants in Lisboa, Cascais, Oeiras, Lagos
B2B export	Buyers abroad evaluate you online first	Manufacturers, software, wine producers, industrial services
Relocation services	The entire audience searches in English	Immigration lawyers, real estate for foreigners, international schools

And when it usually does not pay:

• A local service business whose customers are Portuguese — a plumber in Braga, a driving school in Viseu, a daycare in Almada. English pages here get near-zero visits.
• "Maybe someday we'll export." Build it when the export plan exists. Adding a language later is a normal, plannable project.
• "It looks more professional." An empty English section with three translated pages looks less professional, not more.

A quick test: open Google Analytics (or ask whoever manages your site). If under roughly 10% of your sessions come from non-Portuguese browsers or foreign locations, and you have no concrete plan to change that, skip English for now and spend the money on the Portuguese site.

Machine translation is a trust-killer

The tempting shortcut is a free translation plugin or pasting your pages through an automatic translator. Here is what actually goes wrong, with the kind of errors we see on real Portuguese sites:

• Menu items become nonsense. "Entradas" rendered as "Entrances" or "Tickets" instead of "Starters". A wine list where "vinho da casa" becomes "wine of the house" — readable, but instantly signals nobody checked.
• Legal-ish pages get dangerous. Automatic translation of your terms, cancellation policy or AL house rules can change meaning. A refund clause that says something different in English is a real dispute waiting to happen.
• Forms and buttons break trust at the worst moment. The visitor gets to the booking form and hits "Send the demand" instead of "Send request". Right at the point of conversion, your site suddenly reads like a scam page.
• Tone collapses. Portuguese formal register ("Entre em contacto connosco") machine-translated turns stiff or oddly casual. Your carefully written positioning becomes generic mush.

The pattern: machine translation fails most often exactly where money changes hands — menus, prices, policies, forms. A visitor who spots one absurd translation stops trusting the prices and the reviews too. Studies on multilingual e-commerce consistently show a strong majority of buyers prefer purchasing in their own language when the text reads natively; a broken translation removes that advantage and adds doubt.

Machine translation is fine as a first draft that a fluent human then rewrites. It is not fine as the published product.

Adaptation, not translation

The English version of your site should be written for the English-reading customer, which is not the same as translating the Portuguese sentences:

• A Portuguese visitor to a Cascais clinic knows how the SNS works; a British expat needs one paragraph on private appointments and insurance.
• Your PT homepage might lean on being "empresa familiar desde 1987"; the EN version for tourists may need "10 minutes from Lisbon airport" first.
• Prices: Portuguese customers expect IVA-included prices by default. Foreign B2B buyers often expect prices stated ex-VAT — say which it is.
• Keywords differ. Portuguese users search "alojamento local Lagos"; English users search "holiday apartment Lagos". Translated keywords are usually the wrong keywords.

Practical rule: translate facts, rewrite persuasion.

The technical side, in plain words

Two things matter and both are cheap when planned from day one.

URL structure. Each language needs its own addresses. Three sane options:

Option	Example	Best for
Subdirectories	feitura.pt/pt/... and feitura.pt/en/...	Most small businesses — one domain, one authority, simplest
Subdomains	en.yoursite.pt	Rarely worth the extra setup for SMEs
Separate domains	.pt + .com	Companies with genuinely separate country operations

Subdirectories win for almost everyone. Avoid the fourth option some plugins use — the same URL that swaps language via cookies — because Google effectively sees only one language.

Hreflang. A small tag in each page's code that tells Google "this page has a Portuguese twin here and an English twin there". Without it, Google may show your Portuguese page to a searcher in London, or treat the two versions as duplicate content. It is a few lines per page, trivial when the site is built with it in mind, tedious to retrofit onto a site that never planned for it. If a developer builds you a bilingual site, hreflang should be included without you asking — it is table stakes, not an extra.

The parity trap

The most common failure of bilingual sites is not launch — it is month six. You update the Portuguese menu, add a new service page, change your prices... and the English version quietly falls behind. Now your English visitors see last year's prices. That is worse than having no English version.

Defend against it with a simple rule set:

1. Decide what is bilingual and what is not. Core pages (home, services, prices, contact, booking) — always both languages. Blog and news — Portuguese only is a legitimate choice; just don't show empty English sections.
2. One update = two updates. Whoever changes a price on the PT page changes it on the EN page in the same sitting. Make it a checklist item, not a memory item.
3. Quarterly parity check. Twenty minutes: open both versions side by side, scan prices, hours, and services.
4. Keep the second language small if maintenance capacity is small. Five accurate English pages beat thirty stale ones.

What it honestly costs

• Content: budget roughly +30–40% of the content effort. Not +100% — the structure, photos and design are shared — but real adaptation by someone fluent is not free.
• Technical: small, if planned from the start. Language switching, hreflang, and duplicated page templates are a modest line item on a new build. Retrofitting multilingual support onto a site built single-language can cost more than the original site did.
• Market context (2026): in Portugal, freelancers charge roughly €350–1.750 for a site and agencies €3.000–6.000 for an institutional one — and multilingual support is frequently a paid extra or an afterthought at both ends. Ask explicitly what "bilingual" includes: hreflang? language-specific URLs? who writes the English?

At Feitura we build sites bilingual by default — PT/EN structure, hreflang and language-aware URLs are part of the base build, not a paid extra — so the only real decision left is the content: which pages, and who writes the English well.

If you're weighing whether your site should speak English — and what that changes in the budget — get a concrete estimate for your case and we'll tell you honestly if one language is enough.