Adding a second language to a website looks like a translation job and behaves like a second website. Every page, form, confirmation email and campaign landing page needs a twin, and the twin has to stay in step for as long as the business exists. Growth Bully, a Malta performance marketing agency, gets asked for one most often by businesses who have never checked whether anybody is searching in that language.
The question is not whether a second language would be nice to have. It is whether there is demand you are measurably losing today, and whether you can keep two versions alive once the launch excitement fades.
Does your website need a second language?
Only when there is measured demand in that language that you are currently losing. A market being bilingual is not evidence. Evidence is search queries, ad impressions or enquiries arriving in that language. Without it, a translated site adds maintenance, dilutes your pages and changes nothing about lead volume.
Malta is the clearest example of the trap. Almost everyone here reads two languages, so a second version feels obviously correct. In our own search research across the local market, Maltese language commercial queries returned no Maltese language results at all, and the pages that ranked were English without exception. Buyers were searching in the language of business, not the language of conversation. That finding is directional rather than absolute, but it points the same way every time we test it.
How do you check whether the demand is real?
Three places tell you quickly. The search terms report in your ad account shows the exact words people used before clicking. Search Console shows the queries you already appear for. Your own enquiry inbox shows the language buyers write to you in. If all three are one language, you have your answer.
- The search terms report. Already paid for, already specific, and it needs no estimation. The same report that drives keyword selection answers the language question for free.
- Search Console queries. Filter for the second language. If impressions are near zero after a year of publishing, there is no hidden audience waiting.
- Your inbox and your calls. Count how many enquiries arrive in the other language. This is the only sample that includes people who chose to buy.
- Ad tests before builds. Run a small set of ads in the second language to an existing page. If nobody clicks, nobody was waiting for the website.
None of this takes a week. It takes an afternoon, and it is the difference between a translation project with a reason and one with a hope.
Which pages should get a second language first?
Start with the pages that take money or capture leads, never the blog. A buyer can tolerate reading about your history in a second language, but not a quote form they cannot complete. Translate the service page, the contact path and the confirmation messages, then stop and measure before adding more.
- The one service page that generates the most enquiries.
- The enquiry or booking form, including its error messages.
- The confirmation page and the automated email that follows it.
- The pricing or process page buyers read before they commit.
- Everything else, only once the first four have produced enquiries.
That order exists because a half translated journey is worse than an untranslated one. A visitor who reads a page in their language and then hits a form in another concludes the business is careless, not bilingual.
What does a second language cost after launch?
Roughly double the ongoing work, not double the build. Every price change, new service, campaign landing page, legal update and confirmation email now happens twice, forever. The build is a one off cost that people budget for. The maintenance is a permanent one that almost nobody budgets for, and it is where second languages quietly rot.
The technical side is the easy part. A stable URL pattern per language and correct hreflang relationships are a morning of work for a competent developer, and your web build should include them by default. The hard part is editorial ownership: one named person who updates both versions in the same week, every time, or the second version drifts into a museum of last year promises.
Should ads run separately for each language?
Yes, separate campaigns or at least separate ad sets, because you cannot read performance otherwise. Mixed language traffic inside one campaign hides which audience is converting and which is burning budget. Separating them also lets each language have its own landing page, which is the part that decides whether the click becomes an enquiry.
Keep the budgets honest too. A second language campaign that gets a tenth of the spend will look like a failure when it was never given enough volume to be read. Decide a test budget, run it long enough to see conversions rather than clicks, and judge it on cost per enquiry against your existing main language campaigns.
When is a second language the wrong call?
When the second language is aspirational rather than observed. Common versions are translating for a market you have never sold to, translating because a competitor did, or translating to look international. Each one adds pages that nobody searches for, splits your authority across duplicate content, and gives you twice as much to keep current.
There is almost always a better use of the same budget. Fixing the enquiry path, writing the pages you are missing in the language that already converts, or improving the visibility you already have will usually produce more leads than a mirrored site nobody asked for. Translate when the evidence tells you to, and spend the money on demand you can see until it does.
If you are weighing a second language against other growth work this year, tell us what the evidence shows and we will tell you which one gets you more enquiries.