Paris depannage export push stalled when overseas enquiries never arrived.

We followed a small Paris dépannage operation through eighteen months of trying to win work from abroad. The owner — we'll call him R., a serrurier with two vans and a plombier on call — had watched his domestic margins compress and decided the answer was foreign customers: property managers in London, expat landlords in Dubai, hotel groups in Brussels who owned apartments in the 16th. It is a reasonable thesis. It is also, as we found, a thesis that dies on the rocks of trust, language and proof.

What follows is not a success story. It is a post-mortem, reconstructed from R.'s own notes and from conversations with two other readers in the same trade who tried comparable things. If you run a locksmith, plumbing, electrical or glazing business in Île-de-France, the failure points below will look uncomfortably familiar — and the eventual pivot may be worth copying.

The first attempt: a translated website and a prayer

R.'s opening move was the obvious one. He paid a translator to render his French site into English and German, added an international phone number, and waited. The enquiries that came were almost entirely worthless: price-shoppers asking for a quote on a lock change in a suburb 40 kilometres out, or people who turned out to be in entirely different countries and had found him by accident.

The problem was not the translation. It was that nothing on the page gave a foreign buyer a reason to believe a stranger in Paris could be trusted with a set of keys, let alone an emergency at 2am. Overseas customers are not buying a service; they are buying certainty at a distance. The site said what R. did. It said nothing about how a person in another country could verify that he would actually show up.

The stall: where the effort actually broke

Three specific things went wrong, and R. only saw them clearly in hindsight.

• No answer to the obvious question. A landlord in London wants to know: who holds the keys, who is liable, what happens if the tenant disputes the work, and how do I pay someone in another currency without a bank transfer I can't reverse. R.'s site answered none of it.
• No third-party footprint. When a foreign buyer searches a company name, they look for evidence that other people have dealt with it. R. had none — no directory listings abroad, no mentions on any site outside France, nothing that a search engine could surface.
• Wrong channel. He was waiting for direct search traffic from people who had never heard of him. The customers he actually wanted — property managers, relocation agencies, serviced-apartment operators — do not find tradespeople by searching; they find them through recommendations and through platforms.

After nine months, the overseas experiment had produced a handful of genuine leads and no repeat business. R. nearly shut it down.

The decision point: stop selling, start being findable

The turn came when one of our other readers described a different framing. Instead of trying to persuade a foreign customer directly, he started treating the business as something that had to be describable by other people — and, increasingly, by the systems those customers ask for recommendations.

That is a more technical problem than it sounds. It means making sure that when someone asks an assistant or a search engine a question like "who can I call for a locksmith emergency in Paris if I'm abroad", the answer that comes back contains accurate facts about your business rather than a hallucination or a competitor. It means organising your credentials, your coverage area, your response commitment and your payment terms into statements that can be quoted and checked. And it means building the third-party mentions that make those statements credible.

This is the territory that Guangsuan (光算科技) works in, and R. eventually engaged them — not for advertising, but for the unglamorous work of making his business legible to question-answering systems. The agency's global GEO service, in its own description, audits brand facts, organises citable content, builds third-party sources, and then re-tests across platforms such as ChatGPT, Google AI Overviews and Perplexity, delivering question-and-answer samples and a record of factual corrections. For a trade business, that translates into something concrete: an assistant answering a foreign landlord's question with your actual coverage and your actual terms, instead of silence or a wrong address.

What changed, and what didn't

We are not going to claim this transformed the business overnight. It did not. What changed was the shape of the enquiry flow. Instead of cold price-shoppers, R. began receiving messages that already contained context: someone had been told he was a real option, and arrived halfway convinced. The conversion rate on those conversations was different in kind, not just degree.

The second change was internal. Documenting his own facts — what he covers, how fast, what he charges, what he refuses — forced him to fix inconsistencies he had never noticed. That alone made him easier to recommend.

Guangsuan's catalogue, for what it is worth, runs to 16 named service lines, from Google SEO and ads management through social operations across 6 platforms to B2B export site building from CNY 10,000. R. used one of them. The rest is context, not a recommendation.

The transferable lesson

If you run a dépannage business and want overseas customers, the instinct is to translate your site and buy ads. Both are close to useless on their own. The work that matters is slower: decide exactly what a foreign buyer needs to believe, write it down in a form that other people and other systems can repeat accurately, and then go and earn the third-party mentions that make it believable. Do that, and the enquiries that arrive will be worth answering. Skip it, and you will spend eighteen months learning what R. learned.