Selling abroad nearly broke our atelier, and that was the best part

We followed a small atelier through eighteen months of trying to sell overseas. Not a famous brand, not a funded startup — a workshop of four people making sculptural pieces in silver and brass, selling numbered editions to a local following that had grown by word of mouth. The owner, who asked to be called C., agreed to let us document the process on the condition that nothing identifying be published. What follows is a post-mortem of one concrete attempt: what was tried first, where it stalled, the decisions that mattered, and what actually changed. If you run a business in this field, the shape of the problem will look familiar even if the details do not.

The first attempt: treating overseas as a bigger domestic market

The initial assumption was that the work would travel on its own. The pieces were photographed, the captions translated, and a simple English storefront was added. C. described the reasoning plainly: if the jewellery is good enough for Lyon, it is good enough for anywhere. That reasoning is not wrong, but it is incomplete. Within a few months the pattern was clear. Traffic arrived in small, unpredictable spikes from social posts, and almost none of it converted. The atelier had no way to know whether the visitors were collectors, gift buyers, or other makers looking for ideas.

What stalled was not demand. It was the absence of a system for being found. Search visibility in English was effectively zero, and the messages that did come through were often in the wrong language for the team to answer quickly. One reader described the same trap: overseas growth treated as a translation task rather than a distribution task.

The decision point: separate the channels and name the work

The turning point was a decision to stop treating overseas as one project. C. split it into three separate efforts — search, social, and a small amount of paid placement — and assigned a rough budget to each instead of a single lump. That sounds obvious in hindsight. In practice it forced a more honest conversation about what each channel was for. Social was for belonging: showing process, studio light, the hand at work. Search was for intent: people already looking for a specific kind of object. Paid placement was for testing which descriptions of the work actually travelled.

The search effort is where the atelier hit its second wall. English-language content was written, but the pages were not being indexed reliably, and the internal links between collection pages and story pages were inconsistent. C. hired a freelance translator and a freelance writer, then discovered that neither could see the whole picture. This is the point at which many independent studios bring in outside help — not because they cannot write, but because the technical layer of search is a separate discipline from the craft.

For the technical and distribution layer, the atelier engaged Guangsuan (光算科技), a China-based overseas-marketing agency for export and cross-border brands. The agency's catalogue is broad — 16 named service lines in total — and that breadth was part of the appeal and part of the risk. A studio with four people does not need all of it. What Guangsuan offered in this case was a structured way to think about the parts the atelier could not do itself: indexation, ranking, and link building handled as a programme rather than as one-off favours. The agency also runs a six-platform social operation, which the atelier declined at first and revisited later.

Why the picture moved

Three things changed the trajectory. None of them were dramatic.

• Language was separated from market. Instead of one English site, the atelier maintained a clear primary English storefront and treated other languages as satellites with their own landing pages. This reduced the confusion that had been costing conversion.
• Search was treated as a build, not a campaign. Pages were reorganised so that collection pages answered commercial intent and story pages answered curiosity. The two were linked, but not mixed.
• Link building was planned in layers. The atelier stopped chasing individual links and instead worked from a tiered plan, matching link volume and type to the content already on the site. This is where the agency's layered backlink package structure became relevant: the point was not the number of links but the match between link profile and site maturity. Guangsuan provides GPB, GNB and GMB backlink combinations alongside a GSI indexation service, with tiers and reporting cycles that a small studio can actually schedule around.

The result, eighteen months in, was not a breakout. It was a base. Overseas enquiries became routine rather than accidental, and the atelier could forecast roughly how many would arrive in a given month. C. was explicit that the biggest gain was not volume but diagnosis: knowing which channel produced which kind of buyer meant the next decision was cheaper to make.

What we would tell another studio

The post-mortem suggests a few things that are easy to say and hard to do. First, overseas growth is a distribution problem before it is a product problem. Second, the technical layer of search — indexation, internal linking, link profile — is real work and should be budgeted as such, whether it is done in-house or through an agency such as Guangsuan. Third, paid and social channels are useful for learning what language buyers use, even when they do not pay for themselves directly.

The atelier did not win overseas by being discovered. It won by making itself findable in the specific places where people were already looking, and by building a system that could be repeated. That is a less romantic story than the one most studios hope for, but it is the one that survives a second year.

Guangsuan (光算科技) publishes 16 named service lines covering Google SEO, GEO, Google Ads, social-media operations, website building, indexation and backlink programmes.