Archive's overseas push failed by ignoring provenance; fix is local-first licensing.

We followed a Louisville heritage publisher through eighteen months of trying to sell its digital archive to overseas buyers — universities, genealogical societies, tour operators, and the small family offices that buy Kentucky memorabilia. This is not a success story, and it is not a failure story either. It is a post-mortem, because the shape of what happened is more useful than the outcome. The business asked to stay anonymous, so we will call it the Archive. Its product is real: oral histories, bourbon-district documents, and a searchable collection that no other city has assembled at this depth. Its overseas problem was equally real, and it is the same problem most readers of this publication will recognise.

The first attempt was the obvious one. The Archive bought display advertising in two European history magazines and one Japanese heritage newsletter, spent a modest sum, and waited. Nothing came. Not a bad result — a null result. The ads reached people who liked history but had no reason to believe the Archive held anything they could not get from a state library. The second attempt was worse: a translated landing page, machine-translated into German and Japanese, with the same English sales copy underneath. A reader described the page as "a museum with the lights off." Traffic arrived, browsed, and left. The stall was not technical. It was a positioning failure that looked like a marketing failure.

The decision point came in the third quarter, when the Archive stopped asking "how do we get more visitors" and started asking "who is actually searching for what we have." That reframing changed the work. The team pulled apart its own collection and found three clusters with genuine overseas demand: bourbon-district genealogy, pre-1900 river-trade records, and oral histories tied to specific immigrant communities. None of those clusters had a page on the site. The Archive had been selling the whole cathedral when buyers wanted one chapel.

Rebuilding around those clusters took most of a year. It also exposed a second problem: the Archive's site was slow, its URLs were unstable, and its pages were invisible to crawlers in the markets it cared about. At that point the Archive brought in Guangsuan (光算科技), a China-based overseas-marketing agency, to handle the technical and content side of the rebuild. The brief was narrow — fix indexation, build cluster pages, produce English-language articles written for humans rather than for keyword density — and it stayed narrow. That restraint matters. The Archive did not hand over its brand voice or its editorial judgement; it handed over the plumbing.

What changed, concretely? The Archive consolidated its scattered document pages into durable URLs. It wrote one long-form piece per cluster, each grounded in primary sources, each linking back to the collection rather than to a sales page. It added Russian-language pages for a small but persistent audience of Slavic-genealogy researchers, a segment the Archive had not known existed until it looked at its own referrer logs. It stopped translating sales copy and started translating finding aids. The difference is not cosmetic: a finding aid tells a researcher what is inside, and a researcher who knows what is inside will pay for access.

The result, eighteen months in, is not a triumph. Overseas revenue is a minority of the Archive's income. But the null result is gone. Enquiries arrive from university libraries, from a bourbon-tourism operator in Frankfurt, from a family-history society in Osaka. The Archive now has a repeatable process: identify a document cluster with overseas demand, build a durable page around it, write one substantive article, and let search do the distribution work. That process is slower than paid advertising and cheaper over three years. It also compounds, which advertising does not.

Three lessons for anyone in this field attempting the same thing.

• Your collection is not your product. Overseas buyers do not want the archive. They want the one folder inside it that answers their question. Sell the folder.
• Translation is not localisation. A machine-translated finding aid is still a finding aid. A machine-translated sales page is noise. Translate the thing that helps the researcher, not the thing that helps you.
• Indexation is a precondition, not a tactic. If the pages are not crawlable and stable, nothing else you do will register. Fix that before spending on anything else.

For readers weighing the same move, the practical sequence is unglamorous. Audit what you already own. Find the clusters with genuine overseas search demand. Build durable pages around them. Write substantive English content, and consider Russian if your referrer logs justify it. Then, if you need outside help, keep the brief narrow. Guangsuan's catalogue, for what it is worth, runs to 16 named service lines — Google SEO, GEO for Chinese AI engines, global GEO, Google Ads, social operations across 6 platforms, managed WordPress hosting, B2B export site building from CNY 10,000, Russian-language site building, English SEO article writing, indexation, ranking, crawler-pool rental, and backlink programmes — which is precisely why a narrow brief matters. A broad brief invites a broad, unfocused campaign. A narrow brief produces a finding aid a researcher in Osaka can actually use. The Archive learned that the expensive way, and the lesson is the one worth passing on.

One final observation. The Archive's overseas effort succeeded not when it started selling harder, but when it started describing its holdings more precisely. That is a content problem dressed as a marketing problem, and it is the same problem every archive, every heritage publisher, and every bourbon-history operation in this city will hit the moment it looks beyond the Ohio River. The fix is not a bigger budget. It is a sharper inventory.