Nozawa Hospitality

The Mission

Nozawa Hospitality runs fifteen places to stay in Nozawa Onsen, plus a few bars and restaurants, and manages properties for owners who live elsewhere. They came to us with a brand they liked and a set of goals it could not reach on its own: read clearly as an accommodation business, rank in two languages, bring every property under one roof, and be editable by the people who work there.

The Outcome

A rebuild in Framer that keeps the existing brand and changes everything under it. A homepage that opens on booking a room, calls to action that reach the booking platform, property pages with real depth to them, a page structure built from what people actually search for, a Japanese version written for Japanese search, a migration that carried the old site's traffic across, and guides so the team can run it without us.

The Impact

It is now obvious within a second what the company sells, and a guest who is ready to book can do it. Fifteen properties sit together under one brand, each with a page detailed enough to book from, and the team can edit all of it in both languages without calling anyone. The search position built up over years survived the move, which is the part nobody notices when it goes right.

Plenty of redesigns start by throwing out the brand. This one did not need to. Nozawa Hospitality already had a clean look, a logo that worked, and photography of the properties that had recently been reshot. Rebuilding that would have been spending the client's money on something they already had.

So the visual language stayed and the work moved underneath it. What the site needed was structure: a clearer job to do, every property in one place, a Japanese version that could rank on its own, and a setup the team could edit themselves. None of that is solved by a new colour palette.

One thing genuinely was not landing. Nozawa Hospitality is a company you book a bed with, but guests kept reading it as something closer to a village welcome service, a general source of information about Nozawa Onsen rather than a place with fifteen properties to sell. The name invites that reading, and a site full of good village content reinforces it.

So the homepage now opens on booking a room. That is the first thing in the hero, before the village, before the seasons, before anything else, and it resolves in a second what used to take a few scrolls to work out. The village content did not go anywhere, it just moved to where it helps, which is underneath the offer rather than in front of it.

A hero that leads with booking has to actually book. So the calls to action across the site now go through to the company's booking platform rather than to a contact form. A form is the right tool for a question about a group, a late arrival or an owner enquiry, and it is still there for those. It is the wrong tool for someone who already knows their dates, because it turns a booking into an enquiry and puts a person and a wait in the middle of it.

The other half of that is depth. Every one of the fifteen properties now has its own gallery, floorplans, a full amenity list, where it sits in the village, and a description long enough to tell you what staying there is actually like. One has private hot baths. Another is a short walk from the lifts. Those differences are the entire reason a guest picks one over another, so the templates were built to carry them rather than flatten every property into three lines and a photo.

That depth is also what let the properties come home. A few had drifted onto separate websites of their own, and consolidating them only works if the page they arrive at is at least as good as the site they left. The same thinking runs through activities, events, bars and restaurants, and a village FAQ well over a hundred questions deep.

Before a page got designed we went through twelve months of their search data to find what the site already earned and where demand was going unanswered. That became the page structure, the titles, and the metadata for every page, each one aimed at a query the site was already showing up for.

The most useful finding was one nobody expected. Taiwan is their third biggest market and their best converting one, ahead of the US, UK and Canada combined, at the highest click-through rate of any market they have. The query data also settled practical questions, like which word Taiwanese guests use for a hotel, and the fact that they already search the properties by their Latin names, so those names are better left alone than translated. Research like that is cheap, and it decides a lot.

A redesign is the easiest way in the world to throw away years of search traffic, and this one had more exposure than most, since the old site ran across four language hosts and every surviving page needed its redirect on all four.

The blog was the interesting part. It was a small share of clicks but a quarter of all impressions, and several posts ranked better than the site pages covering the same subject. The plan was to retire the blog, which was still right, but only if the pages absorbing those posts inherited the actual content instead of a link to it. That turned a redirect list into a content plan, and it fed straight back into the property and village pages. The structured data got the same treatment: there is another company in the village with a nearly identical name, the one that runs the ski resort, and spelling out the legal entity and its subsidiaries is what keeps search engines from merging the two.

Japanese is around 40 percent of their clicks and it leads the searches for summer, transport, food and festivals. That is too much to hand to a language toggle, so the Japanese version is built as its own site with its own search presence.

It was written rather than translated. The headings carry the company name, because that is the phrase people search for. The founding story reads as a story instead of a specification. The stat about staff numbers swinging with the season keeps the swing, since the swing is the point. Around four hundred CMS entries came across with it: properties, activities, events, restaurants, amenities, map points and the FAQ.

Winter and green season are not the same audience twice. Winter is international, heavily Australian and New Zealander, and books a year ahead. Green season is domestic Japanese, more competitive, and books late.

Rather than build a site that splits the difference, the hero and the imagery swap between a winter mode and a green season mode. The part that matters is that switching it is one dropdown, so it happens when the snow arrives rather than whenever an agency gets around to it.

This was the actual product. Everything on the site that changes lives in a CMS collection, so adding an event or updating a property's gallery is filling in a field, not booking developer time.

On top of that we wrote guides for the jobs the team will really do: switching the seasonal hero, editing pages, updating galleries, working in the Japanese locale. They are written around the traps rather than the happy path, because the sentence worth having is the one warning you that the canvas will not appear to change when you flip the season, and that Publish does not stop to ask whether you are sure.