Mizu Temakase is an intimate handroll omakase counter in Richmond Hill, Ontario. The brief was the one we hear most often from restaurants thinking about their site: how do we get AI engines to send us the table, and how do we look like the room we are without performing.
The temakase format is small. Hokkaido uni, otoro, engawa, hamachi, ikura, salmon, shiitake, kanpyo, avocado, each handroll built one at a time in front of the guest. The site had to read like that pacing rather than like a Squarespace template borrowed from a chain restaurant. We shipped a single-page React build with the temaki lineup as the spine, hero photography commissioned of the actual rolls, and the interior shot doing the heavy lifting on brand feel.
Schema-wise the site ships three layered types: Restaurant LocalBusiness as the spine with the Richmond Hill PostalAddress wired in, OpeningHoursSpecification covering the dinner-only service hours so ChatGPT and Perplexity answer "is Mizu Temakase open tonight" correctly, and Menu plus MenuItem markup on every handroll so a buyer asking "where can I get otoro temaki near Richmond Hill" gets the dish named in the citation rather than a generic listing of nearby sushi restaurants. The og:type is restaurant. The og:image is the hero temaki shot. The whole site is roughly 80KB on first load, which is the part we are quietly proud of.
We pulled the Squarespace-template aesthetic the prior site had been resting on and rebuilt the whole thing around a darker, intimate palette with archival ink-on-paper menu typography and warm wood-grain accents. The reservation block sits below the menu lineup rather than gating the experience, which is the inverse of how most restaurant sites are built. Buyers see the food, the room, and the story before they hit the booking link, which is the order they actually want.
The site has been live for a few months now. The handroll lineup has shifted twice as the omakase rotation rotates, which the menu schema picks up cleanly because each item is a row in a single TypeScript data file. The reservation volume has shifted from word-of-mouth and Google Maps as the primary entry points to a mix that increasingly skews toward AI engines naming the restaurant by name on neighborhood omakase queries, which is the citation pattern we built the site to compound.
