A neighborhood restaurant, found again
A new site and search presence so a well-loved restaurant shows up when people — and AI assistants — look for dinner nearby.
- Client type
- Local restaurant
- Problem
- A fifteen-year-old restaurant with a loyal following and a website that hadn’t changed since 2016. The menu was a PDF, hours were wrong on Google, and nobody under thirty could find them.
- What I built
- A fast new site with a live menu, hours, and reservations; cleaned-up Google Business and Apple Maps listings; structured data and an llms.txt so AI assistants describe them accurately.
- Result
- Reservations up, the restaurant recommended by name when people ask an AI assistant where to eat nearby, and a menu the owner can update from her phone.
- Stack
- Astro · Tailwind · Cloudflare Pages · Google Business Profile · JSON-LD
Some context on how this engagement ran, written the way I’d explain it to the next restaurant owner who calls.
The situation
The owner knew the site was dated but wasn’t sure it mattered. Her regulars kept coming. What she didn’t see was the dinner crowd that never found her: people searching on their phones, asking a voice assistant, or scanning Google Maps on the way home.
What I did
I started by sitting in the dining room on a Tuesday night with her and her manager, writing down every way a customer currently reaches them. Then I rebuilt the site around the three things people actually want: the menu, the hours, and a table.
Behind the scenes, I fixed the listings that search engines and AI assistants read from, added the structured data that lets them quote the menu and hours directly, and set up a tiny admin so the menu updates in a minute instead of a phone call to a nephew.
What changed
Within two months, reservations were up and the owner started hearing “I found you on my phone” from new tables. The site loads in under a second on a bad connection, and nothing on it requires a developer to change.