Every category-defining live-data product I can think of started in a single city. Uber in San Francisco. Yelp in San Francisco. DoorDash in Palo Alto. Resy in New York. Foursquare in New York. Glossier in New York. The reason isn't accident — real-time, network-effect products need dense early data to be useful at all. A live nightlife map of 12 disconnected metros at launch is worse than a live nightlife map of one good city, because in the 12-metro version, every individual city is too thin to be reliable.
San Diego is the city. Here's why.
The right size, the right density, the right demo.
San Diego has roughly 1.4 million people in the city proper, 3.3 million in the county, and a 21-35 nightlife-active demo that's both large enough to power a city-wide dataset and small enough to be navigable as one social graph. The nightlife corridor between Pacific Beach and Gaslamp Quarter is one of the densest concentrations of bars, lounges, and clubs in the country per square mile. North Park, Hillcrest, and Little Italy add a craft / cocktail / wine-bar layer on top. La Jolla and Coronado add the upscale and date-night surface. Carlsbad, Del Mar, and Encinitas extend the map north along the coast.
The full set of 19 neighborhoods we cover on day one:
These aren't aspirational. Every one of them has live polygons in our backend with venue coverage curated by hand — the map you see at launch is the map we've been mapping for over a year. The neighborhood count matters because density inside each neighborhood is what makes the live signal robust. If we declared "we cover 40 neighborhoods" but most had three venues each, the energy levels would be useless. 19 neighborhoods with deep venue coverage in each is the version that actually works.
One city right, then the next.
Our expansion order is fixed and short-term: Orange County, then Los Angeles, then the major US nightlife metros — New York, Miami, Austin, Nashville, Chicago, Atlanta, Las Vegas — over the next five years. Each new city only opens once the live dataset in the previous city is mature enough that a user moving from one to the other gets the same quality of signal.
The temptation to launch in five cities at once is real. Investors love it. PR loves it. Users hate it — because thin live data in five places is worse than thick live data in one. So we resist.
The dataset is the moat. The dataset grows one city at a time.
If you live in SD.
You're the early-data flywheel. The first 1,500 users on the waitlist get Pro free for six months at launch (summer 2026). If you've got friends in PB, Gaslamp, North Park, Hillcrest, La Jolla, Little Italy, Mission Beach, OB, East Village, South Park, Bird Rock, Point Loma, Uptown, Old Town, Mission Valley, Coronado, Del Mar, Encinitas, or Carlsbad — share the link. Every friend who joins moves you ahead in the launch queue.
We're getting one city right. This one.
Sources + further reading.
- All 19 SD neighborhoods — the hand-finished landing pages for every hood referenced in this post.
- Glossary — Neighborhood polygon — how we define each hood's boundary.
- About Jellyfish — the founder origin story and the patience-as-strategy framing.