13 Cities, 105 Websites, One Build: The BellyUp! Case Study
BellyUp! is 13 city website families, 105 websites in all, and they went live in 63 days. Twelve of them started as a copy of a city that already worked. The useful part for anyone running many sites is what happened when a bug turned up in one. Every city below links to a live site you can open.
Context
Every platform looks good at site one. The real question for anyone who builds websites for other people is what happens at site twenty, and at site a hundred.
BellyUp! is a family of local discovery websites. Every city gets eight of them, a city home plus seven discovery sites. BellyUp! Jax, for example, has its city home plus Bars, BBQ, Family, Nightlife, Over 50, Vegan, and Sports & Game Day. Every city is built the same way, so a new one starts from a city that already works.
It is the second reference build on this system, next to the 160 domain portfolio case study. That one shows what runs every night. This one shows what happens when the same build is copied again and again with real places, real neighborhoods, and real mistakes.
Methodology
Atlanta went live on August 18, 2026 as the working prototype. It was finished and tested before anything was copied from it. Jacksonville, on September 18, was the first city built from it.
Each new city got its own colors, neighborhoods, restaurant and bar list, and local guides. All of them sit on the same system as the rest of the portfolio, so the nightly job that reads every page back covers them too.
When a fix was made in one city, it was carried to the others instead of staying where it was found. On October 16, all 81 sites that existed by then had a confirmed sitemap and Search Console access.
Findings
Three things stood out.
- Cities started arriving in batches. Three went live on September 19 alone, and two more on October 19.
- A bug found in one city was fixed in all of them. One pass carried four fixes to twelve live sites.
- The biggest city needed a fix of its own. Charleston launched with 347 places and a change that made very long restaurant lists load reliably.
When each BellyUp! city went live
Thirteen cities went live between August 18 and October 20, 2026. That is 63 days. A month passed between the first city and the second. After that, twelve cities went live in the 32 days from Jacksonville on September 18 to New York on October 20, and the last four arrived in six days.
| # | City site | Went live | Places listed |
|---|---|---|---|
| 1 | BellyUp! ATL | August 18 | 80 |
| 2 | BellyUp! Jax | September 18 | 93 |
| 3 | BellyUp! St. Augustine | September 19 | 74 |
| 4 | BellyUp! Orlando | September 19 | 112 |
| 5 | BellyUp! Tampa | September 19 | 175 |
| 6 | BellyUp! Charlotte | September 20 | 223 |
| 7 | BellyUp! St. Pete | October 6 | 189 |
| 8 | BellyUp! Savannah | October 6 | 237 |
| 9 | BellyUp! Charleston | October 7 | 366 |
| 10 | BellyUp! Birmingham | October 15 | 257 |
| 11 | BellyUp! Miami | October 19 | 89 |
| 12 | BellyUp! Austin | October 19 | 57 |
| 13 | BellyUp! NYC | October 20 | 29 |
Together the 13 cities list 1,981 places and 74 local creators across 105 websites. That is 104 city and discovery sites plus the national home, BellyUp! America.
The lists keep growing after launch. Atlanta opened with 31 verified places and lists 80 today. Charleston opened with 347 and lists 366.
What broke in BellyUp! and how one fix reached every site
Copying a site also copies its bugs. Two are worth telling, because both were quiet.
The forms that looked fine
While Jacksonville was being built, a warning showed up on the test server. The first city ran the same code, so it was checked live instead of assumed. Every page loaded normally. But the contact, submit, and claim forms rejected every message.
The site started the visitor's session too late, after the page had already begun to load. So the browser never got its cookie, and the form's security check failed every time. The fix was one small change, made on the live first city and in Jacksonville before it opened.
A status code would never have caught it. A real test message sent through the real form did.
The edit screen that erased fields
The second bug lived in the screen used to edit a place. Saving a venue quietly erased the fields that the edit form did not show. It was live on the Atlanta and Jacksonville sites.
The fix was made in the St. Augustine build, then copied back to all twelve live Atlanta and Jacksonville sites on September 19, before St. Augustine opened. Four shared fixes went out in that one pass.
That is the rule underneath the whole family. A fix is not done when it works in one city. It is done when every site that has the problem has the fix.
Charleston and the very long list
Charleston opened on October 7 with 347 places, 27 neighborhoods, and 16 guides, the biggest list so far. The launch came with a change so very long restaurant lists would load reliably. The bigger the city, the harder it leans on the build.
What an agency can copy from the BellyUp! build
Three habits, and none of them need 105 sites to pay off.
- Finish one site before you copy it. Atlanta ran as a working prototype first, so every city after it inherited something that already worked.
- Carry every fix back. The older sites got the newer fixes before the next city opened, so no city turned into the odd one out.
- Test the real result. A form that returns a normal page can still reject every message. Send one through.
With ten client sites the lesson is the same. The twentieth site should not need its own research. The multi-site operations routine writes out the nightly checks, and the cost article shows what 100 client sites cost on each kind of pricing. See what you get on this system, or read how to start an agency on it.
Visit the BellyUp! sites and check the work
Everything on this page can be checked by clicking. Start at BellyUp! America, which keeps the full timeline and the directory of every city. Then open any city, like BellyUp! Charleston or BellyUp! NYC, and one of its discovery sites, like Sports & Game Day in Jacksonville.
The honest limits of the BellyUp! case study
This page counts websites, places, and creators. It does not count readers. A live site says nothing about how many people use it, and this page makes no claim about audience, reach, or results for any restaurant, creator, or sponsor.
The newest cities are too new to judge. Miami and Austin opened on October 19 and New York on October 20, so there is almost no Google data for them yet. Even the oldest city is only a couple of months old.
A bot visit is not a reader. The server logs record crawlers and people, and even the visits that look like people include scrapers. So this page leaves visit counts out.
Software does not decide which places belong. A live website is not proof that a restaurant is still open, so every list still depends on someone checking the newest dated posts and local news. That work costs the same in city thirteen as in city one.
And the cities are not identical. Same build does not mean same content. Each one has a different list, different neighborhoods, and its own local people.