I pulled every independent business inside a 2500 metre circle around the Quad, dropped anything carrying a chain brand, and then tried to open each one's website. There were 121 of them. Twenty-seven had a site that answered.
What did I measure?
The source is the OpenStreetMap website tag, which volunteers maintain, and a business with no website recorded there might still have one. The data cannot show whether ninety-four businesses near campus have a website.
It does show that for ninety-four of them, a machine looking for a website finds none. Ninety list none at all. Three list one that no longer answers. One points at a page its owner deleted.
A student with a phone runs into the same gap. Someone searching for a business wants its hours and phone number, and the search comes back empty.
By category
| Category | Found | Reachable site |
|---|---|---|
| Restaurants | 64 | 15 |
| Clothing | 12 | 1 |
| Barbers and salons | 11 | 3 |
| Cafes | 8 | 2 |
| Gift shops | 4 | 2 |
| Bakeries | 3 | 1 |
| Bookshops | 3 | 1 |
| Car repair | 3 | 0 |
| Phone repair | 3 | 0 |
| Bike shops | 2 | 2 |
| Ice cream | 2 | 0 |
| Laundry | 2 | 0 |
| Florists | 2 | 0 |
| Jewellers | 1 | 0 |
| Tailors | 1 | 0 |
Restaurants do best, and they still clear only one in four. Every laundry, florist, jeweller, tailor, ice cream shop, phone repair place and car repair shop in the circle came back with nothing.
The number that surprised me
Twenty-three of the 121 have opening hours recorded anywhere a machine can read them.
For eight in ten businesses within walking distance of campus, only someone standing at the door can tell whether they are open right now.
Where does a website not help?
I build websites, so you could read this as a pitch for my own service. Two things cut against that reading.
Some of these businesses do not need a website. A barber shop with a line out the door every Saturday gets enough customers without one, and a restaurant that fills every night on word of mouth already has the customers a website would bring.
For many of the rest, the cheapest fix is to claim the Google listing and enter the hours, which is free and takes twenty minutes.
What happens next
I am visiting these places and writing about them one at a time. For each one I write who runs it, how long it has been here, what it does well, and what a student trying to find it runs into. The series is called the Storefront Project. I feature businesses whether or not they are clients.
If you run something near campus, I would like twenty minutes at your place. You get a written piece about your business either way, and you owe me nothing. Reach me at (205) 462-7839 or elijah@purcellventures.co.
Method
I ran one Overpass query against OpenStreetMap for shops and food businesses within 2500 metres of the Quad, on 20 August 2026. I dropped anything tagged with a chain brand. I sent an HTTP request to every listed website and counted a 403 or a 405 as working, because plenty of live sites refuse a bare HEAD. The script is scout.py and re-running it reproduces every figure above.
Correction, same day. The first version of this piece said twenty-six working sites and ninety-five without. The script was sending only a HEAD request, and two restaurants whose sites are perfectly alive answer GET but refuse HEAD, so the script counted them as broken. I corrected the figures above to twenty-seven and ninety-four, and I left this note up so readers can see what changed.
Part of the guide AI for small business.