A few years ago, cities across the country were signing deals to blanket themselves in free or cheap WiFi, and the pitch was straightforward. String up enough access points, subsidize the network through advertising or a small subscription fee, and every resident gets internet access regardless of whether they can afford a cable bill. Several of the biggest projects are now quietly falling apart, and the reason has almost nothing to do with the radio technology.
The company doing much of this building has started pulling out of city after city, citing a business model that never generated the revenue anyone modeled when the contracts were signed.
The physics turned out to be harder than the sales pitch admitted. Blanketing an entire city with outdoor WiFi at a signal strength that reliably reaches inside people’s homes requires far more access points per block than the original proposals assumed, because WiFi was designed for a room, not a neighborhood, and every wall between a router and a laptop cuts the range further. Cities kept discovering they needed two or three times the hardware density originally budgeted, and each additional access point needs power, backhaul, and eventual maintenance.
Then there is the advertising model underneath most of these deals, which assumed a volume of users and engagement that never materialised, because most people who already have a broadband connection at home saw little reason to switch to an outdoor network with worse reliability, and the population the project was actually meant to help, people without an existing connection, could not always afford the laptop needed to use it in the first place.
I do not think the underlying goal was wrong. Getting internet access to people who cannot currently afford it is a real problem worth solving, and I would like some version of this to work eventually.
What I think was wrong is treating it as a network engineering problem when it was mostly an economics problem, and building the infrastructure before anyone had proven the money actually worked at city scale. A handful of smaller, targeted deployments- a park, a downtown business district, a specific low-income neighborhood- have done fine, because the density and the funding matched something achievable. The city-wide version kept getting sold as the same idea at ten times the scale, and scale is exactly where this stopped adding up.