In the Connectivity Explorer a Census block is considered well-served based on a minimum level of service (e.g. 100Mbs↓/20Mbs↑) and a minimum number of service providers. This definition should be extended to actual service-level data (e.g. “x% of homes have verified service at”) and pricing data (e.g. “paying less that $35/month”) as soon as reliable data comes available!
The concepts of under-served and un-served are defined relative to well-served; so they too require a target speed, the service provided, and the the number of providers who offer service at that speed.
For example: “A residential block is well-served at 100↓/20↑ if it has a minimum of 3 providers”.
You can then define “un-served” (e.g. no providers at 100↓/20↑) or “under-served” (e.g. 1 or 2 providers if your target is 3) relative to your goals. This approach easily extends to other measures such as population or households. The settings can be changed for each Notebook. The Connectivity Explorer presents this data both for the notebooks locations and as a state-wide measure, using the active notebooks target speed and number of providers.
In this formulation, you can focus on the issues of "what speeds to we really need to be well-served?” and “how many providers make for a healthy market?” A single provider is sufficient if you’re planning/measuring a municipal or regulated network, but not acceptable if you’re relying on competition to drive your economic model.
The FCC’s current (8/2020) stance would be that “A well-served block is one that has access to 25↓/3↑ (or above) with only a single provider.” Jon Sallet says it should be 100↓/50↑. Most economists would prefer to see at least 3 or 4 providers, unless you’re discussiong a regulated or municipal model.
FCC:
- 25Mbs↓/3Mbs↑ is well-served? Really?
- 1 provider is well-served? Really?