This article is excerpted from “Co-ops Connect FYI,” September 30, 2022

What do a Mexican restaurant, a Baptist Church, an orthodontist’s office and a school bus barn have in common? I’ll take suggestions for a good punchline. Unfortunately, all I’ve got is this: when Conexon submitted these establishments as broadband serviceable locations (BSL) in rural Georgia, they were rejected.

Across the country, restaurants, car dealers, shops, grocery stores, pharmacies, independent schools, rural libraries and countless small businesses were rejected. They are not part of the location fabric, defined as locations that could receive mass-market broadband.

The bottom line: These hundreds of thousands, perhaps millions, of locations will not appear on the FCC’s new broadband maps. According to the FCC, “The locations in the fabric that have a BSL Flag of ‘False’ are structures that have or should have broadband service but likely do not take or would not take mass-market service (and therefore do not fall within the definition of a BSL) based on available data. Examples of such locations include certain Community Anchor Institutions and large enterprises.”

Understanding Cost Models

About a decade ago, when FCC staff were putting together cost models for building fiber networks, a determination was made not to include large businesses in the calculations.

Though fiber would be built for large businesses, the FCC assumed that such firms could pay for dedicated access and thus would not need to be part of the modeling process.

Schools, libraries and health care institutions had their own funding programs, so these too were excluded. Thus, only housing units and small businesses were part of the Connect America Cost Models.

Lack of SMB Visibility

The FCC’s consultant for the cost model was CostQuest. I know the CostQuest team. It is a first-rate, technically proficient group. CostQuest did not possess a database of the locations of all small businesses in the country. It licensed such a database, which turned out to be inaccurate and distorted the Connect America Fund and Rural Digital Opportunity Fund programs. But that’s another topic for another time.

For the Broadband Location Fabric, the FCC used CostQuest to map the locations for the new broadband maps. Once again, the decision was made to exclude large enterprises and community anchor institutions.

Once again, CostQuest, not possessing such a database, licensed one. As I understand the methodology, businesses with a certain number of employees are treated as large enterprises, and the location fabric excludes such businesses from the broadband map.

But there is a flaw in how the data is used. The data set aggregates all employees at every location and counts individual locations as one large enterprise. Effectively, each restaurant in a chain is treated as if every employee works in every restaurant. Each branch in a bank is treated as if every employee works in every branch.

You get the idea. Every small part of a chain is treated as if it were a large enterprise. The result is these locations are excluded from the location fabric.

Go Deeper

It makes sense to exclude large enterprises from the calculations of rural broadband, right? And to follow the same approach with the new location-based broadband map, right? Wrong.

Hundreds of thousands – maybe millions – of data points were wrong. According to the FCC, “[Though] these locations may subscribe to mass-market services in some instances, we currently lack data that would enable us to make reliable determinations of when this is the case.”

Such an approach is precisely wrong for businesses in the rural U.S. To provide a concrete example, I will share results from queries into the location fabric. As with the cost models, the public inexplicably does not have access to this data.

Example: Baxley, Georgia

Baxley is the county seat in rural Appling County, with a population of 4,942 according to the 2020 census:

  • We found 54 locations in Baxley that are not counted under the FCC’s approach to mapping.
  • Retail shops, restaurants, independent schools, churches and small business offices would likely take a mass market service, especially the gigabit and
    multi-gigabit services fiber optic networks offer.
  • The broad assumption by the FCC that every business establishment in the rural U.S. would buy dedicated internet access is passing strange.

The designation “FALSE” in Table 1 means the location is not a broadband serviceable location under the Infrastructure Act. Congress punted the definition of the BSL to the FCC. Even though Congress deferred to the FCC, I’d be surprised if any member of Congress meant that a Baptist church or a small store or restaurant in a rural area should not be counted. (Are any rural members of Congress paying attention?)

Table 1. The designation “FALSE” in this table means the location is not a broadband serviceable location under the Infrastructure Act.

Why it Matters

Does it matter whether the FCC misses 40 or 50 locations in every rural town and community on the broadband map?

Suppose one’s focus for mapping is the 140 million or so locations, mostly in urban and dense suburban areas. Probably not. But since the first use of this map is the determination of unserved locations, mostly in rural areas, it certainly matters.

Conexon’s initial analysis found that the location fabric misses more than 10 percent of rural locations on farms and in trailer parks and new subdivisions.

The Final Word

Excluding nearly every small business in the rural U.S. is a big deal. Suppose an entity counting locations in the rural U.S. is undercounting farms and trailer homes and excluding small businesses. If that happens, the most underserved segments of the most underserved parts of the country are missed.

Such an approach distorts the picture of broadband availability in the rural U.S. and will distort the funding allocation.

When Conexon asked whether the FCC would correct its errors, it was told to submit challenges. The FCC has neither the time nor the capacity to adjudicate hundreds of thousands of challenges. The flaws are systemic and should not depend on the challenge process.

Sight unseen, the National Telecommunications and Information Administration has already decided that it cannot use the first iteration of the FCC’s maps for the Broadband, Equity, Access, and Deployment (BEAD) Program allocation. There is still time to correct this fundamental flaw in the FCC’s approach to rural small businesses and anchor institutions.


Jonathan Chambers is a partner at Conexon, a rural broadband leader working with electric cooperatives to deliver fiber-to-the-home internet solutions to rural communities across the country.

Jonathan Chambers