Sizing Up Another BEAD Round Using the Ever-Changing BEAD Map
Nearly five years after Congress created it, BEAD is still very much a work in progress.
Key takeaways
- The number of locations still eligible for BEAD has fallen 69% since funding was initially allocated, extending the 65% decline we reported in July 2025. None of this is BEAD’s doing as ISPs have not sat idly by waiting for funding from this program.
- Looking only at the locations that needed BEAD in December 2022, 79% are now served, are no longer serviceable locations, or are covered by a federal program other than BEAD.
- Round One’s projects have been overtaken by this ceaseless buildout. 69% of projects receiving funds via BEAD Round One now have more than a fifth of their locations receiving 100/20 service, and 89% have more than a fifth at 25/3.
- Looking ahead to BEAD Round Two, we estimate that 1,041,099 locations will appear on NTIA’s supplemental lists.
- Further, we estimate that the amount of supplemental funding that will likely be made available by NTIA to serve these locations could be anywhere from $5.3B to $8.8B. An exact figure depends on several variables, notably whether provider matching funds count toward NTIA’s “average cost per location.” Dipping into remaining BEAD will reduce BEAD “savings” by 25% to 42%, leaving $12.2B to $15.7B unspent.
Another BEAD Round
BEAD was designed to close the digital divide in the U.S., but the size and shape of that divide have changed dramatically since Congress created the program in November 2021. Private investment and other broadband programs have continued to connect locations while BEAD, the largest ever broadband grant program, has followed a circuitous path across two very different presidential administrations to finally get to the point of being able to dole out funds. As a result, far fewer locations remain unserved or underserved today, and those that remain tend to be harder and more expensive to reach.
In short, the contours of the problem that BEAD was designed to solve have changed substantially, and in many ways the mission – to close the digital divide – has become more difficult. The ACLP has been doggedly tracking these myriad changes and challenges over the last few years and has recommended, among other things, that NTIA use unspent BEAD funds to make sure no locations fall through the cracks.
NTIA is finally catching up to the reality of an ever-changing BEAD map and has acknowledged that the data demand a course correction.
NTIA’s Supplemental Deployment Policy Notice, released on September 3, announced a new BEAD round aimed at addressing these issues. This analysis looks at three related questions:
- How much of the original BEAD challenge remains?
- How much of it is already covered by existing awards or other federal programs?
- How many locations—and how much funding—are likely to be involved in this new round?
The answers illustrate both how much progress has occurred outside BEAD and how complicated the final stretch of closing the digital divide may be.
A location needs BEAD if it is a broadband-serviceable location listed in the FCC Fabric without reported adequate service at it. Service is adequate when it advertises at least 100 Mbps down and 20 Mbps up, is classified as low latency, and runs over copper, cable, fiber, licensed or licensed-by-rule fixed wireless.
Two variables in that test are treated differently across published work, so we provide a range of estimates based on both.
- Unlicensed fixed wireless. This technology has been deemed both reliable and unreliable over the course of BEAD. The Biden NTIA classified it as unreliable and excluded it from BEAD. The Trump NTIA reclassified it as reliable as part of Benefit of the Bargain (BoTB) but only if certain criteria were met. We report both treatments.
- Which side of the market an offer must reach. The FCC’s own unserved-and-unfunded file counts a location as “served” if either a residential or a business offer reaches it. Other work has used a residential-only test, or matched the offer to the building type recorded in the Fabric.
Where a single figure is needed, we use the FCC’s own convention: the qualifying technologies, and either side of the market.
Our calculations are based on state Final Proposal data aggregated by the ACLP, and published FCC data: the Broadband Data Collection, the Broadband Serviceable Location Fabric, and the Broadband Funding Map. Some states’ Final Proposal data is still a draft or submitted version rather than an NTIA-approved one, so totals need not match NTIA’s public announcements of Final Proposal approval. NTIA generates the Supplemental BEAD Eligible Location Lists from inputs we do not have. Everything here is an estimate of what those lists will contain, not the lists themselves.
NTIA has said that over 5,000 additional locations could become eligible in Alaska. Our estimate for Alaska runs from 6,419 to 7,605. NTIA’s figure is a floor rather than a count, and our estimate precedes the abbreviated challenge process, which can only remove locations from the list.
Where things stand
From every analytical vantage, the problem BEAD was established to solve has shrunk sharply.
| Unlicensed fixed wireless | Market test | Dec 2022 | Unserved today | Still BEAD-eligible today | Decline in eligible |
|---|---|---|---|---|---|
| Excluded | Any market (FCC) | 12,104,450 | 5,120,087 | 3,266,502 | 73% |
| Excluded | Residential only | 13,458,918 | 6,212,309 | 4,255,947 | 68% |
| Excluded | Matched to structure | 12,891,335 | 5,241,985 | 3,356,189 | 74% |
| Included | Any market | 10,600,688 | 4,370,104 | 2,820,835 | 73% |
| Included | Residential only (prior report) | 11,886,942 | 5,326,181 | 3,682,974 | 69% |
| Included | Matched to structure | 11,326,704 | 4,482,055 | 2,903,025 | 74% |
Using the methodology from our July 2025 report regarding the shrinking digital divide, the count of eligible locations has fallen from the 11,939,924 to 3,682,974, a decline of 69%, compared with the 65% we reported then.
Under each counting method shown above, the number of BEAD-eligible locations has declined between an estimated 68% and 74%. “Eligible” means a location is unserved or underserved, not covered by another federal program’s enforceable commitment, and was not removed by the state in its Final Proposal.
NTIA allocated BEAD on unserved locations alone – i.e., those that lacked internet service of at least 25/3. The same data shows there were 8,518,832 unserved locations in December 2022 compared to 3,199,376 today.1 BEAD funds unserved and underserved locations alike, which is the larger count the rest of this piece uses.
Of the locations that needed BEAD when funds were allocated, 54% are now served and a further 13% turned out not to be serviceable locations at all. Counting locations since covered by a federal program other than BEAD, 79% of that original cohort has been resolved or funded without BEAD. Another 16% sit inside a BEAD award.
Whichever way one approaches these figures, the problem points in the same direction: the universe BEAD was created to address has shrunk dramatically. However, the number of locations that still lack adequate service is not the same as the number still needing a funding solution.
What is already funded
Needing service and being unfunded are different things, and the locations that still need service divide across three funding buckets.
Of the 5,120,087 locations that need service today, BEAD is funded to reach 2,225,650. Of the 3,787,410 locations identified as receiving BEAD funding in states’ Final Proposal data, 1,426,088 (roughly 38%) already have service, and a further 135,672 are no longer locations at all.
Other federal programs cover 1,672,591, led by Rural Digital Opportunity Fund at 809,187 locations.
That leaves 1,221,846 locations with no current funded path to service, roughly 24% of the locations that need service today.
The changing map creates a second issue as well. Private buildout has not only reduced the number of locations left for a supplemental round – it has also changed the composition of many projects BEAD has already selected.
Round One projects and the 80 percent rule
The IIJA, which created BEAD, defines an unserved service project as “a project in which not less than 80 percent of broadband-serviceable locations served by the project are unserved locations,” and an underserved service project as “a project in which not less than 80 percent of broadband-serviceable locations served by the project are unserved locations or underserved locations.” In practice, this means that upwards of 20% of a BEAD project area can already be served, thus allowing for some measure of overbuilding. To date, it does not appear that anyone has checked to see if this 20% served rule has been followed or if certain project areas now run afoul of it given the significant changes to the BEAD map documented here. With BEAD Round Two upcoming, we take a closer look as part of this analysis.
Of the 5,728 projects that received $18.4B in funding via BEAD Round One:
| Share of locations now served | Projects at 100/20 | Support | Projects at 25/3 | Support |
|---|---|---|---|---|
| More than 20% | 3,976 (69%) | $13.7B | 5,102 (89%) | $17.4B |
| More than 50% | 2,116 (37%) | $5.7B | 3,984 (70%) | $13.0B |
| More than 80% | 959 (17%) | $1.1B | 2,008 (35%) | $4.3B |
This means that 89% of BEAD projects are not compliant with the 20% served rule for an unserved service project, and 69% are not compliant with the rule for an underserved service project, because too many of their funded locations are now served. Moreover, what is left in these projects are often the more difficult and expensive locations to serve. BEAD award locations that are still unserved sit in areas 52% less dense than award locations somebody else has since served, 30 nearby locations against 62.
This may result in project descoping or contract revisions to comply with the statute and could add to the time, costs, and complexity associated with wrangling the ever-changing BEAD map to pinpoint remaining unserved locations.
Against that backdrop, the practical question becomes: what will be left for NTIA’s new supplemental round?
The supplemental list
NTIA will build each state’s list of locations eligible for supplemental deployment funds by a sequence of exclusions detailed in the Notice and applied to the current FCC data. On the FCC’s own convention:
NTIA will hand each state a Supplemental BEAD Eligible Location List built on Fabric v8, excluding locations that have since fallen out of Fabric v9; 76,849 of the locations that need BEAD today are no longer locations in v9.
The list also excludes everything the approved Final Proposal already addressed, which removes 2,205,798 BEAD-awarded locations and 1,651,324 covered by another federal program.
States removed a further 93,667 at Final Proposal under reason codes 1 through 3 and 5 through 7.
NTIA leaves state and local enforceable commitments off its own list and requires states to apply them, which would take out another 51,350.
Based on the latest data available to us, and likely less current data than what is available to NTIA, we estimate 1,041,099 locations.
This may be an underestimate: 1,129,257 new locations appear in Fabric v9 and not v8, and none can be tested for service because the required BDC data (June 2026) is not yet public. Removals by default, descope or rescission will likely further increase this total.
Satellite, ULFW, and the challenge process
A challenger can strike a location off the list with evidence of service on copper, cable, fiber, licensed or licensed-by-rule fixed wireless, unlicensed fixed wireless, and non-geostationary (i.e., LEO) satellite.
Non-geostationary satellite is reported as covering 99.7% of broadband-serviceable locations in the country, and its treatment as not-qualifying but valid-bidder throughout the BEAD process has been a source of continued ambiguity. Per FCC Data, LEO offers are available to 1,038,504 locations, or 99.8% of those on the supplemental list.
Unlicensed fixed wireless is available at 155,826 locations, 15.0% of the list.
Consequently, 99.8% of locations on the list could be challenged by either LEO or ULFW, leaving only 2,561 locations with no challengeable technology of any kind according to the latest FCC data. Presence in the FCC map is not on its own enough to remove a location for LEO and ULFW ISPs, and providers using those technologies may be required to show they have the capacity and ability to actually serve locations they challenge.
What the list looks like
The locations left on the list differ from the ones BEAD has already funded, both in what they are and where they’re located. Of the remaining locations to potentially be served by BEAD Round Two:
815,039 are homes and 220,497 are businesses, and a further 5,563 are mixed-use buildings the Fabric records as both.
489,985 were already waiting when BEAD was allocated, 200,748 were reported as having qualifying service in December 2022 and have since lost it, and 350,366 were not in the Fabric at all back then.
They are in areas 84% less dense than locations that already have service, but, curiously, 15% denser than the locations BEAD funded in its first round.
That leaves one final question: what will it cost to address this much smaller—and increasingly difficult-to-serve—universe?
What it will cost
NTIA will provide each state an upper limit of funding, set by its own average cost per location under its approved Final Proposal, multiplied by the number of locations on its list.
The notice is unclear whether this “average cost per location” includes only BEAD funding, or also includes matching funds from ISPs. As such, we present estimates using both metrics.
Those rates run from $1,466 to $29,314 per location with a median of $8,339, or from $1,099 to $21,986 with a median of $4,919 if provider matches are excluded.
Applied state by state, the upper limit comes to between $5.3B and $8.8B, depending on whether provider match is included in the average cost per location.2
This translates to using 25% to 42% of the $21B in Benefit of the Bargain “savings,” leaving anywhere from $12.2B to $15.7B in the “savings” bucket after the Second Round.
These calculations reflect dynamics evident in BEAD’s First Round, where locations are sparser than these on the same density measure. Going above the upper limit set by NTIA would require a state to demonstrate extraordinary circumstances.
What’s next
The Supplemental Notice addresses one of two recommendations the ACLP made in October 2025. We proposed a Reserve Fund seeded with “savings” realized through the Benefit of the Bargain round, to make certain the digital divide actually closes. The latest Notice does this in part by running a second round for locations that were excluded from BEAD for reasons that had nothing to do with whether they needed service. The Supplemental Notice does not address our second recommendation, which was to keep funding available for defaults and for costs that have risen since the applications were priced.
That gap matters because the money is sitting there. On our estimate, the second round consumes 25% to 42% of the Benefit of the Bargain “savings,” leaving $12.2B to $15.7B unallocated. Meanwhile, BEAD subgrantees are under real cost pressure, from the changes NTIA made to the program, from rising labor and material costs, and from delays and expense at utility poles and in permitting. Defaults are a question of when rather than whether and are most likely among the long tail of smaller providers holding BEAD awards. The notice says nothing about how they will be covered. In addition, a minority of states appear to have signed contracts with subgrantees and there has been considerable descoping of locations in those states. With most states still working to complete their Round One contracts, NTIA’s policy notice is ambiguous vis-à-vis what will happen to those locations.
The figures here will also move, in both directions and over a long clock. The abbreviated challenge process can only remove locations from the lists. The next release of the FCC’s maps will add them, and it will add locations that cannot be tested today because the required data is not yet published.
Taken together, this analysis highlights a sobering reality: Nearly five years after the Infrastructure Investment and Jobs Act created it, BEAD is still very much a work in progress.
By state
Each state’s supplemental list and its funding cap.
| State | Supplemental locations | Funding cap | Funding cap, excl. match |
|---|---|---|---|
| AK | 7,206 | $108.8M | $97.6M |
| AL | 33,616 | $320.3M | $167.6M |
| AR | 21,272 | $149.1M | $83.5M |
| AS | 1,929 | $2.8M | $2.1M |
| AZ | 30,676 | $184.6M | $85.3M |
| CA | 33,193 | $280.5M | $174.4M |
| CO | 21,933 | $145.6M | $96.0M |
| CT | 1,461 | $20.0M | $14.7M |
| DC | 13 | $0.4M | $0.3M |
| DE | 1,404 | $26.2M | $4.0M |
| FL | 43,144 | $277.7M | $134.0M |
| GA | 17,668 | $122.6M | $45.7M |
| GU | 97 | $0.4M | $0.1M |
| HI | 1,089 | $6.4M | $4.8M |
| IA | 13,225 | $101.5M | $64.2M |
| ID | 14,567 | $100.2M | $68.0M |
| IL | 26,430 | $256.5M | $158.8M |
| IN | 28,294 | $200.0M | $111.3M |
| KS | 8,872 | $76.0M | $55.6M |
| KY | 35,320 | $245.3M | $148.8M |
| LA | 25,415 | $176.7M | $102.0M |
| MA | 2,725 | $30.0M | $20.0M |
| MD | 6,903 | $121.1M | $63.1M |
| ME | 9,329 | $67.5M | $20.7M |
| MI | 34,542 | $257.5M | $160.1M |
| MN | 13,982 | $106.4M | $70.9M |
| MO | 15,439 | $88.0M | $57.2M |
| MP | 880 | $4.9M | $2.9M |
| MS | 5,384 | $47.9M | $29.3M |
| MT | 13,143 | $86.4M | $58.4M |
| NC | 69,141 | $568.8M | $303.3M |
| ND | 551 | $16.2M | $12.1M |
| NE | 7,477 | $35.1M | $23.7M |
| NH | 3,087 | $26.3M | $11.2M |
| NJ | 3,631 | $70.1M | $20.5M |
| NM | 15,787 | $198.4M | $141.6M |
| NV | 5,572 | $50.1M | $36.3M |
| NY | 21,204 | $217.0M | $114.9M |
| OH | 49,234 | $268.5M | $152.8M |
| OK | 21,665 | $325.8M | $242.8M |
| OR | 12,896 | $116.1M | $72.4M |
| PA | 23,592 | $279.8M | $147.0M |
| PR | 2 | — | — |
| RI | 1,466 | $14.5M | $9.0M |
| SC | 8,418 | $33.3M | $17.3M |
| SD | 3,179 | $41.5M | $33.1M |
| TN | 13,654 | $125.2M | $62.8M |
| TX | 120,707 | $931.4M | $632.2M |
| UT | 7,453 | $81.4M | $51.2M |
| VA | 91,613 | $1,054.8M | $590.0M |
| VT | 4,609 | $49.0M | $28.0M |
| WA | 39,561 | $242.3M | $174.9M |
| WI | 18,189 | $113.7M | $72.4M |
| WV | 21,062 | $232.0M | $161.2M |
| WY | 8,198 | $57.3M | $42.2M |
| Total | 1,041,099 | $8,759.8M | $5,254.7M |