Why performance testing is so critical for BEAD-funded LEO projects: What states and localities need to know
By Ziggy Rivkin-Fish, V.P. Broadband Strategy, CTC Technology & Energy
February 23, 2026
The BEAD Final Proposals are all in—and a significant share of states’ BEAD funding has been awarded to low-Earth orbit (LEO) providers. Given how many households will be served by those awards, the LEO testing guidance released by NTIA in its February 2 update to the BEAD Frequently Asked Questions (FAQ) is not just a minor clarification—it directly affects how public funds and the interests of rural households will be protected by state broadband offices.
Performance testing is one of the few tools available to validate LEO service obligations, so states need to understand what has changed in the guidance, what remains non-negotiable, and where contradictions in NTIA’s guidance may create challenges during deployment and closeout.
This new LEO testing guidance matters for local governments as well; localities that understand the testing requirements can help gather customer complaints and performance data that state broadband offices will need to support enforcement actions.
Testing is the key to confirming LEO service commitments
LEO Capacity Subgrant projects are fundamentally different from terrestrial broadband projects. To monitor terrestrial construction, a state’s inspectors can examine physical assets—fiber and cabinet installations and the like—and inspect as-built maps and photographic evidence. But for LEO deployments, there is no infrastructure the state can inspect.
Unlike for terrestrial projects, therefore, performance testing is a state’s only objective means to verify claims of completed milestones for a LEO project. Without testing, the state bears the risk of accepting incomplete work.
A robust testing framework for LEO networks becomes even more important during the extended period of performance NTIA has required for LEO subgrants. Neither the state nor subscribers hold any real visibility into whether LEO networks are maintaining sufficient capacity as the years pass. Similarly, the states have no way to verify that LEO subgrantees’ vast satellite constellations are functioning properly, or even to ensure that customer premises equipment remains correctly placed or angled to secure BEAD-compliant service.
The risks to continued BEAD‑compliant LEO service are significant precisely because there’s so little transparency into how a LEO network is performing at any given location. For example, if LEO providers have to change the orbital altitudes of their satellites to avoid congestion or debris, those shifts could affect the speed, capacity, and coverage available to BEAD-funded locations.
And because launching replacement satellites takes significant time and resources, states may face long delays before major service issues, once identified, can be fixed.
First things first: NTIA’s new guidance confirms LEO providers’ service obligations—so states know what they need to test
With the February 2 release of the FAQ update, NTIA has clarified that the BEAD performance requirements established in the BEAD NOFO, the BEAD Restructuring Policy Notice (RPN), and governing documents such as NTIA’s Performance Measures for BEAD Last-Mile Networks (Performance Measures) cannot be changed.
This guidance appears to be a direct response to Starlink, which tried to sidestep some of NTIA’s BEAD rules by proposing revised contract terms to the states that would redefine when service is considered “complete” and when a test should count as a failure.
NTIA’s rejection of Starlink’s novel terms protects the states’ interests, because a primary value of BEAD-funded LEO deployments lies in the service commitments that NTIA requires the LEO providers to meet—including 100/20 Mbps (or greater) service, free customer premises equipment, and access to a lower-cost plan for eligible households.
More good news: NTIA presents clearer rules on what counts as “completed” service
In a win for accountability, the February 2 NTIA FAQ update states that LEO providers cannot avoid responsibility for their service commitments by redefining when a location is “completed” or ready for service.
Starlink’s proposed term revision would have treated a location as complete the moment the customer premises equipment was mailed by Starlink to the customer—rather than when it was installed—and would have invalidated performance tests if a customer did not properly install or calibrate the customer premises equipment.
The FAQs confirm that LEO subgrantees such as Starlink must still ensure end-to-end connectivity and cannot shift the burden of ensuring proper installation onto subscribers.
Likewise, performance validation must reflect the actual activated connection, regardless of who installs the customer premises equipment. Providers cannot simply mail equipment to a customer, declare the job done, and blame “user error” for poor test results.
Here is what the FAQs now say:
“NTIA emphasizes that no BEAD statutory requirements or other program rules may be altered by a BEAD subgrant agreement. BEAD subgrantees have an obligation to provide at a minimum ‘qualifying broadband’ to all locations covered by the BEAD project with (i) a speed of not less than 100 Mbps for downloads; and (ii) a speed of not less than 20 Mbps for uploads; and (iii) round-trip latency less than or equal to 100 milliseconds. Furthermore, the obligation to provide qualifying broadband cannot be made contingent on successful installation by subscribers or requiring subscribers to make modifications to their own or surrounding property.” (FAQ 9.4)
This answer reinforces that testing remains central to documenting BEAD-funded LEO compliance. However, as we describe next, NTIA’s FAQs raise new questions around when and how performance testing applies to LEO projects.
Where things get murky: What “testing” means and what standards states can apply
The FAQs rely on NTIA’s Performance Measures guidance, which sets minimum performance standards. The FAQs tell states they cannot impose higher standards than those outlined in Performance Measures:
“…the test requirements and thresholds detailed in the NOFO and the Policy Notice may not be changed. The requirements that may not be changed by the Eligible Entity include the sample size, number of sample sets, and the speed, latency and availability requirements.” (FAQ 11.13)
But this restriction does not align with how states should administer BEAD projects to protect the public funds. The FAQs do not explicitly differentiate annual performance testing (required for the states’ semiannual reports to NTIA) with the project-specific testing that states will require to document successful completion of milestones or closeout—completion sufficient to justify payment of public funds to the LEO subgrant awardee:
- Annual subgrantee testing: States will conduct testing to measure a subgrantee’s network performance for the purposes of annual reporting to NTIA . The testing is conducted across all of a subgrantee’s projects in that state. The tests lead to a pass-or-fail result and can trigger remedial actions. These tests are subject to the strict constraints referred to in FAQ 11.13 and documented in Performance Measures on how testing should be conducted and with which standards.
- Project milestone testing: States will conduct testing to verify whether a specific project has met a set of project milestones for the purposes of reimbursement or closeout compliance. This testing applies at the individual project level in accordance with 2 CFR 200. For LEO, this type of testing will occur throughout the 10-year extended period of performance.
States define their own project milestone-based testing requirements and are allowed to impose higher standards than for annual subgrantee testing, including specifying the sampling area and increasing the sample size to ensure the provider is delivering on all its commitments. These tests are not subject to NTIA standards or reviews. And NTIA specifically calls out that states are allowed to impose alternative testing requirements for such purposes:
“An Eligible Entity may elect to include in subgrant agreements additional testing and reporting requirements to monitor construction progress, such as construction validation or acceptance testing. The results of these tests are not required by NTIA and, therefore, will not be collected or reviewed by NTIA.” (FAQ 11.2)
More confusion about testing: A shift in NTIA’s language about the deployment period
The February 2 FAQs advise states that obligations in the Performance Measures do not apply past the deployment period, including for LEO projects and their 10-year extended period of performance following the deployment period:
“Further, these obligations only apply during the deployment period of performance (which does not include the extended period of performance for LEO subgrantees).” (FAQ 11.13)
But this answer contradicts other guidance. Closeout for LEO is defined as occurring during the extended period of performance—when the final payment is issued. And the FAQs also state:
“Subgrantees must submit their final performance measures results report by the due date for the final report prior to grant close-out and final payment.” (FAQ 11.1)
This creates a tension: LEO service obligations (and the states’ obligation to monitor LEO subgrantees’ performance) continue past deployment—but the FAQs also imply testing obligations end at deployment.
Fortunately, the guidance in Performance Measures includes a footnote acknowledging that states may require continued testing, even after closeout if needed, which establishes a broader role of testing as a means of ensuring that LEO providers adhere to the continued service commitments in the 10-year extended period of performance following the deployment period:
“…under certain circumstances, an Eligible Entity may have legitimate basis to request a subgrantee to provide performance data… after closeout.” (Performance Measures footnote 8, p. 4)
This footnote is a key to states’ ongoing testing for both terrestrial and LEO projects. The states have a “legitimate basis” for monitoring performance after closeout within the Federal Interest Period of 10 years after deployment for terrestrial networks and after deployment in the equivalent extended period of performance for LEO networks.
How states can strengthen their approach to LEO testing
States should clarify for their subgrantees, including their LEO subgrantees, that there are two distinct types of testing:
- Testing for NTIA reporting (provider-level, annual)
- Milestone and closeout testing (project-specific, tied to payments)
These types of testing serve different purposes, even if they use similar methodologies. States should consider documenting this distinction in their subgrant agreements, milestone reimbursement procedures, and technical assistance to subgrantees to provide clear authority for project-specific testing, including after closeout.
Clarifying a state’s testing approach will have an additional benefit: It will communicate to local governments that the localities have a role in holding the LEO providers to their commitments. Local governments can field complaints from local businesses and residents, assist those LEO subscribers with conducting their own speed tests, and aggregate the results for the state broadband office. Receiving those data will enable the state broadband office to require testing from LEO providers focusing on these trouble spots, prior to releasing the next scheduled funding allocation to the LEO provider for the project(s) affected.
What should LEO testing look like during the extended period of performance?
Even if NTIA has not clarified the full framework, states still should ensure that LEO providers remain compliant before releasing ongoing payments.
States that implement clear testing protocols now will be better positioned to verify compliance, support enforcement actions if needed, and protect their BEAD investments over the extended period of performance. States could adopt one of several approaches:
- Targeted testing: The state could test only when issues arise (e.g., customer complaints, public speed-test data).
- Streamlined testing: The state could test across projects during the extended period of performance, followed by targeted supplemental tests when problems emerge.
- Full project-level testing: The state could treat each payment during the extended period of performance like a recurring milestone and require mandatory verification.
Regardless of the specific approach, the clearest path forward is to establish—in subgrant agreements and compliance documentation—that project-specific milestone and closeout testing is distinct from annual NTIA reporting requirements. This distinction is grounded in federal grant requirements and reflects how states already manage terrestrial projects.
And states can work with local communities to ensure LEO service commitments are honored. Whether fielding and collecting customer complaints and/or sponsoring and aggregating customer speed tests to establish patterns of trouble spots, local governments and community organizations can alert states to where formal testing could be focused to validate service commitments.