We check the facts your next move depends on.
Blue Prysm is a boutique consultancy for telecom, satellite and technology businesses. Most of the decisions we see turn on a technical fact nobody in the room can check and a commercial question about who will pay. We work both from the evidence, then stay until a first offering is in the market with your team running it.
35 published analyses, every figure sourcedA free spectrum pricing tool132 strategy frameworks
$0.198 per MHz-POP
CBRS Priority Access, 3.55 to 3.65 GHz
Auction 105 · 2020 · FCC
What the industry says and what the filings show
Four things the industry repeats, set beside the filings and the arithmetic.
-
What is saidReported
“Satellite is for the last 2% of coverage.”
What the record showsFiled
79 MHz
of US spectrum SpaceX has contracted to own: 65 MHz of mid band from EchoStar and, since 8 October 2026, up to 14 MHz of low band from Grain
Nobody needs 79 MHz to supplement someone else’s network. The 800 MHz block is the band Sprint built on, it gets a phone’s signal out of a building, and most phones already support it.
Verdict: The supplement is buying a carrier’s spectrum position.
Direct-to-device → -
What is saidReported
“AI is cutting operators’ costs.”
What the record showsFiled
$150.9bn
spent in 2025 by AT&T, Verizon and T-Mobile running their networks and serving their customers
Not one annual report we read puts a currency figure on what AI has saved, and that spending is within 1% of what it was two years earlier.
Verdict: The saving is claimed often and counted nowhere.
AI in telecom operations → -
What is saidReported
“CBRS is the home of the private network.”
What the record showsFiled
86%
of the money in the CBRS auction was paid by mobile and cable operators
And on NTIA’s count, fewer than one radio in twenty in the band is indoors.
Verdict: The band was sold as private and bought by operators.
Private networks and neutral host → -
What is saidReported
“A 6G network will be a radar.”
What the record showsCalculated
≈1.5 m
the finest range cell of a 100 MHz carrier, against the metre the standard asks for
And a radar echo fades with the fourth power of distance, where a phone’s signal fades with the square.
Verdict: It is a radar, but a coarse one.
Integrated sensing →
The series
Six series, thirty-five parts, each question answered from the licence file, the filing or the standard. Every series is also an ebook.
-
Seven questions about direct-to-device
Two satellite operators may transmit on American mobile spectrum, and both rent it.
7 parts · ebook
-
Eight questions about 6G
6G has a calendar with meeting numbers on it, and nothing yet that a carrier can sell.
8 parts · ebook
-
Five questions about integrated sensing
A mobile network that senses is a radar, and a poor one next to the real thing.
5 parts · ebook
-
Five questions about AI in telecom operations
Every operator names AI in its cost programme, and not one annual report we read puts a currency figure on what it has saved.
5 parts · ebook
-
Five questions about non-terrestrial networks
Altitude makes the first four design decisions before anyone at a satellite company makes one.
5 parts · ebook
-
Five questions about private networks and neutral host
The shared band at 3.5 GHz was sold as the home of the private network, and mobile and cable operators paid 86% of the money.
5 parts · ebook
Price a band
Set a bandwidth and a population, and see what the licence would cost at each of the five US prices we use as comparables.
Implied price range
$1.35bn to $8.10bn
Low: 20 MHz × 340,110,988 × $0.198 = $1,346,839,512
High: 20 MHz × 340,110,988 × $1.191 = $8,101,443,734
Prices per MHz-POP from FCC auction results and EchoStar’s Forms 8-K, as part 4 of the direct-to-device series sets them out. CBRS is priced per licence, and the part explains why.
Open the full tool →The dates already fixed
Dated events from our analysis, counted down from today in your browser, soonest first. Each one links to the part that sets it out.
| Date | Event | Series | Days to go | Calendar |
|---|---|---|---|---|
| 30 Nov to 1 Dec 2026 | ITU-R Study Group 5 meets to approve the IMT-2030 requirements and evaluation guidelines | 6G | ||
| 7 to 10 Dec 2026 | 3GPP RAN#114 in Boston, the next decision point on sensing at Level C or Level D | Integrated sensing | ||
| 15 to 19 Mar 2027 | 3GPP RAN#115 in Rotterdam, where the four sensing specifications are due for approval | Integrated sensing | ||
| 27 Apr 2027 | Bidding opens in FCC Auction 115, for 160 MHz of upper C-band at 3.98 to 4.14 GHz | 6G | ||
| 10 Jun 2027 | Google’s CBRS database closes, and every customer on it must move to another | Private networks | ||
| 4 Jul 2027 | NTIA must identify its first 200 MHz of federal spectrum under the 2025 pipeline law | Private networks | ||
| 5 Nov 2026 | Grain must tell the FCC which satellite operator its 800 MHz process chose; it agreed to sell to SpaceX on 8 October | Direct-to-device | ||
| 4 Dec 2026 | Satellite applications for the 800 MHz band are due at the FCC under the order that approved its sale | Direct-to-device | ||
| 30 Nov 2027expected | The second-step closing of EchoStar’s spectrum sale to SpaceX | Direct-to-device | ||
| 2 Dec 2027 | The EU’s duties for AI in critical digital infrastructure start | AI in operations | ||
| 9 May 2028 | All 1,197 of AST SpaceMobile’s American authorisations run out | Direct-to-device | ||
| Dec 2028month only | 3GPP freezes the first 6G specifications, in Release 21 | 6G | ||
| 30 Nov 2029 | SpaceX’s first performance test on its EchoStar spectrum: a 5 dB downlink signal-to-noise ratio and 70% availability over 90% of each area | Direct-to-device | ||
| 12 Dec 2030 | The renewal window opens for the 17,460 CBRS Priority Access Licences that expire on 12 March 2031 | Private networks | ||
| 7 Oct 2032 | Half of SpaceX’s 15,000-satellite direct-to-device constellation must be in orbit and operating | Direct-to-device |
The decisions we work on
The six decisions we are asked about most. Open one to see the first three things we would look at.
-
01“A competitor just launched FWA in our footprint.”
The first three things we would look at
- How much of your base is addressable by their offer, by speed tier and location, rather than the headline coverage claim
- Which cohorts churn first, and what they are worth over the contract, not the month
- Where their fixed-wireless capacity runs out as they add subscribers, and what that does to their own service
-
02“We’re being asked for our satellite plan.”
The first three things we would look at
- What a satellite link can realistically deliver to an unmodified handset today, versus the roadmap
- Who owns the customer relationship, the SIM and the data in each partnership model
- What the economics look like at renewal, not at signing, which is where the leverage moves
-
03“We need a spectrum position before the auction.”
The first three things we would look at
- What comparable licences cleared at, per market rather than a national average
- How much depth you need for the service you are selling, and where the next MHz stops adding capacity
- What the band does not buy you: indoor penetration, device ecosystem, equipment timing
-
04“We’re buying or selling a network.”
The first three things we would look at
- The spectrum position: what is owned, what is leased, term, encumbrances and expiry cliffs
- Whether the coverage and capacity claims hold up against the licence terms and the propagation
- The build obligations and clearing costs attached to each licence
-
05“Our 6G roadmap needs a business case.”
The first three things we would look at
- Where each 3GPP and ITU-R milestone falls today, and what a slip in any one of them would do to the plan
- Which capabilities carry a business case before the Release 21 freeze, on 5G-Advanced hardware already in the field, which need a partner or a permission outside the carrier, and which only pay once the standard exists
- What the US spectrum pipeline releases and when: the upper C-band in Auction 115 from April 2027, the 800 MHz that section 40002 of P.L. 119-21 requires by 2034, and the 7.125 to 8.4 GHz range WRC-27 is studying, with the statute excluding its upper 1,000 MHz and the US advisory committee split three ways on the rest
-
06“Our AI plan needs prioritising.”
The first three things we would look at
- Where AI changes the unit economics in a telecom operation, and where it stays a demo
- What data and integration each use case needs before it works
- The sequence: what to do first, what it opens up, and what to wait on
How an engagement runs
The first step is a one-day teardown. Each step stands on its own, has a fixed fee agreed before it starts, and you decide whether the next one is worth doing.
Step 1
One-day teardown
A 90-minute working session on the decision in front of you, a written answer in two working days, and the model behind it, yours to keep.
Step 2
Two-week assessment
Two working sessions, a 12 to 18 page memo with the working shown so your team can check every number, and a one-page board summary.
Step 3
Project
Scoped to the engagement, with weekly checkpoints, until a first offering is in the market with your team running it. Every model and source is yours at the close.
Every figure we publish carries its source and its date
We read the filings rather than the press releases, and we publish what we find. Six series are out, thirty-five parts in all, on direct-to-device, 6G, integrated sensing, AI in telecom operations, non-terrestrial networks, and private networks and neutral host. Each part asks one question and answers it from the licence file, the SEC filing, the auction result, the standards report or the statute, and each says what it does not cover.
Analysis · part 4 of 7
Price per MHz-POP · exhibit 3 of the part
What spectrum costs
SpaceX leases 10 MHz of paired PCS G block from a carrier today and has agreed to pay $19.6B for 65 MHz of its own. That is six and a half times the bandwidth and no landlord, at a price within 4% of what the C-band auction cleared at per MHz-POP. The second-step closing is expected in November 2027.
EchoStar Forms 8-K · FCC licence file, 13 Sep 2026
Read it →Analysis · part 6
Who can afford to play
The three national carriers generated $105.4B of operating cash flow in FY2025 and had $33.8B left once the network and the shareholders had been paid, against about $337B of net debt.
SEC FY2025 cash-flow statements, retrieved 17 Sep 2026
Tool · live
What does spectrum cost?
Give it a band, a market and a bandwidth and it returns what comparable spectrum has cleared at, with the auction result behind every figure.
All of it
Every series and every tool
Thirty-five parts in six series, each series also as an ebook, with the tools and templates built from the same sources.
Put 132 strategy frameworks to work on your own business
The library runs from SWOT and Porter’s Five Forces to Wardley mapping and spectrum valuation. Each framework tells you when to use it, takes you through it question by question and works out the answer as you type. The result exports as a PowerPoint deck, a Word document or an Excel workbook, and nothing you enter leaves your browser.
Start from your question
Why has growth stalled?
Start with Revenue bridge →Is this market worth entering?
Start with Porter’s Five Forces →Will this investment pay back?
Start with Break-even analysis →Where does AI pay in this business?
Start with AI opportunity portfolio →Start with a diagnosis
Tell us what you are trying to decide. We reply in writing within two working days, with how we would approach it, what it would take and what it would cost. If the right answer is to change nothing, the reply says so.
If you would rather write directly, the address is hello@blueprysm.com.