Launch bottleneck & launch economics
THESIS
Global launch supply — not satellite manufacturing, not spectrum — is the binding constraint on every non-SpaceX constellation through ~2028. SpaceX's internal Starlink demand crowds out external customers on Falcon 9, and no alternative (Ariane 6, New Glenn, Neutron, Vulcan, Chinese commercial) yet offers comparable cadence at comparable $/kg. Falsifier: two consecutive quarters where a non-SpaceX provider flies >=1 dedicated LEO-constellation mission per month at a disclosed price within 2x of Falcon 9 rideshare rates.
STATE OF PLAY
Falcon 9 flew 59 Starlink missions out of 74 total in H1 2026 (~80% internal use, ~7 Starlink launches/month as of June 24 sweep). External LEO customers compete for the remaining slots. Rocket Lab's Iridium acquisition (June 29, $8B) is partly a launch-capacity play: captive Neutron capacity for Iridium replenishment removes that demand from the open market. Qianfan's cadence stall (no launch since June 5, ~7 launches short of its July target) shows the constraint operates inside China too — Long March supply is contended by Guowang. SES exposure: meoSphere (28 K2 platforms, first launch 2029) will buy launches in exactly the window when Kuiper replenishment, IRIS², and Guowang/Qianfan second shells all compete for capacity.
STRUCTURAL DYNAMICS
dossier for the group-financials mechanics; pricing can stay below
competitors' cost indefinitely as long as capital-markets access holds.
- Captive launch (SpaceX→Starlink, Rocket Lab→Iridium, CASC→Guowang) is
becoming the norm; the open launch market is what's left over. See
industry-consolidation dossier: captive-launch acquisitions are the
verticalization half of the current M&A wave, a direct structural
response to this bottleneck.
- Rideshare (Transporter/Bandwagon) is the marginal-capacity signal: slot
availability and pricing there move before dedicated-mission pricing does.
POLITICAL & REGULATORY
political launch demand is price-inelastic and further tightens supply.
- Chinese launch capacity is state-allocated; Qianfan vs Guowang contention
is a political queue, not a market.
IMPLICATIONS FOR SES
meoSphere's 2029 first-launch commitment sits in the projected peak-contention window. Threads: "O3b mPOWER sovereign pipeline", "Non-Western sovereign constellations", "Chinese BRI Digital Silk Road". Watch: any K2/SES launch procurement disclosure; Neutron/New Glenn cadence proof points.
ANALYTIC STANCE
Trust planet4589 (McDowell) over Chinese official figures for satellite counts — cross-check every Xinhua/CASC number against it before citing. Trust disclosed 8-K/S-1 launch-contract terms over analyst $/kg estimates, which are usually back-of-envelope. Known blind spot: no independent verification yet of the Transporter-deprioritization hypothesis (Open Question 1) — it originated as an analyst hunch, not an observed pattern; don't let it harden into stated fact before evidence arrives.
OPEN QUESTIONS
Starlink? Evidence: Transporter mission gaps or manifest slips on the
SpaceX site / planet4589 log. (Flagged by desk owner 2026-07-05 —
unverified.)
2. What is the real $/kg on recent dedicated LEO-constellation contracts?
Evidence: any disclosed launch contract value with mass manifest.
3. Does Qianfan's stall resolve via cadence recovery or target slip to 2027?
Evidence: next Spacesail launch or official timeline revision.
SUPERSEDED
(none yet — dossier created 2026-07-05)
TRACKED SERIES
| Date | Metric | Value | Source |
|---|---|---|---|
| 2026-06-24 | F9 launches YTD (total / Starlink) | 74 / 59 | brief 2026-06-27, planet4589 |
| 2026-06-24 | F9 Starlink cadence | ~7/month | brief 2026-06-27 |
| 2026-06-05 | Qianfan cumulative sats | ~200 | brief 2026-07-04, planet4589 cross-check |
| 2026-06-17 | Guowang cumulative sats | ~177 | brief 2026-06-26 (LM-12 Y7, group 22) |
| 2026-06-29 | Rocket Lab–Iridium EV | $8.0B | brief 2026-06-30 (8-K) |