Standards
Corrections
When we get something wrong, we fix it and say so. This page is how to tell us, what we do about it, and a dated list of the material corrections we have made. The wider rules are in our editorial standards, and how each computed number is made is on the methodology page.
Report a correctionHow to report an error
- Use the "Report a correction" button on this page or on any article: it opens our contact form with a correction template. Tell us what is wrong and, if you can, what the right figure or date is and where it comes from.
- Sponsored and partner content is covered too: we fact-check it before it publishes and correct it afterwards.
- Promise Ledger entries (/promises) each have their own correction link. Send the primary source for the words or the outcome; the organisations listed use the same form and the same standard of evidence. A confirmed change to an entry's quote, date or outcome is logged below.
How we correct
- A factual error is fixed where it appears, and the piece carries a dated correction note. We do not quietly edit a claim out of existence.
- When the same wrong fact appears on several pages, we fix it everywhere it appears and, where we can, add an automated check so it cannot come back.
- If a piece is wrong in its premise rather than its details, it is withdrawn, and its address says so instead of showing a blank error page.
- Material corrections (a wrong fact, figure or label a reader could have relied on) are listed below with the date. Typos and wording changes are not.
Who handles it, and how long it takes
- Corrections are handled by the founder, Jay Griffiths, not by the pipeline that wrote the piece.
- We acknowledge every correction report within 2 business days: a person replies to you. How fast a confirmed error is fixed depends on the kind of error (the service levels below, the same for everyone), and a material correction is logged here with its date.
Service levels, by kind of error
The same for everyone: readers, companies with a claimed profile and companies without one. No account is needed, and nobody can pay for a faster correction.
| Kind of request | For example | Service level |
|---|---|---|
| Every request | Any correction, from anyone, with or without an account | Acknowledged within 2 business days |
| Errors a filing proves | Listing status, ticker, legal name, headquarters, a Form D round | Fixed, or we write to you with what we are checking, within 2 business days; applied within 5 |
| Judgement items | Sector, subsector, tags, headcount band | Decided in a monthly batch; fact-checked where they move a score gate |
| Funding | Rounds, amounts, investors | Our standing rule: an independent fact-check before any financial figure is published |
We will publish the median time from report to correction once there are enough corrections reported through this process to measure it honestly.
Listed companies: please cite public filings or exchange releases only. Do not send material non-public information; anything that looks like it is deleted and not used.
For companies
Claiming a company profile, and every claim function, is free, forever. We never charge to claim, to verify or to correct a profile.
You do not need to claim a profile to correct it: use the same report link as everyone else and include a source (a filing, a registry record or your own published page). Not sure an email really came from us? See how to recognise email from SpaceNexus.
Corrections log
Newest first. Dates are the day the correction went live (US Pacific time). 22 entries.
Found in our own review
Monthly Hiring Index, October 2026 edition: the same-employer change, the Hiring Index level and the fastest-mover tables; and the October fastest-hiring ranking built on them
What was wrong: The October edition reported a same-employer change of -2,765 postings (-54.8%) across 40 employers and a Hiring Index of 55.4. On the morning of 31 October, the day of our month-end snapshot, every Workday job board answered with Workday's maintenance page, so eleven of the employers in the comparison, all read through Workday, had no count that day, and the comparison counted each of them as zero, including Blue Origin, which had 1,764 postings at the end of September. Separately, the counting-rule changes we made later on 5 October (21:56 UTC), after that morning's 07:45 UTC snapshot, were applied from 5 October, so the movers compared counts made on two different rules: Voyager Technologies led the fastest decliners at -146 (206 to 60), SpaceX read +175 (2,645 to 2,820), and Northwood Space read -14 (74 to 60).
What it says now: An employer whose board could not be read on a month-end snapshot is compared at its latest complete read from the month's last 3 days, carried forward and labelled as such, or left out of that month's comparison when there is none; a failed read never counts as zero. The eleven are compared at their 30 October reads, and with Firefly Aerospace left out (the correction below), October now reads +482 (+9.9%) across 39 employers, Hiring Index 134.6. A rule change now takes effect from the first snapshot taken after we make it: Voyager Technologies reads -14 (74 to 60), SpaceX +205 (2,615 to 2,820) and Northwood Space -12 (72 to 60), all from 6 October. The headline totals and the top 10 did not change. Both rules are on the methodology page, dated 2 November 2026.
Hiring Index, October 2026Fastest-hiring, October 2026How we make our numbers
Found in our own review
Monthly Hiring Index, October 2026 edition: the fastest-decliners table and the same-employer comparison; and the 30-day decliners on the hiring trends dashboard
What was wrong: The October edition listed Firefly Aerospace among the fastest decliners at -86 (157 to 71 postings, -55%) and counted its 158 postings at the end of September against 71 at the end of October in the same-employer comparison, and the 30-day decliners on /hiring-trends led with Firefly at 147 to 35. Firefly did not cut its hiring: from about 5 October it posted its roles on a new Workday careers site that our jobs sync did not read, and the ClearCompany board we did read emptied out as it moved.
What it says now: Firefly is left out of the October same-employer comparison and the fastest-mover tables, as the method says for an employer whose counting basis changed during the month, and it stays out of the 30-day movers until it has 30 days of history on its new board. Our jobs sync now reads Firefly's Workday site. An employer's move to a careers site we do not read is now recorded as a dated basis change, and our daily jobs check warns when a board loses half its roles in 28 days with no rule change on record. The headline totals and the top 10 did not change.
Found in our own review
Launch Cadence Index (/launch-cadence), its embed, and the launch pace figures that read from it (space stats, the 2026 launch schedule and rockets guides, SpaceNexus AM)
What was wrong: The index said it counted orbital launch attempts, but it counted every launch that lifted off, including suborbital flights: Blue Origin New Shepard, Rocket Lab HASTE and the Starship test flights before Flight 14. That added about 4 to 6 flights to 2026 and about 10 to 15 to 2025 to date, so the page showed 259 attempts against 265 (-2.3%) when the orbital pace was level with 2025, and the US and Rocket Lab counts were too high. Last year was also cut at the end of the same UTC day rather than the same moment.
What it says now: Only orbital attempts are counted, using the same rule as the Space Score (a suborbital orbit, New Shepard, HASTE, SpaceShipTwo, or a Starship flight before 2026-09-28 is not an orbital attempt), and both years are cut at the same moment. The page states the rule and how the full-year projection is made.
Found in our own review
Monthly Hiring Index, September 2026 edition: the same-employer change, the Hiring Index level and the fastest-decliners table
What was wrong: The September edition said postings at the same employers fell by 2,089 (-65.7%) and put the Hiring Index at 34.3 (August 2026 = 100), and its fastest-decliners table led with Anduril at -2,061. Hiring did not fall: on 26 September we narrowed Anduril, Hadrian and Hermeus to space-segment roles (Anduril went from 2,375 postings to 124; Hadrian and Hermeus to none). Those changes were made a day before our jobs sync began logging counting-rule changes, so the same-employer comparison treated them as a hiring collapse instead of leaving those employers out, as its method says it does.
What it says now: The three employers are left out of the September same-employer comparison, and Anduril's movers reading starts on 26 September. The same-employer change reads +177 (+22.5%) across 21 employers tracked on the same basis at both month ends, and the index 122.5. The headline totals (9,905 postings at month end, +3,172 on August) and the top 10 did not change. Rule changes made before the ledger existed are now kept in a fixed list that every comparison reads.
Found in our own review
Company profiles of private companies (the funding figure, now 'Raised privately'), the /company-profiles headline (now 'Private funding on record'), company cards, compare pages, the company CSV and open-data file, the MCP company tool
What was wrong: A private company's funding figure was the total recorded on its profile, not the sum of its funding rounds, and an independent check of every private company's rounds against filings and the companies' own announcements found figures that did not hold up. Rivada Space Networks' $2.5B rested on a single row no source supports, with no disclosed equity round; Vast's $1.3B included four rounds before 2026 that did not happen (its earlier money came from its founder); MinoSpace's $466M is not established as private funding (its listing prospectus shows its 2023 to 2025 capital increases were led by state or local-government funds); Exolaunch's $70M came from two unsourced placeholder rows; and Mynaric's $35M was one unsourced row dated while the company was listed. Rows were also recorded twice, rounds were missing, and debt, grants and state capital were recorded as equity.
What it says now: From 7 October 2026 a private company's figure is labelled 'Raised privately': the stated amounts on all its published private funding rounds, each with its source. Debt, grants, state capital, owner funding and acquisitions are listed in its funding history by class and not added in, and a company with no such round shows no figure. Profiles count every company, including defense companies; the funding tracker's totals count space companies only. We corrected 192 funding rows, removed 76 (duplicates, rounds that did not happen and unsourced entries) and added 195 missing rounds, each independently fact-checked; rounds dated in the frozen Q2 and Q3 2026 editions were corrected on those editions with a dated note. 183 private companies' figures changed, for example Rivada Space Networks no figure (was $2.5B), Vast $300M (was $1.3B), Hadrian $1.96B (was $2.36B), CesiumAstro $409.7M (was $682M) and Loft Orbital $324.6M (was $503M). The directory headline is now 'Private funding on record', $58.7B (it read $65.2B), and says what it leaves out.
Found in our own review
Company profiles of some private companies ('Total Funding'), the /company-profiles 'Funding Tracked' headline, the company CSV and open-data file, the MCP company tool
What was wrong: Several private companies' 'Total Funding' figures were not money the company raised from investors. Amazon Kuiper's $10.00B was Amazon's own budget for the project; Maxar's $6.58B was its $6.4B take-private by Advent plus a public offering; Terran Orbital's $3.57B added up two acquisitions and its SPAC merger; SpaceSail's $2.60B matched no source, and its known raises were led by state capital; Orbcomm's $460M was a debt refinancing; MaiaSpace's $135M was its parent's commitment; SpaceIL's $100M was donations to a non-profit; Relativity Space's $1.95B included an amount for an undisclosed 2025 investment; Ursa Major's $631M had no source; Privateer's Series A was counted twice; Neuraspace's figure included a Portuguese public grant; Anduril's Series F was recorded twice; and Spaceflight Inc, Bayanat and TiSpace showed totals no source supports. Together they added $26.3B to the 'Funding Tracked' headline.
What it says now: Each of these figures is now the sum of the company's private funding rounds on record, or no figure where there is none: Terran Orbital $33.0M (its 2017 Series A and the seed rounds of its subsidiary PredaSAR; a 2018 filing recorded as equity was debt, and a 'Series C' that never happened was removed), Orbcomm $108.8M (its 2004 and 2005 preferred rounds, from its IPO prospectus), Relativity Space $1.32B, Ursa Major $371.3M, Anduril $10.5B, Privateer $56.6M and Neuraspace $9.7M. Amazon Kuiper, Maxar, SpaceSail, Spaceflight Inc, Bayanat, MaiaSpace, SpaceIL and TiSpace now show no figure. The 'Funding Tracked' headline went from $91.4B to $65.2B. Our nightly updates now fill an empty figure only from private funding rounds on record, never from acquisitions, debt, grants, state capital or an owner's own money. Other private companies' recorded totals are being checked against their rounds.
Found in our own review
Company profiles of listed companies ('Total Funding'), the funding rows behind them, the /company-profiles 'Funding Tracked' headline, the company CSV and open-data file, the MCP company tool
What was wrong: A listed company's 'Total Funding' added up every funding row we held for it, including IPO proceeds, public offerings, debt facilities and tender offers, so it did not measure the private funding it raised. Telesat's $1.60B was one debt facility, Eutelsat's $4.06B was all debt, and SpaceX's figure included its IPO and debt. The 'Funding Tracked' headline on /company-profiles, which adds up every profile's figure, read $201.0B. Checking the rows behind the new figures against SEC filings and the companies' own releases also found rows that were wrong: Rocket Lab's two 'Series F' rows ($140M and $75M) describe rounds that never happened; BlackSky's, Planet's and Spire's only rows were unsourced and misdated or mis-sized; Astroscale had an $80M 'Series G' that was a government contract; York's three rows were unverifiable or a change of owner; Booz Allen's row was employee stock awards; and missing rounds left Firefly, Voyager, Planet, Spire, Synspective and Astroscale understated.
What it says now: A listed company's figure is now labelled 'Raised privately before listing': the stated amounts on its private funding rounds dated before it listed. IPOs, offerings, debt, convertible notes that did not convert into a priced round, tender offers and grants are listed in its funding history by class and are not added in, and a listed company with no such round reads 'No private round on record' (Telesat, Booz Allen and York now read that way). We corrected 15 funding rows, removed 8 and added 15 missing rounds, each with its filing or source, so that, for example, SpaceX reads $8.69B, Firefly $750.8M, Voyager $370.1M, Astroscale $348.7M (one more round with no stated amount), Planet $262.0M, Rocket Lab $158.2M, Synspective $140.7M, Spire $127.5M and BlackSky $100.7M. 29 listed companies' figures changed. The 'Funding Tracked' headline now reads $91.6B and says that listed companies count private rounds before listing only. Figures for companies that are not listed did not change. In the open-data file the column is renamed funding_usd, with funding_basis and listed_since beside it.
Found in our own review
Monthly Hiring Index, August and September 2026 editions, and the Fastest-Hiring Space Companies rankings for those months
What was wrong: When we changed how an employer's postings are counted (on 5 October, for example, SpaceX's count stopped including xAI and X roles), the change was also applied to months before it: every earlier snapshot of that employer was dropped, so it disappeared from past editions. From 5 October the September edition's top 10 left out SpaceX (2,610 postings at month end), Rocket Lab (533), K2 Space (245), Vast (204) and Voyager Technologies (192), and started with Blue Origin. The August edition lost employers in stages from late September; by 5 October its top 10 left out SpaceX (2,248), Anduril (2,179), Rocket Lab (430), Relativity Space (351), K2 Space (206), True Anomaly (182), Vast (171), Voyager Technologies (150) and Hadrian (139), and was filled out with employers holding 1 to 4 postings. The fastest-mover tables and the fastest-hiring rankings for both months left the same employers out.
What it says now: Each edition shows the month as it was recorded: an employer's figures for a month are counted under the rule in force that month, and a later rule change no longer reaches back. September's top 10 starts with SpaceX at 2,610 and again includes Rocket Lab, K2 Space, Vast and Voyager Technologies; August's starts with SpaceX at 2,248 and Anduril at 2,179. Fastest movers still start from a rule change made inside the month. Employers whose rules changed after the month are named under the top-10 table with the date. The headline totals did not change.
Hiring Index, September 2026Hiring Index, August 2026Fastest-hiring, September 2026
Found in our own review
The Launch Cadence and Slip Report, August 2026 edition (/releases/launch-cadence/2026-08)
What was wrong: The schedule-movement section counted every raw change in our launch date ledger, including changes to placeholder dates. Launch Library records a date known only to a month or a year as the last day of that period, so a placeholder being replaced read as a move. The August edition reported 5 manifest changes on 3 launches, net -12 days, with 4 moving a launch later and 1 earlier. Two of those figures came from placeholders: Crew-13 moving from Sep 12 to a Sep 30 end-of-month placeholder was counted as 17 days later, and Ariane 6 Metop-SG B1 going from "NET 2026" to "NET November" was counted as 31 days earlier. The other 3 were time changes of under a day to one Electron launch.
What it says now: The report uses the Launch Ledger method, the same one as /launch-slips: moves are netted per launch per day, and only a move between two dates known to the day counts as a slip. For August that is 4 changes and 0 slips: 2 changes of under a day to the Electron launch (its 3 raw changes, netted per day), 1 launch that lost its firm date (Crew-13) and 1 placeholder roll (Ariane 6). Placeholder rolls and lost dates are listed and labelled, never counted as slips.
Found in our own review
The /funding-tracker headline and the homepage funding tile
What was wrong: The "last 12 months" private funding figure added up every row in our funding table. On 25 September it read $119.0B, of which about $75.4B was the SpaceX IPO, $10.0B a SpaceX tender offer and about $4.1B debt. It also counted a $5B defense round (Anduril Series H) as space funding.
What it says now: The headline counts private funding rounds into space companies only. IPOs, other public offerings, tender offers, debt, grants, acquisitions and rounds into companies not classed as space are listed beside it, labelled, and never added in.
Found in our own review
The /investors directory
What was wrong: The headline tile read "Combined AUM $7.3T": the firm-wide assets of 49 investors added together, $6.3T of it from Fidelity and T. Rowe Price, none of it space money. Each card's "Total Invested" summed the full size of every round a firm appeared in, IPOs and tender offers included, which read as the firm's own cheque.
What it says now: Both figures are gone. The directory shows what we can count: private space rounds each firm led or joined, by date.
Found in our own review
/industry-trends and the benchmark cards on /space-stocks
What was wrong: The trends were labelled "figures as of 2025" with no source on any figure. A re-check against 2026 sources found only 9 of 119 metric and evidence lines held up as written. Some were invented (for example "Descartes Labs acquired by AntiSec"), and every market forecast lacked a named source. Two /space-stocks benchmarks were wrong or unsourced.
What it says now: Every remaining metric and evidence line carries a source a reader can open and a date. The unsupported lines and forecasts were removed, and an automated check fails on any line without a source or date.
Found in our own review
Starship facts on 36 pages and data files across the site
What was wrong: Pages across the site described Starship as having reached orbit, treated Flight 13 as a revenue flight, or described a planned ship catch.
What it says now: All 13 Starship flights flown by then were suborbital and no Starship had been in orbit; Flight 13 was not a revenue flight; Flight 14 was NET 28 September pending FAA approval, with no ship catch planned. The site now reads these facts from one source file.
Found in our own review
Intelligence briefs and a history page
What was wrong: Three weekly briefs dated February 2026 were not drawn from real reporting and described events that did not happen (for example, a Starship booster reflown within 48 hours). A history page said Starship Flight 4 (June 2024) reached orbit; it was suborbital.
What it says now: The three briefs were deleted. The history page was corrected, and its old address forwards to the corrected page.
Found in our own review
/news-aggregator and /videos
What was wrong: /news-aggregator showed about 30 invented articles credited to real outlets. On /videos, 11 of 12 curated YouTube videos did not exist, including an invented Starship "Flight 6 launch to catch".
What it says now: /news-aggregator was taken down and now forwards to /news. /videos lists verified official uploads only, and an automated check guards the list.
Found in our own review
Lunar facts across the site, and /moon
What was wrong: Pages described Artemis III as a late-2027 Moon landing, carried SLS Block 1B and Block 2 figures after NASA dropped the Exploration Upper Stage, gave an out-of-date Gateway status, and froze the Artemis Accords count at 43. A 60-item fact-check of /moon also corrected dates, details and 16 dead source links.
What it says now: Artemis III is a 2027 Earth-orbit docking test and the first landing is on Artemis IV. The Accords count reads from one dated site figure. No tracker score, lead or projection changed in the fact-check.
Found in our own review
The Space Score rankings
What was wrong: The scores were hand-set values from February, while the page said every figure was computed and dated that day. The quarter label came from the calendar, so each quarter republished the same numbers.
What it says now: Scores are computed from sourced data by a published method, and each carries its coverage, confidence and measurement date.
Found in our own review
Mission debriefs
What was wrong: Some debriefs stated outcomes the launch record did not support: successful launches published as "scrubbed", and press events, a static fire and a rollout written up as scrubbed launches. The pages also said the debriefs were "human-curated"; they are AI-drafted.
What it says now: 18 debriefs were unpublished. Debriefs are written only for real launches with a recorded outcome, the outcome comes from the launch record, and the pages say they are AI-drafted.
Found in our own review
/jobs, /space-environment and Mission Control
What was wrong: 62 illustrative job postings were re-created daily with fresh dates and listed as real openings. The solar flare forecast on /space-environment was randomly generated. Mission Control's Deep Space Network, EPIC and NASA image panels fell back to invented captions and telemetry.
What it says now: The illustrative postings were deleted. The flare forecast is NOAA SWPC's published 3-day probability with its issue date, or says it is unavailable. The Mission Control panels show real data or nothing.
Spotted by our editor
Launch pages
What was wrong: A panel headed "Live Telemetry" with a pulsing LIVE badge showed generated numbers over one generic two-stage ascent curve, drawn the same for every vehicle. It was neither live nor telemetry.
What it says now: The panel is labelled "Illustrative Ascent Profile" and says it is a generic reference curve, not this vehicle's flight plan and not measured data.
Found in our own review
Search snippets and page descriptions
What was wrong: Seven served figures were wrong, among them: Artemis Accords signatories (said 61, was 71), 2025 orbital launches (said 230+, was 324), active satellites (said 10,500+), and a Rocket Lab vs SpaceX comparison that said 55+ Electron and 300+ Falcon launches.
What it says now: Each figure was corrected against its source, including the comparison table on the Rocket Lab vs SpaceX page.
Found in our own review
The Blue Origin vs SpaceX guide and New Glenn comparison pages
What was wrong: The guide said New Glenn "lost its third vehicle in April 2026" and "has not yet reflown a booster".
What it says now: NG-3 on 19 April 2026 was New Glenn's first booster reflight. The pages were corrected against the record.
Corrections to Launch Ledger editions
Found in our own review
Slip Watch, week 2026-W39
What was wrong: The edition counted 28 date changes that our sync only caught up with during this week as if they had happened in it. Our sync had not been refreshing those launches when their dates changed (until 2026-09-24 it refreshed only the next 100 upcoming launches), so it recorded each backlog change on the day it next saw the launch, and the day each change was made is not known from our records. Launch Library 2's own records date 25 of them before this week, 19 of those before our recording began on 2026-08-29. As frozen, the edition read: 48 date changes: 1 slip (9.3 days), 0 pulled in, 6 same-day changes, 6 firmed, 3 lost a firm date, 32 placeholder rolls.
What it says now: Counting only changes our sync saw when they happened (method 1.1): 20 date changes: 1 slip (9.3 days), 0 pulled in, 6 same-day changes, 4 firmed, 1 lost a firm date, 8 placeholder rolls. The slip and pull-in figures are unchanged. The 28 caught-up changes are listed on the edition page, labelled, and not counted.
Corrections to Research release editions
Spotted by our editor
SpaceNexus Supply-Chain Concentration Report, Q2 2026
The supplier map in this edition was AI-written and is removed; see the sourced supply chain.
Replaced: Mapped relationships 30; Top supplier country United States; Top three countries 66.7%; Single-source dependencies 10.
Now reads: .
Spotted by our editor
SpaceNexus Supply-Chain Concentration Report, Q2 2026
The coverage statement on this edition was wrong about how the map was made. It said the relationships were curated by hand from public sources and reviewed against sources, that criticality and risk were ratings made by hand, and that nothing on the page was model-generated. The supply-chain map and the bill-of-materials register were written into our code by an AI assistant in February 2026 from its general knowledge, not from cited sources, and were last reviewed for errors (an AI-assisted review) on 2026-03-25; no relationship, rating or shortage carries a source. The figures are unchanged. The coverage statement now says how the map was made, and the figures describe that sample, not sourced research.
Replaced: Mapped relationships 30; Top supplier country United States; Top three countries 66.7%; Single-source dependencies 10.
Now reads: Mapped relationships 30; Top supplier country United States; Top three countries 66.7%; Single-source dependencies 10.
As-published headline restated
SpaceNexus Federal Space Awards, Q2 2026
Published 2026-09-16 at 02:33 UTC, before this quarter’s defense records were complete and while the first sweep of federal awards was still running: the edition could then see 25 attributed federal awards, and its headline read $48.6M of space-coded obligations, 99.5% of it from NASA. The sweep finished later; on 2026-09-24 SpaceNexus added the registrations (Unique Entity IDs) of company subsidiaries, which brought in most Department of Defense records; and defense awards reach USAspending 90 days after the action date. Attribution was also tightened on 2026-10-08 (a name alone no longer attributes an award). The edition is a published edition held provisional: its figures are recomputed from current data, with the agency split and the concentration measures withheld, until 2026-10-29, when the quarter’s records are complete and the edition is frozen with its final figures.
As published: Space-coded obligations $48.6M; Companies winning space work 4; NASA share of space-coded 99.5%; Top five share 100.0%.
Now reads (live provisional figures on 2026-10-08; the edition is frozen when its source records are complete): Space-coded obligations $1.49B; Companies winning space work 32; NASA share of space-coded –; Top five share –.
As-published headline restated
SpaceNexus Space Score Top 25, Q3 2026
This table was first published on 2026-09-15 as a live table that runs through its quarter. On that day it was the hand-set table that preceded the computed Space Score: a scored pool of 97, SpaceX top at 953, entry score 694. Computed scores replaced it on 2026-09-23 (methodology v2), and the edition froze as the table stood at the end of the quarter, computed under methodology v3 at 23:48 UTC on 2026-09-30 after that day’s audit corrections to the inputs. Later methodology versions (v4 to v6, October 2026) and later data repairs do not change this edition. From 2026-10-08 until this note, the frozen copy’s company count read "From 148 ranked of 356 scored": those were the counts of the October board, not the quarter-end one. The quarter-end archive did not record its counts, so the edition now says so instead of showing a number.
As published: Companies ranked 25; Top score 953; Entry score 694; Rank movement First edition.
Now reads: Companies ranked 25; Top score 876; Entry score 678; Rank movement First edition.
First freeze of an edition published before editions were frozen
SpaceNexus Hiring Index, August 2026
Published 2026-09-15, before SpaceNexus froze release editions. Until this first freeze the edition page recomputed from live data, and its figures moved as records were added or corrected. The headline as published is kept here. From 2026-10-08 an edition is frozen when it is published.
As published: Open roles at month end 6,733; Month-over-month –; New postings in the month 3,834; Remote share 1.4%.
Now reads: Open roles at month end 6,733; Month-over-month –; New postings in the month 3,885; Remote share (live) 3.7%.
Quoted as published; dashes are reformatted to house style.
First freeze of an edition published before editions were frozen
SpaceNexus Hiring Index, September 2026
Published 2026-10-04, before SpaceNexus froze release editions. Until this first freeze the edition page recomputed from live data, and its figures moved as records were added or corrected. The headline as published is kept here. From 2026-10-08 an edition is frozen when it is published.
As published: Open roles at month end 9,905; Month-over-month +3,172; New postings in the month 5,345; Remote share (live) 3.8%.
Now reads: Open roles at month end 9,905; Month-over-month +3,172; New postings in the month 5,348; Remote share (live) 3.7%.
First freeze of an edition published before editions were frozen
SpaceNexus Launch Cadence and Slip Report, August 2026
Published 2026-09-15, before SpaceNexus froze release editions. Until this first freeze the edition page recomputed from live data, and its figures moved as records were added or corrected. The headline as published is kept here. From 2026-10-08 an edition is frozen when it is published.
As published: Launch attempts 26; Outcomes 25 / 1; Year to date 212; Manifest changes observed 5.
Now reads: Orbital launch attempts 26; Outcomes 25 / 1; Year to date 206; Launch slips 0.
First freeze of an edition published before editions were frozen
SpaceNexus Launch Cadence and Slip Report, September 2026
Published 2026-10-06, before SpaceNexus froze release editions. Until this first freeze the edition page recomputed from live data, and its figures moved as records were added or corrected. The headline as published is kept here. From 2026-10-08 an edition is frozen when it is published.
As published: Launch attempts 27; Outcomes 27 / 0; Year to date 239; Launch slips 7.
Now reads: Orbital launch attempts 27; Outcomes 27 / 0; Year to date 233; Launch slips 7.
First freeze of an edition published before editions were frozen
SpaceNexus Space Insider Activity, August 2026
Published 2026-09-16, before SpaceNexus froze release editions. Until this first freeze the edition page recomputed from live data, and its figures moved as records were added or corrected. The headline as published is kept here. From 2026-10-08 an edition is frozen when it is published.
As published: Open-market insider purchases 5; Open-market insider sales 195; Disclosed purchase value $1.7M; 5% holder filings 102; Filings indexed 413; Lines excluded from the two figures above 182.
Now reads: Open-market insider purchases 5; Open-market insider sales 198; Disclosed purchase value $1.7M; 5% holder filings 113; Filings indexed 442; Lines excluded from the two figures above 211.
Quoted as published; dashes are reformatted to house style.
First freeze of an edition published before editions were frozen
SpaceNexus Most Active Space Investors, Q2 2026
Published 2026-09-15 while the edition page recomputed from the live funding record. Two things moved it, and the counting rule is by far the larger. The edition as published counted every transaction recorded in the quarter: $75B of its $84.9B was one SpaceX row then recorded as an IPO, and the rest included tender offers, public offerings, debt, grants and companies doing no space work. From 2026-09-16 only private space rounds count, and every transaction the rule excludes is listed on this page with the reason (25 transactions carrying $118.6B in this edition). Separately, rounds dated in the quarter were added and corrected after publication: 69 transactions are now recorded for the quarter, against 39 when it was published (the dated corrections to the Q2 2026 funding edition are listed on /corrections). From 2026-10-08 an edition is frozen when it is published.
Note revised . As first published: Published 2026-09-15 while the edition page recomputed from the live funding record. Rounds dated in the quarter were added and corrected after publication (the dated corrections to the Q2 2026 funding edition are listed on this page), so the page moved. From 2026-10-08 an edition is frozen when it is published.
As published: Rounds recorded 39; Disclosed capital $84.9B; Rounds naming a lead 27 of 39; Distinct lead investors 26.
Now reads: Space rounds counted 44; Disclosed capital $3.42B; Rounds naming a lead 36 of 44; Distinct lead investors 54.
Quoted as published; dashes are reformatted to house style.
Corrections to Chart of the Week editions
Found in our own review
Chart of the Week, 2026-W39 (week of Sep 21, 2026 – Sep 27, 2026)
This edition counted 61 "slips" in four weeks under an earlier definition: any launch date that moved by more than a minute, including placeholder rolls ("NET 2026" to "2027") and same-day changes. On 26 September 2026 we changed the definition to real slips only: a move of a day or more later between two dates known to the day, netted per launch per day. That is the definition the Launch Cadence and Slip release uses, which counted 7 real slips in all of September. The subtitle said "last 8 weeks", but our slip ledger began recording on 29 August 2026, so the chart has four weekly points.
The chart and its caption are kept exactly as published; the correction is printed on the edition. The edition, with the correction
Found in our own review
Chart of the Week, 2026-W40 (week of Sep 28, 2026 – Oct 4, 2026)
Titled "Orbital launches per month", this edition counted every recorded launch, including 11 suborbital flights in the 12 months (3 New Shepard, 5 Rocket Lab HASTE and 3 Starship test flights before Flight 14). Orbital launch attempts were 329, not 340. From 8 October 2026 the chart counts orbital attempts only.
The chart and its caption are kept exactly as published; the correction is printed on the edition. The edition, with the correction
Found in our own review
Chart of the Week, 2026-W41 (week of Oct 5, 2026 – Oct 11, 2026)
The caption said SpaceX's 34 launches were "52.3% of the total", but it divided by 65, the launches of the eight providers shown. 75 launches were recorded in the 90 days, across 18 providers, so SpaceX's share was 45.3%; the table's "Total" row (65) is the total of the eight providers shown. The count also included one suborbital Starship test flight (Flight 13, 24 July 2026): counting orbital attempts only, as the chart does from 8 October 2026, SpaceX had 33 of 74 (44.6%).
The chart and its caption are kept exactly as published; the correction is printed on the edition. The edition, with the correction
Corrections to quarterly funding editions
Spotted by our editor
Space funding edition, Q3 2026
Corrected after the October 2026 Research accuracy review and the full re-verification of 2026 funding rows (Q3 2026). Orbit Fab's "more than $25M in interim financing" is held out: no source states the instrument (rule 1, instrument not stated). Interlune's $5,000,000 September raise is a SAFE that has not converted, so it is debt under rule 1 and not a private round; with no company announcement it is dated to its Form D first sale, 31 August 2026. Xingyi Lianxin's A+ and A++ rounds are state capital, not private funding: both were led by government-owned investors (rule 3). The edition now reads $3,294,843,403 across 45 rounds (was $3,299,843,403 across 48).
Replaced $3.30B across 48 rounds; the edition now reads $3.29B across 45. The corrected edition
Spotted by our editor
Space funding edition, Q2 2026
Corrected after the October 2026 Research accuracy review and the full re-verification of 2026 funding rows (Q2 2026). Zenk Space's CNY 180M round ($26,528,463) is state capital, not private funding: no lead is named and three of its four named investors are government-controlled (rule 3, no-named-lead test). ICEYE's EUR 450M Series F is converted at the ECB rate of 9 June 2026: $520,785,000 (was $520,000,000). Isar Aerospace's EUR 270M Series D is converted at the ECB rate of 9 June 2026: $312,471,000 (was $312,000,000). Optera's GBP 3M round is state capital, not private funding: it was led by the UK Innovation & Science Seed Fund, financed by UK government bodies (rule 3, government-financed funds); it is dated 15 June 2026. Sophia Space's $7M June raise is a SAFE that has not converted, so it is debt under rule 1, not a private round (it was counted because its round type was blank). Dispatch's $500,000 is held out: no source says what instrument it was (rule 1, instrument not stated). The edition now reads $3,423,804,785 across 44 rounds (was $3,460,599,648 across 48).
Replaced $3.46B across 48 rounds; the edition now reads $3.42B across 44. The corrected edition
Spotted by our editor
Space funding edition, Q3 2026
The Q2 2026 edition was corrected on 7 October 2026; the comparison now reads the corrected figure: $3,460,599,648 across 48 rounds, was $3,344,329,319 across 46, so the change against Q2 2026 is -4.6%, was -1.3%; the sector comparisons read the corrected Q2 2026 edition too. This edition's own rounds and totals are unchanged.
The comparison with Q2 2026 replaced $3.34B across 46 rounds and now reads $3.46B across 48; the edition's own figures are unchanged. The corrected edition
Spotted by our editor
Space funding edition, Q2 2026
Corrected after the October 2026 fact-check of the Starfish Space, QOSMIC and Skybound Medtech rows (Q2 2026 rows). Starfish Space's April 2026 Series B, announced as "over $100 million" and counted until now with no dollar figure, counts at $111,690,329, the total amount sold on its SEC Form D (0001886318-26-000001), the exact figure for the same round. QOSMIC's $3.33 million seed round, announced on 24 June 2026, is added: the row had no source link, so the open-data rule left it out; it now cites the company's own release. Skybound Medtech's $1,250,000 equity raise on SEC Form D (0002080612-26-000042), first sale 8 June 2026, is added: it was recorded after the edition was frozen. Rows counted in the private-funding total: Starfish Space, QOSMIC, Skybound Medtech.
Replaced $3.34B across 46 rounds; the edition now reads $3.46B across 48. The corrected edition
Spotted by our editor
Space funding edition, Q3 2026
Corrected after the October 2026 independent fact-check of private companies' funding rows (Q3 2026 rows). Antares's July 2026 Series C is added: $370 million of equity; the $100 million of debt announced with it is not added in. Chang Guang Satellite's round of nearly RMB 5 billion keeps no dollar figure (an approximation); the stated figure, its co-lead investors and the company's own announcement are recorded. Galactic Energy's July 2026 D+ round of several hundred million yuan, mostly from state investment platforms, is added as state capital with no dollar figure. Galactic Energy's Pre-IPO round of several billion yuan is added with no dollar figure: no exact amount was stated. HyImpulse's Series A extension of more than EUR 50 million keeps no dollar figure (a floor); the stated figure and its co-lead investors are recorded. Insera Solution's Series A lead investor, SV Investment, is confirmed against its source. Orbit Fab's "$20 million Series B" of March 2026 was a signed term sheet, not a closed round: the row now records the more than $25 million of interim financing led by Stride Capital that the company announced on 7 July 2026, with no dollar figure (a floor), so it moves into this quarter. Space Pioneer's July 2026 round of CNY 200 million from Hengdian Capital is added, converted at the ECB rate of 7 July 2026: $29,439,938. Rows counted in the private-funding total: Antares Industries, Chang Guang Satellite Technology, Galactic Energy, HyImpulse, Insera Solution, Orbit Fab, Space Pioneer (Tianbing Technology). Rows listed beside the total, never added in: Galactic Energy (state capital).
Replaced $2.90B across 44 rounds; the edition now reads $3.30B across 48. The corrected edition
Spotted by our editor
Space funding edition, Q2 2026
Corrected after the October 2026 independent fact-check of private companies' funding rows (Q2 2026 rows). Astranis's $155 million delayed-draw credit facility from Trinity Capital, announced with its May 2026 Series E, is recorded as debt. ElevationSpace's Series B, announced as JPY 6.4 billion, is converted at the ECB rate of its announcement day: $39,695,370, not the rounded $40 million. Hermeus's April 2026 Series C equity was recorded twice, as a press-reported $200 million and as its Form D ($200,589,992): the press row is removed and the Form D row is dated to the 7 April announcement, labelled Series C, with lead investor Khosla Ventures. A second source is attached to Orbcomm's $460 million debt refinancing. Turion Space's Series B was stated as more than $75 million, a floor, so it keeps no dollar figure; the stated figure and lead investor Washington Harbour Partners are recorded. Rows counted in the private-funding total: ElevationSpace, Turion Space. Rows listed beside the total, never added in: Astranis, Orbcomm (debt); Hermeus (not classed as a space company by its recorded sector).
Replaced $3.34B across 46 rounds; the edition now reads $3.34B across 46. The corrected edition
Spotted by our editor
Space funding edition, Q3 2026
Corrected 2026-10-05 (second): SpaceSail's CNY 6.976 billion round, mostly from state-owned funds, is now listed beside the total as state capital; the private-funding headline is unchanged.
Replaced $2.90B across 44 rounds; the edition now reads $2.90B across 44. The corrected edition
Spotted by our editor
Space funding edition, Q3 2026
Corrected 2026-10-05 after an independent fact-check of every row. First frozen on 2026-10-05 at $14,500,175,077 across 23 rounds; now $2,900,403,465 across 44 rounds (16 with no dollar figure). Blue Origin's reported $10 billion round has not closed (its CEO to CNBC, 2026-09-30), so it is listed under Raising now and not counted. Hadrian ($1.37 billion) is recorded as a defense and industrial manufacturer, and its round announcement describes defense, aerospace and industrial work rather than space: it is listed beside the total, as Castelion and Venus Aerospace are. Open Cosmos's EUR 300 million includes a venture debt tranche of undisclosed size, and Orienspace's "several hundred million yuan" is not an exact figure: both count as rounds with no dollar figure. Neuraspace is converted at the ECB rate of its announcement day (+$9,323). 23 closed rounds the first freeze missed are added after a second independent check (11 with an exact figure, +$131,549,065; 12 with no dollar figure). Beside the total: Castelion is $800 million of equity plus a $250 million credit facility (debt), and BlackSky's $150 million of share sales took place in Q2 and moved there. Announcement dates, round labels and lead investors are corrected for Interlune (a SAFE), Space Epoch, Array Labs, Pixxel, EnduroSat, Ravee Optics and others.
Replaced $14.50B across 23 rounds; the edition now reads $2.90B across 44. The corrected edition
Spotted by our editor
Space funding edition, Q2 2026
Corrected 2026-10-05: BlackSky's $150 million of at-the-market share sales took place during Q2 2026 (its Q2 results: "During the quarter, the Company raised $150 million"), so the offering is listed beside the Q2 total (it had been dated August, in Q3). Offerings by listed companies are never in the total, so the Q2 figure does not change.
Replaced $3.34B across 46 rounds; the edition now reads $3.34B across 46. The corrected edition
Questions about a correction: contact us.