The 2026-2027 AI/Space/Data-Infra IPO Pipeline as a Quarterly Sanity Check Lead-Magnet Series
The question
What's the realistic 2026-2027 IPO pipeline across AI/space/data-infra names (SpaceX, Anthropic, Databricks, etc.), and which make the strongest "junior-analyst IC memo" lead-magnet subjects for Sanity Check? Context: the SpaceX IPO-memo lead magnet is an open RDCO board task; ARK's "SpaceX IPO wave" note and an Anthropic IPO filing both landed the week of 2026-06-07. The goal is to turn one-off memos into a QUARTERLY lead-magnet series that ties the investing cluster to the content cluster. This is a content lead-magnet brief, NOT investment advice — no buy/sell.
What we already know (from the vault)
- The IC-memo-as-lead-magnet idea already has a worked exemplar in-vault: [[2026-05-12-mostlymetrics-cerebras-ipo-s1-breakdown]] is, structurally, exactly the junior-analyst IC memo format we want to ship — revenue trajectory, customer concentration, cap-table, red flags, valuation comps, all source-cited from the S-1. CJ Gustafson's S-1 breakdown is the template to imitate (and Cerebras is now public, so the data is fully available).
- The strongest original framing for the series is already in the vault: Packy's "Tickered vs scarce" idea ([[2026-04-30-not-boring-scarce-assets-abundance-driven-scarcity]]). Pre-IPO names feel scarce (narrative, secret, status); the same company gets "Tickered" months after listing — sortable in a screener alongside 248 peers. A pre-IPO IC-memo series sells readers the scarce-asset moment before the Tickering. That is the editorial hook that clears the no-derivative bar ([[feedback_no_derivative_sanity_check_pieces]]).
- The investing-cluster anchor doc [[2026-05-12-innermost-loop-ai-infrastructure]] already maps the frontier-lab Layer 1 (Anthropic ~$1T secondary, OpenAI $852B→$1T, xAI $230B) and the chip/data-infra layers — so the memo subjects double as instrumentation for an active RDCO thesis, not just content.
- Lead-magnet doctrine in-vault: readers pay for "holdable" objects — a memo is exactly that ([[2026-02-24-ship30for30-paid-newsletter-checklist]]); and personality-driven framing ("Sanity Check by the founder") is the scarce alternative to generic AI-written takes ([[2026-04-30-not-boring-scarce-assets-abundance-driven-scarcity]]).
- Vault note: the Cerebras memo already concluded Cerebras is NOT an RDCO position (86% revenue from two related Abu Dhabi entities, upside-down cap table). That's fine — a lead-magnet memo's job is to be a credible analytical read, not a buy rec, so even "pass" names make good memos.
What the web says
Pricing/timing as of 2026-06-12 (favoring the most current signals; SpaceX is pricing/trading as this brief is written):
| Name | Sector | Filing status | Rough timing | Data availability for a memo |
|---|---|---|---|---|
| SpaceX | Space | Public S-1 filed 2026-05-20 (confidential 4/1); roadshow 6/4 | Pricing 6/11, first trade 6/12 (Nasdaq: SPCX) | HIGH — full S-1; ~$18.7B 2025 rev (unaudited), $1.75T target valuation, $135/sh, 556.6M shares |
| Anthropic | AI lab | Confidential draft S-1 filed 2026-06-01 | IPO target ~Oct 2026 (~90% 2026 odds) | MEDIUM — confidential S-1 (not yet public); ~$965B last mark, ~$47B revenue run-rate cited, $50B raise in talks |
| OpenAI | AI lab | No S-1 filed (preparing) | Late 2026-2027 | LOW-MED — no S-1; $852B (Mar 2026), ~$25B annualized rev, ~$14B projected 2026 loss |
| Databricks | Data infra | No S-1; CEO calls 2026 "a terrible year to go public" | H2 2026 filing → late-2026/early-2027 list (~41% 2027 odds) | MED-HIGH (private) — $134B, $5.4B annualized rev +65% YoY, FCF-positive, NRR >140%; ONLY profitable name in the wave |
| Cerebras | AI chips | Already IPO'd ~mid-May 2026 | Public | HIGH — full S-1; $510M 2025 rev, $24.6B backlog, ~$48.8B FD valuation; already covered in-vault |
| Canva | Design SaaS | No S-1 | 2027 listing confirmed | LOW-MED (private) — ~$42B |
| Stripe | Fintech | No filing | 2026+ / unclear | LOW (private) — ~$90-159B secondary, profitable |
| Discord | Social/dev | Confidential S-1 reported | Q2-Q3 2026 | LOW-MED — confidential only; ~$15B |
- Reader-pull signal (Benzinga viewer poll, "most-anticipated 2026 IPO excluding SpaceX"): Anthropic 63%, OpenAI 24%, Databricks 7%, Canva 6% (benzinga.com). Anthropic has overtaken OpenAI in investor preference.
- SpaceX is the largest IPO in history (~$75B raise, $1.75T valuation) and is pricing/trading right now (Capital.com, CNBC, TECHi).
- Anthropic confidentially filed its draft S-1 on June 1, 2026 — the first major AI lab to formally start the process, beating OpenAI (Yahoo Finance, Univest).
- Databricks is deliberately waiting for a quieter window so SpaceX/Anthropic/OpenAI don't absorb the ~$200B of IPO capital first; it is the only FCF-positive name in the cohort (TNW, TechTimes).
- The whole pipeline is described as a "$3 trillion AI IPO race" with 25+ venture-backed names targeting listings, most in H2 (CNN, Value Add VC).
Convergences and contradictions
- Convergence: Vault and web agree Cerebras is the cleanest already-public IC-memo template (full S-1, signal-dense). The vault's "Tickered vs scarce" framing is independently confirmed by the web pattern — the 2025 venture IPO cohort underperformed while pre-IPO names command scarcity premiums.
- Contradiction (resolved): One AI-funding-tracker source still lists Cerebras as "pricing May 14," but the vault S-1 doc (2026-05-12) plus the Moonshots episode and the convergent web pricing data confirm Cerebras has already IPO'd (~mid-May, ~$48.8B FD). Treat Cerebras as public.
- Timing nuance: SpaceX is not "upcoming" — it is pricing/trading as of today (6/12). A SpaceX memo is now a recently-IPO'd memo, not a pre-IPO one. That changes the angle (lessons from the largest IPO ever) but does NOT kill it; SpaceX remains the highest-pull subject.
Synthesis for RDCO
The lead-magnet thesis is sound and the timing is unusually good: a genuine IPO supercycle (~$3T across the pipeline) is unfolding in real time, and Sanity Check's data/AI audience is the exact reader who wants a credible, jargon-light "what does the S-1 actually say" read. The format is already proven in-vault (the Cerebras S-1 breakdown). The differentiating editorial hook — what keeps these from being generic IPO explainers and clears the no-derivative bar — is the Tickered-vs-scarce frame: each memo catches a name in its scarce, pre-Tickered moment and reads the filing like a junior analyst prepping an IC. The angle is "here's what the screener won't tell you once this gets sortable," not "here's a company going public."
Ranked top IC-memo subjects (reader-pull x data-availability x audience-fit):
- SpaceX — highest reader pull by a wide margin (the headline event), full public S-1 = maximum data availability. Weakness: lower data/AI-audience fit (it's space, not data infra) and it just listed, so the angle must pivot to "anatomy of the largest IPO in history + the xAI consolidation read." Ship it FIRST anyway, riding the news wave — it is the open board task and the traffic magnet that earns subscribers for the rest of the series.
- Anthropic — best blend: 63% reader pull (highest in the post-SpaceX poll), strongest audience fit (it's the AI lab Sanity Check readers actually use), and a confidential S-1 now exists. Data caveat: the S-1 is confidential, so a memo must be built from the public Series H-1 mark ($965B), reported revenue run-rate, and secondary-market data — frame it as a "pre-S-1-reveal preview," then refresh when the public S-1 drops (~Oct). This is the keystone of the series.
- Databricks — strongest analytical memo subject because it's the only profitable name in the wave and lands squarely in Sanity Check's data-infrastructure wheelhouse. Lower raw pull (7% poll) but highest audience fit and a rich private-data story (FCF-positive, NRR >140%, $5.4B rev). No S-1 yet, so it's a "private-data IC memo" — defensible because the metrics are well-reported; flag the no-S-1 caveat explicitly.
- Cerebras (recap/comp) — already public with a full S-1 and already analyzed in-vault. Best used as the benchmark/comp memo the series references rather than a standalone headline, or as a "how the first AI-wave IPO is trading" follow-up.
- OpenAI — huge pull but currently too data-poor for a credible IC memo: no S-1, contested financials, and the biggest losses in the cohort. Hold until it files; a memo now would be speculation, not analysis.
Too data-poor to support a credible memo yet: OpenAI (no S-1), Canva (private, 2027, thin public data), Stripe (private, no filing), Discord (confidential only, small). These are watch-list candidates for future quarters, not Q1 subjects.
Proposed quarterly cadence (one IC memo per quarter, refreshable on filing events): Q3 2026 — SpaceX (ship now, ride the listing). Q4 2026 — Anthropic (preview memo now off the confidential filing, full memo when the public S-1 drops ~Oct). Q1 2027 — Databricks (private-data memo, timed to its expected H2-2026 filing / late-2026 list). Q2 2027 — OpenAI or Canva, whichever has filed by then (event-driven). Each memo doubles as instrumentation for the [[2026-05-12-innermost-loop-ai-infrastructure]] thesis, tying the content and investing clusters. The series brand could be "The IC Memo" or "Before It Gets Tickered."
Open follow-ups
- Did SpaceX actually price/trade on schedule today (6/12), and at what valuation? The memo's opening data needs the confirmed first-day numbers, not the $135 target.
- Does the founder want the series gated behind email (lead magnet proper) or run as free issues with an upgrade CTA? Affects format (PDF one-pager vs newsletter section).
- Is there an RDCO position-disclosure obligation if a memo covers a name in an active paper-trade thesis (e.g., anything in the Innermost Loop basket)? Set a standing "not investment advice + any RDCO position" disclosure footer.
- Should the Cerebras vault memo be repackaged as the series' first published artifact (it's already written, just internal), to test the format with low marginal cost before the SpaceX build?
- When Anthropic's public S-1 drops (~Oct), schedule a memo-refresh trigger off the EDGAR filing.
Related
- [[2026-05-12-mostlymetrics-cerebras-ipo-s1-breakdown]]
- [[2026-04-30-not-boring-scarce-assets-abundance-driven-scarcity]]
- [[2026-05-12-innermost-loop-ai-infrastructure]]
- [[2026-02-24-ship30for30-paid-newsletter-checklist]]
- [[feedback_no_derivative_sanity_check_pieces]]
Sources
- Vault:
~/rdco-vault/06-reference/2026-05-12-mostlymetrics-cerebras-ipo-s1-breakdown.md - Vault:
~/rdco-vault/06-reference/2026-04-30-not-boring-scarce-assets-abundance-driven-scarcity.md - Vault:
~/rdco-vault/01-projects/investing/theses/2026-05-12-innermost-loop-ai-infrastructure.md - Vault:
~/rdco-vault/06-reference/2026-02-24-ship30for30-paid-newsletter-checklist.md - Web: SpaceX IPO timeline/pricing — https://capital.com/en-int/learn/ipo/spacex-ipo ; https://www.cnbc.com/2026/05/20/spacex-ipo-live-updates.html ; https://www.techi.com/spacex-ipo/
- Web: Anthropic confidential S-1 (6/1/2026) — https://finance.yahoo.com/markets/stocks/articles/anthropic-files-confidential-1-joins-161008569.html ; https://univest.in/blogs/anthropic-ipo-confidential-s1-sec-filing-june-2026-five-things
- Web: Most-anticipated post-SpaceX IPO poll — https://www.benzinga.com/markets/ipos/26/06/52975380/anthropic-openai-databricks-or-canva-most-anticipated-ipo-after-spacex
- Web: Databricks waiting / only profitable name — https://thenextweb.com/news/databricks-ceo-calls-2026-a-terrible-year-to-go-public-as-spacex-anthropic-and-openai-prepare-to-absorb-200-billion-in-ipo-capital ; https://www.techtimes.com/articles/318244/20260611/databricks-only-profitable-name-ai-ipo-wave-outlier-that-makes-money.htm
- Web: $3T AI IPO race / pipeline overview — https://www.cnn.com/2026/06/09/tech/openai-ipo-anthropic-wall-street ; https://aifundingtracker.com/ai-ipo-tracker/ ; https://valueaddvc.com/blog/ipo-pipeline-2026-every-company-expected-to-go-public-this-year-and-current-status