"Bending Spoons IPO: S1 Breakdown" — @CJ Gustafson
Why this is in the vault
Definitive S1 analysis of the most operationally extreme software acqui-scaler to attempt a public exit — directly maps to RDCO's SaaS roll-up thesis research and the capital cycle investing framework.
⚠️ Sponsorship
Sponsored by Abacum (FP&A tool, NetSuite-stack adjacent). Explicit sponsor block at top of issue: "Mostly metrics is proudly powered by Abacum." CTA links to a HubSpot-tracked demo URL (hubs.li/Q04klk530). No Mostly Talent self-promotion in this issue.
The core argument
What Bending Spoons is: An Italian acqui-scaler (founded Copenhagen 2013, HQ Milan) that buys digital businesses with large loyal userbases, guts them (82% headcount cuts are typical), rebuilds on shared proprietary infrastructure, then lever-and-repeat. Portfolio as of the S1: AOL, Brightcove, Evernote, Eventbrite, Harvest, komoot, Remini, StreamYard, Vimeo, WeTransfer — 500M+ MAUs, 9M+ paying subscribers. They have never sold a single acquisition.
The playbook in three steps: (1) Acquire disciplined on price — target businesses with durable demand that don't need ad spend. (2) Transform aggressively — slash headcount, collapse management layers (4→2), rewrite tech on shared infra, raise prices. (3) Reinvest — combine operating cash flow with fresh debt to fund the next deal.
The financials:
- FY2025 revenue: $1.31B (+95% Y/Y); Q1 2026: $601M (+132% Y/Y); TTM ~$1.6B
- But organic growth was only 13% in FY2025 — 82 of the 95 reported percentage points were acquired revenue
- Gross margin: 66-68%, below pure-software benchmarks due to app store tolls (Apple/Google take 15-30% on ~a third of electronic revenue)
- Blended NRR: 94% (Evernote 99%, AOL 95%, StreamYard 91%, Remini 87%)
- Revenue per employee: $2.57M in FY2025, up from $1.12M in FY2023 — among the highest in software
- GAAP net income FY2025: -$0.2M. Adjusted net income: $376M. The $335M+ gap in add-backs is larger than GAAP operating profit
The debt machine: $4.36B total debt, ~$3.6B net. Interest expense FY2025: $143M (+337% Y/Y). Q1 2026 alone: $93M interest (+382%), annualizing to ~$370M/year. Interest consumed 51% of FY2025 operating income; in Q1 2026, 78%.
The house of adjustments: GAAP operating income $278M → Adjusted $613M, via $151M intangible amortization, $85M transaction costs, $79M reorganization costs (the layoffs). CJ's critique lands hard: reorganization is the business model, not a one-time item. Q1 2026 alone had $76M in reorganization costs, roughly a quarter of that quarter's Adjusted Operating Income.
AI as triple-edged sword: (1) Internal force multiplier — AI-authored PRs went from <10% in Q1 2025 to >90% in Q1 2026 (70% AI-only), directly driving the revenue-per-employee doubling. (2) Portfolio threat — Evernote competes with ChatGPT/Notion; file transfer is becoming a feature not a product. (3) Deal pipeline accelerant — AI pressure on incumbents creates motivated sellers and compresses acquisition multiples.
Red flags CJ surfaces:
- Material weaknesses in accounting controls disclosed to SEC (absent oversight structures, weak journal entry controls, limited SOX/GAAP expertise). S1 statements filed "as restated."
- No long-term contracts — mostly month-to-month consumer subscriptions
- Foreign Private Issuer status — less disclosure, can follow Italian governance norms
- Never divested anything — impairments, not exits, surface bad deals
Valuation: Targeting ~$20B equity / ~$23.6B EV. EV/2025 revenue ~18x, EV/TTM ~15x, EV/2026E CEO-guided adjusted EBITDA ($1.4B) ~14-17x. Ticker: BSP, Nasdaq. Goldman + JPMorgan leading syndicate of 14 banks.
CJ's closing verdict: pricing BSP at $20B is a bet that the operating machine is real, AI keeps widening their productivity edge, and the next 50 acquisitions are as good as the first 50 — before fast-growing interest expense catches earnings. "That's a bet on Luca Ferrari being Henry Singleton."
Mapping against Ray Data Co
Direct thesis hit — SaaS roll-up research. The vault already holds 2026-05-27-agent-first-saas-rollup-unit-economics which cites Bending Spoons by name as the canonical reference for the "buy sticky-but-bloated software, cut 50-75% staff, raise prices, rebuild agent-first" playbook. This S1 breakdown is the primary-source confirmation of that research brief — the unit economics are now public.
Capital cycle / investing thesis. BSP is an interesting capital-cycle signal from the opposite direction of the chip-fab thesis: rather than tracking capital flowing into semiconductor capacity, BSP is aggregating capital flowing into software consolidation at scale. The $4.6B deployed in acquisitions, funded by $4.36B of debt at ~$370M/year interest run rate, is its own capital cycle — peak capacity (50+ acquisitions, $1.6B TTM revenue) against rising cost of capital is exactly the Phase 2→3 transition logic the Markov tracker watches for in other sectors.
AI-operations signal. >90% AI-authored PRs in Q1 2026 with a 500-person company is a real data point for what agent-first operations looks like at scale. RDCO's own COO-agent thesis is the micro version of what BSP's internal infra (Pico/Lumen/Abacus data stack, Minerva CLV model, Juno payments) does at the portfolio level.
Accounting integrity flag. The material weakness disclosures are a useful counterpoint: aggressive roll-up plus aggressive AI-automation plus inadequate controls is a combination that can hide problems for a long time. Relevant risk model for any RDCO investing thesis that includes highly-levered software roll-ups.
Relevance to RDCO investing: Medium. BSP itself is a speculative IPO (high debt, GAAP negative, accounting weaknesses) not an obvious capital-cycle play. More useful as a comps reference and a stress-test of the roll-up thesis than as a direct position candidate.
Related
- [[2026-05-27-agent-first-saas-rollup-unit-economics]] — the roll-up unit economics brief that already cites Bending Spoons; this S1 is the primary-source confirmation
- [[2026-05-28-service-book-rollup-agent-first-conversion]] — micro-rollup variant analysis; BSP is the macro version of the same playbook
- [[2026-05-27-markov-equities-pipeline-spec]] — capital cycle phase tracker; BSP's debt load and interest coverage trajectory is a parallel Phase 2→3 signal worth monitoring
- [[2026-05-31-mostly-metrics-ndr-net-dollar-retention-decline]] — CJ's NDR framework; BSP's 94% blended NRR is a live case study for that thesis