DUOL — Deck
Duolingo is a Pittsburgh-based consumer subscription app that teaches 130M+ monthly users languages, math, music and chess through a gamified freemium model, monetized via a $7–13/month Super tier and ads on free sessions.
The business didn't break — management pressed the margin pedal back up.
- Operationally the best year on record. FY25 revenue grew 38.7% to $1.04B, operating income more than doubled to $136M, and free cash flow hit a record $370M at a 35.6% margin. Nothing in the P&L collapsed.
- Management reset the model on purpose. The February 26, 2026 guide cut FY26 bookings growth from 33% to 10–12% and adjusted EBITDA margin from 29% to ~25% — framed as ~$50M of foregone bookings reinvested into free-tier users to chase 100M DAUs by 2028.
- The multiple carried the entire decline. EV/EBITDA compressed from 353x at year-end 2023 to ~25x today; normalized P/E near 26x is the cheapest DUOL has ever traded. The Street is pricing the reset as permanent, not temporary.
A rare consumer subscription profile — record cash flow, fortress balance sheet, honest dilution drag.
Operating leverage on a 72% gross margin moved op margin from -2.5% to 13.1% in two years on almost no incremental sales spend — S&M stayed at 12% of revenue against peers at 40–60%. Two honest adjustments: FY25 net income of $414M included a $232M one-time deferred-tax release (normalized closer to $180M), and stock-based comp at $137M is roughly 11x the $12.6M buyback — cash yield on an SBC-honest basis is 4.8%, not the 7.7% headline.
Five years of beat-and-raise ended in one shareholder letter.
Before: From the July 2021 IPO through mid-2025, Duolingo compounded MAUs from 42M to 117M, paid subs from 1.7M to 9.5M, and margin from breakeven to 25.7% adjusted EBITDA. Initial revenue guides landed 4% light every year; bookings growth ran ~1,000bps above initial guide every year. Stock ran from $139 to a $544 peak in May 2025.
Pivot: On November 6, 2025 the Q3 print missed Q4 bookings guidance (22% vs consensus 30%) and the stock fell 25% — its largest single-day drop ever. On February 26, 2026 management voluntarily re-based FY26 to 10–12% bookings growth and 25% EBITDA margin, authorized a first-ever $400M buyback, and CEO Luis von Ahn wrote a preemptive investor letter — the first across 12 quarters.
Today: $1B+ of idle cash, a newly expanded buyback, a CFO who was the Audit Committee chair three months ago, and a tape 81% off the high trading at its 52-week low. The question the next chapter answers: does the free-tier investment reaccelerate DAUs, or did the 30–35% margin target quietly die?
Orderly distribution, not capitulation — and insiders sold into every down leg.
- Death cross unbroken since August 26, 2025. 50-day SMA crossed below the 200-day at $316 and has not re-crossed across nine months. Spot $105 sits 52% below the 200-day near $219 — unambiguous downtrend regime.
- The two largest single-day declines on record. -25.5% on November 6, 2025 at 7.2x average volume and -14.0% on February 27, 2026 at 7.8x — both earnings gap-downs. Distribution on expansion, not a panic flush.
- Nine months of one-way insider selling. Every named officer — von Ahn, Hacker, Glance, Meese, Chen — filed repeat sale Form 4s from September 2025 through February 2026. Not a single insider purchase through a 68% drawdown. Von Ahn's 10b5-1 sales plan expired December 15, 2025; subsequent sales are discretionary.
Founder-controlled, founder-aligned, and one oversight seat just collapsed.
- Pay for performance, purely. Von Ahn and Hacker take $750K flat salaries with zero new equity since the 2021 IPO. Their entire wealth depends on a single 10-tranche PSU award running to 2031 with price hurdles from $128 to $816 — 8 of 10 have hit, the top two ($544 and $816) are still live.
- Dual-class super-voting stock. The two co-founders hold 75.8% of the vote while owning ~15% of the economics. Classified board with three-year staggered terms; no lead independent director. Public holders cannot remove a director or amend bylaws.
- The audit chair became the CFO. Gillian Munson — Audit Committee chair since 2019 — resigned the board on February 23, 2026 to replace outgoing CFO Matt Skaruppa. Legally permitted, but it erodes one of the cleanest oversight relationships a public company has, three months before the Q1 reset print.
Lean cautious — wait for the May 4 print; the setup is asymmetric but the signal hasn't landed.
- For. The only profitable, growing consumer edtech in its cohort — 38.7% revenue growth with 13% op margin and 35.6% FCF margin against peers losing money at -5% to -34%. Nothing in the operating P&L justifies an 80% derating.
- For. $1.05B of net cash, zero funded debt, a new $400M buyback, and founder PSUs that only pay on stock recovery. Von Ahn's wealth depends on the tape he is sitting inside.
- Against. The February guide walked the 30–35% long-term EBITDA margin target down to ~25% without formally retiring it, and von Ahn wrote a preemptive letter for the first time in 12 quarters — a tell that management knows the reset is larger than the framing admits.
- Against. Nine months of one-directional insider selling through a 68% drawdown, a death cross unbroken since August, and a CFO transition that collapsed an oversight relationship three months before a reset print. No insider has called a bottom with a check.
Watchlist to re-rate: (1) May 4 Q1 DAU growth vs the 20% management assumption — above 35% validates the pivot. (2) Operating margin — below 10% confirms the permanent reset. (3) Buyback execution pace on the $400M authorization — does management buy $100 stock, or just authorize it?