Full Report

How to Read Duolingo

Duolingo is a consumer subscription app dressed as an edtech company: a freemium mobile product that converts about 9 of every 100 monthly active users into paying subscribers, then monetizes the rest with programmatic ads. The economic engine is operating leverage on a fixed content and engineering base — not content, brand, or AI per se — and the metric that matters above all others is the DAU/MAU ratio, because engagement is what converts to subs and compounds word-of-mouth growth. The market is likely overestimating near-term AI threat and underestimating how rare this cost structure is: 72% gross margins, 35% FCF margins, over $1B in net cash, and zero reliance on paid acquisition to grow users.

1. How This Business Actually Works

The revenue engine runs on two rails: a paid subscription (Super Duolingo / Duolingo Max) and ads on free-user sessions, with a small third rail from the Duolingo English Test and in-app purchases. In FY2025, subscriptions were 84% of revenue and growing 44%; "Other" (ads + DET + IAP) was 16% and growing 17%.

The cost structure is what makes this work. Cost of revenue is about 28% of revenue and consists almost entirely of app-store take-rates, hosting, and a rising sliver of generative-AI inference cost. R&D is the biggest expense at 30% of revenue — this is a software shop, not a content shop. Sales and marketing is just 12% of revenue, which is the single most telling number on the page: consumer-subscription peers routinely spend 40 to 60 percent.

Where incremental profit actually comes from: every marginal MAU costs almost nothing to serve — hosting is cheap, content is a fixed investment — and roughly 9% of them eventually pay ~$70–85/year gross. The remaining 91% see ads that monetize at a small but non-zero CPM times billions of daily lessons. This is why revenue grew 39% while operating income more than doubled — classic fixed-cost leverage on a viral distribution model. Bargaining power is asymmetric: Apple and Google are the only meaningful counterparties, and they extract roughly 15 to 30 percent of gross subscription dollars depending on tier and tenure. That is the single largest structural margin drag, and it will not change on its own.

2. The Playing Field

Duolingo is the only profitable, growing, cash-generating public edtech at scale. The peer set is instructive by contrast: Chegg is being rerouted out of business by generative AI, Coursera is stuck in enterprise/consumer limbo with no operating profit, Nerdy shrinks every year, and Udemy barely scrapes breakeven. The "competitive landscape" narrative you will hear — Babbel, Busuu, Rosetta Stone — is mostly private companies at a fraction of the scale. Real competition for user time is YouTube, TikTok, and mobile games.

What the peer set reveals: scale is necessary but not sufficient in this industry. Udemy has roughly Duolingo's revenue and is barely profitable because its unit economics are built on third-party instructors (rev share), paid marketing, and enterprise sales. Coursera has the same revenue as Duolingo and loses money because degrees-and-certificates require partner institutions (more rev share) and enterprise sales cycles. Duolingo is the only operator in the cohort that has found the consumer-subscription model that works in education: own the content, gamify the habit, let virality do the distribution, and monetize with app-store subs. The peer set also reveals what is not a moat in this industry: content libraries, university partnerships, and AI models.

3. Is This Business Cyclical?

In a normal recession, Duolingo is fairly defensive: learners stay on the free tier, sub churn may tick up, but the app is a $7–13/month habit that competes with streaming subscriptions, not discretionary luxury. The real cycle risk is elsewhere: app-store fee regimes, FX, ad-market weakness, and AI platform shifts. Subscription revenue is reported in USD but collected globally; a stronger dollar can drag reported growth several points. Ad revenue (about 12% of total) cycles with digital ad markets. And the Duolingo English Test is tied to international student mobility — it softens in a visa crackdown and strengthens on reopened borders.

The 2022–2023 period is the only real stress test we have. Duolingo kept compounding MAUs through the 2022 ad downturn, grew subs through the 2023 rate shock, and accelerated through the 2024–25 mobile-ads weakness that crushed most consumer internet companies. The sharpest reminder that the market can reprice this name quickly was Q3 FY2025: the stock fell roughly 25% on lighter-than-expected guidance as management telegraphed a deliberate shift toward long-term investment over near-term monetization. That is the cycle in this name — sentiment and multiple, not the cash flows.

4. The Metrics That Actually Matter

Forget the income statement for a moment. Five numbers tell you whether this business is winning or breaking.

Why these over the usual superficial ratios: P/E is nearly meaningless here because FY25 net income includes a $257M one-time tax benefit from releasing the US deferred-tax-asset valuation allowance. GAAP margins will appear to drop next year when that tailwind rolls off. EV/Revenue is too blunt because it cannot distinguish "consumer sub with 35% FCF margins" from "enterprise SaaS with 20% churn." The DAU/MAU ratio is where the story lives: it climbed from roughly 28% pre-IPO to 39.6% in four years. That is not a normal trajectory for any consumer app at this scale — it is the specific signature of a habit-forming product.

5. What I'd Tell a Young Analyst

The trap on this name is letting the narrative run ahead of the math in either direction. Bulls frame it as "edtech TAM $123B" — irrelevant, because Duolingo is not competing with a HolonIQ slide. Bears frame it as "AI will commoditize language learning" — possibly true in five years, definitely not true in two, and the filings show AI is more weapon than threat so far (Video Call, Roleplay, content generation at scale). The things to watch are narrow and specific.

Watch these quarterly, in this order:

1. DAU/MAU ratio. If it breaks 40% and keeps climbing, the subscription story has another multi-year leg. If it plateaus near 40% while MAU growth slows, the pipeline starts to clog.
2. Paid-sub net adds versus MAU growth. If subs outgrow MAUs by two-to-one or better, the freemium funnel is tightening — bullish. If MAUs outgrow subs, management is farming awareness at the expense of monetization. Q3 FY25 was the first hint of the latter, which is why the stock fell 25%.
3. S&M as a percent of revenue. Every point above the current 12% means the viral flywheel is losing efficiency and paid acquisition is filling the gap. Cleanest single tell.
4. Stock-based compensation. $137M in FY25 — about 13% of revenue. Reported FCF is ~$360M; SBC-adjusted is closer to $220M. Anyone quoting a 7% FCF yield is ignoring the dilution.
5. DET and adjacency traction. Math, Music, Chess, and the Duolingo English Test are the call option. If any one reaches 5% of revenue, the TAM re-rates. If none does in three years, this is a single-product company at a multiple that assumes otherwise.

What would change the thesis:

• Apple/Google mandated fee reduction (US, EU, or global): instantly adds 5 to 8 points of operating margin with no product change required. Upside-only catalyst.
• A public, peer-reviewed demonstration that a general-purpose chat assistant beats Duolingo on measured learning outcomes — not vibes. This has not happened; filings cite studies pointing the other way. Worth monitoring independent efficacy research, not social media takes.
• Sub-penetration stalling near 9–10% while DAU growth continues — a "running harder to stand still" signal that compresses the multiple fast.

What I'd ignore:

• Analyst downgrades during guide-downs. The business has beaten quarterly consensus in most quarters since the IPO, and the one guide-down in Q3 FY25 was a deliberate investment-cycle reset, not a demand break. Cycle sentiment, not cash flows.
• Breathless coverage of new AI features. The learning flywheel is the moat; AI is the feature set. Don't mistake the two.

The Numbers

Duolingo compounded from a money-losing consumer app in FY2023 to a \$1.0B-revenue, \$370M-free-cash-flow business in FY2025 — 38.7% revenue growth, 72% gross margins, a \$1.0B net-cash balance, and almost no capex. Yet between November 2025 and March 2026 the equity lost roughly 80% of its value as management signaled a deliberate margin pause to reaccelerate daily active users, and analysts cut targets from \$500-plus to \$85–\$110. The single metric most likely to rerate or derate the stock from here is operating margin in FY2026 — whether the DAU pivot is a one-year tax on profitability or a permanent reset of the margin ceiling the market had underwritten.

Snapshot

Revenue and earnings power

Revenue nearly doubled in two years while operating income went from a \$13M loss to a \$136M profit. FY2025 net income of \$414M is flattered by a \$232M net tax benefit booked in Q3 2025 (release of valuation allowance on deferred tax assets) — strip that one-time and normalized earnings are closer to \$180M, still a 2x jump over FY2024.

Gross margin is essentially flat in the low-70s — the classic consumer-subscription profile where incremental revenue drops directly to gross profit. The entire margin story is operating leverage: R&D and marketing grew slower than revenue, and operating margin expanded from -2.5% to 13.1% in two years. That is the curve the Feb 2026 guide cracked — management now expects FY2026 margins to pause rather than continue climbing.

Quarterly trajectory

Sequential revenue additions peaked at \$21M in 1Q25 and decelerated to \$11M by 4Q25 — the first visible sign that topline is maturing. Operating margin held above 12% for three straight quarters into 4Q25, which is why the subsequent FY2026 guide-down (implying margins fall back toward 9–10%) caught the Street off-guard.

Cash generation

Capex is trivial (\$18M in FY2025, under 2% of revenue) so operating cash flow and free cash flow are effectively the same line. The harder question is how much of reported profit is real cash — in FY2023 and FY2024 the company generated far more cash than GAAP earnings; in FY2025 the relationship inverts because of the one-time tax benefit. On a trailing 3-year basis, cumulative FCF (\$794M) is more than 1.5x cumulative net income (\$519M) — a strong signal that cash generation is legitimate, not accrual-led.

Stock-based compensation ran at 13.2% of revenue in FY2025 — high for a company at this scale. SBC-adjusted FCF (what an owner actually keeps after dilution offsets) is \$232M, giving a cash yield closer to 4.8% on the \$4.82B market cap rather than the 7.7% headline FCF yield. Buybacks (\$13M in FY2025) are not close to offsetting the dilution.

Balance sheet — all cash, no debt

Net cash of roughly \$1.05B is about 22% of the current market cap. Debt is \$94M of finance-lease obligations — essentially leased office space, not funded debt. There is no interest coverage concern; the only capital-allocation lever on the balance sheet is how aggressively management chooses to use the cash pile (very little buyback activity to date).

Returns on capital

ROE hit 30.7% in FY2025, though most of the leap reflects the same deferred-tax release that inflated net income. Underlying ROE (normalized earnings ÷ equity) is closer to 15% — still healthy, but not the three-handle print the reported numbers suggest.

Capital allocation — nothing returned yet

No dividend. Buybacks have been ~\$15M per year — covering less than one-tenth of stock-based compensation. The \$1.1B cash pile is sitting idle. With the stock down 80% from its high, the absence of an activist repurchase program is a capital-allocation tell: management is either preserving dry powder for M&A (recent NextBeat team-buy hint) or remains unconvinced the drawdown is complete.

Valuation — then and now

Multiple compression has been the entire story. Year-end EV/EBITDA went from 353x in FY2023 (when earnings were minimal) to 119x in FY2024 to 36x at end-FY2025 — and with the post-guide sell-off, the current-price EV/EBITDA is closer to 25x. On a normalized-earnings P/E basis the stock trades near 26x, the cheapest it has ever been as a public company. A 7.7% headline FCF yield — or 4.8% after subtracting SBC dilution — is the best entry multiple the business has offered.

Peer comparison

Duolingo is the only peer in this set that is both growing and profitable. UDMY grew 0.4%; COUR grew 9% but burned \$77M at the operating line; CHGG is in a full-blown collapse with revenue down 39% and margins deeply negative; NRDY is shrinking with a negative-35% operating margin. The premium Duolingo has historically commanded on multiples is entirely earned by operating fundamentals — the gap is not subtle. What the market is questioning is not whether Duolingo is better than these peers, but whether Duolingo can stay a 30%+ grower without giving back the margin it just built.

Fair-value range

Sell-side consensus has collapsed from a \$222 average target in late 2025 to roughly \$100–\$110 on the most recent post-guide revisions (JPM \$92, Goldman \$105, Barclays \$110, DA Davidson \$85, Needham \$145). At \$105, the stock is within the fair-value band of the base case but with the skew asymmetric to the downside if the pivot fails to produce DAU reacceleration in the next two prints.

What the numbers say

Confirm: the operating business is an anomaly in its peer group — 38.7% revenue growth on a \$1B base, 35.6% FCF margin, no debt, \$1.05B of net cash, and returns on capital that moved from barely-positive to genuinely high in 24 months. The fundamentals are not impaired.

Contradict: the "structurally profitable category compounder" framing that justified a \$27B enterprise value in mid-2025 was always underwritten on margins expanding monotonically. The FY2026 guide proved margins are a management choice, not an inevitability — which means the premium multiple was.

Watch: Q1 2026 DAU growth (May 4 print) and operating margin. If DAUs reaccelerate back above 35% YoY while op margin prints near 10%, the pivot is working and the base-to-bull case takes hold. If DAU growth stays at the high-20s and margin slides below 8%, the bear case dominates and the stock re-tests the \$88 52-week low.

The People

Governance grade: B+. Founder-led, founder-aligned, and founder-controlled — Luis von Ahn and Severin Hacker still run the company they built in 2011 and together hold 76% of the voting power through Class B super-voting stock, so the board can advise but cannot overrule them. Executive pay is refreshingly simple (salary plus RSUs; founders get only salary plus a decade-long performance PSU), the co-founders have massive skin in the game, and insider trading is disciplined under 10b5-1 plans. The real concerns are structural rather than behavioral: a dual-class structure that neutralizes public shareholders, a CFO who was just recruited out of her own audit committee chair seat, and a steady drumbeat of insider selling by CTO Severin Hacker that has run in the tens of millions of dollars over the past nine months.

The People Running This Company

The operational spine is narrow: two co-founders who have been partners since Carnegie Mellon, a brand-new CFO pulled off the board, and a product/engineering bench drawn largely from ex-Google leaders. The bench is stable but thin at the C-suite — succession for either founder would be a genuine event.

What They Get Paid

Pay is almost entirely equity for working executives and entirely performance-based for the founders. The CEO has taken the same $750,000 cash salary every year since 2021 — the CEO pay ratio of 2.6× the median employee is among the lowest in the S&P peer universe. Where shareholders should focus instead is the mechanical enrichment from the founder PSU grant, which has already paid out roughly $115M to von Ahn alone.

The headline distortion is in the Pay-Versus-Performance table. The SEC's "Compensation Actually Paid" calculation — which marks equity awards to year-end value — shows von Ahn's 2025 CAP at negative $44M because stock-price-hurdle PSUs granted at IPO lost paper value as the stock fell from its May 2025 peak. In 2023 and 2024 the same calculation produced $140M and $129M of "CAP" for the CEO. The accounting number swings violently with the stock; the cash reality is that von Ahn has earned $767,500 a year in cash every year since the IPO.

Pay is earned. What looks like an occasional megapay spike is pure mark-to-market on the founder PSU — a grant the Board sized specifically so that founders receive less than 1% of each marginal dollar of new stockholder value created above the IPO price.

Are They Aligned?

Yes, economically — but on their own terms. The dual-class structure means the founders control the company. The question is whether they behave like owners, and the answer from the disclosed data is mostly yes.

Ownership and Control

Class B common stock carries 20 votes per share and converts 1:1 into Class A. The co-founders together command 76% of the vote with far less of the economic interest. Every Class B share is held by insiders; public holders vote Class A only. Practical consequence: shareholder proposals, activists, and even large institutions have negligible leverage at the ballot box. Baillie Gifford at 12.1% of the float controls just 2.9% of the vote.

Insider Behavior — Hacker Sells, von Ahn Holds, Shelton Buys

The insider-activity picture is the single most telling signal on this tab.

Three observations matter:

1. Hacker is selling mechanically and heavily. Through a 10b5-1 plan running since mid-2025 he has been converting ~10,000 Class B shares into Class A every 1–2 weeks, at prices ranging from $344 down to $172, for more than $20M of aggregate proceeds. The trades are pre-scheduled and compliant, but the cadence is consistent even as the stock collapsed — he is not waiting for better prices. His remaining Class B position is still massive (~3.06M shares), so this reads as diversification, not a statement on fundamentals. It is still the kind of activity that should raise eyebrows during a period when the share price has more than halved.
2. Von Ahn has not sold a share in the open market. His Form 4 activity is limited to PSU settlements and the accompanying tax-withholding at vesting. He exercised 32,000 legacy options in 2025 (proceeds retained in shares), and his year-end direct position was ~7.3M Class B. For a CEO watching his paper net worth drop by hundreds of millions of dollars in a single quarter, that is a meaningful vote of confidence.
3. Director Jim Shelton bought 5,000 shares in the open market at $99.76 on March 3, 2026 — a $499K check. Open-market director purchases at Duolingo have been rare and the timing, right after the February selloff, is pointed.

Dilution, Options, and the Founder PSU

At IPO the Board granted von Ahn 1.2M PSUs and Hacker 600K PSUs, tied to ten stock-price hurdles from $127.50 to $816 over 10 years. Eight of ten tranches have already been hit. That's impressive in one sense — the stock did reach $525 in May 2025 — but it means the founders are now sitting on substantial earned equity, while tranches 9 and 10 (requiring $612 and $816) will almost certainly forfeit unless Duolingo trades near all-time highs again before July 2031. The company's anti-hedging and anti-pledging policies apply to these shares; none are disclosed as pledged.

Outside the founder PSU, the share-count overhang is modest: 859K options outstanding at a weighted $18.38 strike, plus 12.5M shares available under the 2021 Plan and 1.6M under the ESPP. Management additionally has a $400M buyback program authorized in 2026, the first meaningful capital-return step post-IPO.

Related-Party Behavior

The proxy discloses no related-party transactions above the $120,000 Item 404(a) threshold other than standard indemnification agreements. There are no undisclosed loans to insiders, no real-estate deals, and no family members in the C-suite. Within this tab, related-party risk is effectively absent — which is not the same as saying there is no conflict risk. The Munson CFO transition is technically a related-person-type event (a director taking an executive role) but is not a Regulation S-K Item 404 transaction.

Skin-in-the-Game Score

An 8 out of 10. The co-founders have hundreds of millions of dollars of net worth locked in founder-era stock, von Ahn has not trimmed, and pay is overwhelmingly contingent on long-term stock performance. It isn't a 10 because (a) Hacker's mechanical 10b5-1 diversification cadence is aggressive, and (b) the dual-class voting structure tilts the alignment in favor of the two people at the top, not all shareholders equally.

Board Quality

Duolingo's nine-person board is 7-of-9 independent (the two dependents are the founders) and the board is classified with staggered three-year terms, which dilutes shareholder ability to enact change quickly. Committees are staffed entirely with independent directors and the audit chair is an SEC-designated "financial expert." The bench skews toward consumer-tech operators and VC partners who have seen many cycles.

Where the board is strong: consumer tech operators and VCs (Bohutinsky, Clemens, Gordon, Lilly, Schlosser) and gaming/engagement design (Ross, Gordon) — both directly relevant to a gamified consumer subscription business. Jim Shelton brings actual education-policy depth that is unusual at this caliber of board and genuinely useful given Duolingo's brand positioning as a mission-driven education company.

Where it is thin: no one on the board has public-company CFO experience who is not also in management now that Munson moved off the board. Sara Clemens is the audit financial expert but her background is operating (COO of Twitch and Pandora), not finance. For a company with $1B+ in bookings that is now shifting into heavier AI capex and a $400M buyback, adding a career CFO or capital-allocation specialist to the board would strengthen the oversight layer.

The Verdict

Strongest positives. Founders have 76% of the vote and behave in a way that is broadly shareholder-aligned — von Ahn has not sold a share, pay is mostly contingent, and the founder PSU structure has already extracted genuine value creation rather than showering the founders indiscriminately. Executive pay is simple (salary + RSUs) and the CEO pay ratio is 2.6×. The clawback, anti-hedging, and anti-pledging policies are all in place. Seven of nine directors are independent.

Real concerns. Dual-class voting permanently neutralizes outside holders. CTO Hacker has sold more than $20M of stock over nine months on a 10b5-1 plan that has not paused as the stock halved. The CFO seat is held by a director recruited off the audit committee chair — governance hygiene will be tested when her first earnings cycle begins. A plaintiff's-bar investigation (Faruqi & Faruqi, announced April 21, 2026) is looking into potential securities claims tied to the AI-spend disclosures and Q3/Q4 guidance cuts — routine at this stage but a reminder that disclosure discipline is not optional. And the board lacks a current-generation public-company CFO as an independent voice now that Munson has crossed the aisle.

What would move the grade.

• Upgrade trigger (to A−): adding an independent director with public-company CFO or capital-allocation credentials; a meaningful pause or slowdown in Hacker's 10b5-1 selling cadence; execution of the $400M buyback at current prices.
• Downgrade trigger (to B− or below): the Faruqi investigation maturing into a certified securities class action with merit; von Ahn beginning to sell open-market stock; any material accounting revision during Munson's first two quarters; or the 2026 say-on-pay vote landing below 80% support.

The Full Story

Between Q1 2023 and Q4 2025, Duolingo went from a gamified language app selling discipline to investors, to an AI-first learning platform willing to sacrifice near-term growth for a long-term platform bet. Management beat every full-year guide it issued from FY2023 through FY2025 — the biggest by an order of magnitude in FY2024 — then, at the same shareholder meeting that reported a record FY2025, cut FY2026 bookings growth guidance from "nearly 20%" (what they said they could do) to 10–12% and stepped margin back ~450bps. The story now is credible but fragile: a founder who tells investors he is spending ~$50M of bookings to rebuild the free experience around Video Call and AI tutoring, after a spring 2025 "AI-first" memo backlash that management never mentioned in writing to shareholders. Credibility is high on delivery, medium on communication.

1. The Narrative Arc

The arc has three acts. Act 1 (2023): a scrappy AI-forward operator with perfect execution — user growth accelerates past 60% YoY, bookings up 45%, margins tripling. Act 2 (2024): the execution engine keeps compounding; AI moves from slogan to shipped product (Max, Video Call); the first subtle flags — a DAU guide-down and AI cost headwind — appear but get absorbed by beats. Act 3 (2025–2026): the strategy gets re-underwritten. An April 2025 "AI-first" internal memo detonates on LinkedIn and never appears in an IR letter; the product teams dial back the edgy marketing voice; and management formally sacrifices short-term bookings to reinvest in the free experience.

2. What Management Emphasized — and Then Stopped Emphasizing

Because a flat Heatmap only renders a single period, the matrix below is the full series. The pattern is what matters, not each cell.

Themes that rose: AI went from feature-level in Q1 2023 to existential in Q4 2025. Subject expansion went from nonexistent pre-Q3 2023 to the primary new-user engine by Q4 2025 (Chess launched 2025 and reportedly grew faster than any prior subject). Teaching efficacy and learning science — the "Duolingo Method" — quietly moved from defensive talking point to strategic centerpiece in Q3–Q4 2025.

Themes that faded: The pre-IPO founder equity hurdles language (the "$178.50 60-day VWAP" bullet) appeared every letter in 2023 and disappeared after the first hurdle was hit. Cost discipline framing — the defining message of 2023 when Duolingo was proving it could be profitable — got progressively de-emphasized once margins reached 25%+. Family plan callouts, a Q1 2024 centerpiece, quietly dropped by 2025.

The quiet pivot: starting Q3 2025, the word "teaching" replaces "monetization" as the controlling verb. Management explicitly says it shifted A/B test weights toward "teaching better and user growth" rather than subscription conversion. This is the first philosophical re-underwrite in three years of letters.

3. Risk Evolution

Three risk-evolution stories actually matter.

First, AI went from absent to dominant in 24 months. The FY2021 10-K has zero AI-specific risk factor. FY2022 adds a one-paragraph regulatory caveat. FY2023 adds a full product-risk factor explicitly naming generative AI, Duolingo Max, frontier AI launch partnerships, and "hallucinatory behavior." FY2024–FY2025 expand it further with the EU AI Act effective date (Aug 1, 2024), Colorado, Utah, and Texas state AI acts, and the Dec 11, 2025 federal executive order. The competition risk factor was silently edited in FY2023 to insert a parenthetical — "a new product could gain rapid scale at the expense of existing brands through harnessing a new technology (such as generative AI)" — that persists through FY2025. That one parenthetical is the single cleanest signal of how management's threat model changed.

Second, Apple concentration got worse, not better. Apple App Store share of revenue moved from 50% (FY2021) to 62% (FY2025). The risk-factor language is almost verbatim between the two filings, so the heatmap shows a flat 5. But the underlying exposure rose materially — the risk is the same, the magnitude is higher, and management never called it out quantitatively in prose.

Third, the profitability frame flipped. FY2021: "we have incurred operating losses each year since our inception." FY2023 onward: "we have recently achieved profitability, which we may be unable to sustain." This is a genuine de-risking. Transitional IPO risks — lock-up, limited operating history, freely-tradable-shares language — evaporated across the five filings, while the dual-class/controlled-company paragraph remained completely unchanged. Governance has not relaxed as the company matured.

Smaller shifts worth flagging: the Beijing-office risk line ("over 30 employees in Beijing makes it easier for the Chinese authorities to bring enforcement actions") appeared only in FY2023; FY2025 replaces it with softer but broader language about "foreign-affiliated technology companies." The FY2021 AWS outage ("over 5 hours") grows to ("over 12 hours") in FY2025, reflecting an actual October 2025 incident. COVID and Brexit disappear entirely.

4. How They Handled Bad News

For a 12-letter dataset, there is almost no "bad news" to handle. Every FY2023 and FY2024 letter is a beat-and-raise. The notable stress tests are three.

The FY2023 Q4 DAU guide-down (Feb 2024). Management preemptively guided Q1 2024 DAU growth down to "mid-50s" versus the 60%+ trend. Rather than defending the prior number, CEO Luis von Ahn framed it as arithmetic: "we can't expect our user growth to accelerate forever." Actual Q1 2024 DAU growth came in at 54%, matching the guide. Clean handle, no drama.

The April 2025 "AI-first" email. On April 28, 2025, Luis von Ahn sent an all-hands email — then posted it to LinkedIn — announcing Duolingo would become an "AI-first" company, including phasing out contractors and permitting teams to hire new humans only if the work "could not be automated through AI." The backlash was immediate and external: "AI first means people last"; "I can't support a company that replaces humans with AI." A week later, von Ahn walked it back: AI is "a tool to accelerate what we do, at the same or better level of quality" and "I do not see AI replacing what our employees do (we are, in fact, continuing to hire at the same speed as before)." In the Financial Times he later conceded: "I did not expect the amount of blowback."

The Q4 FY2025 strategic step-down (Feb 2026). This is the biggest tonal shift. Management guided FY2026 bookings growth to 10–12% after FY2025 delivered 33%, and cut the Adjusted EBITDA margin target by roughly 450bps. Von Ahn's framing: "we expect bookings growth of around 11% in 2026, compared to the nearly 20% we believe we could achieve if we operated like we have in past years"; "I don't take this decision lightly, and I know it may come as a surprise to some investors." That parenthetical — the first time in 12 letters he explicitly preempts investor pushback — is the key tell. The rationale: invest in the free experience (Video Call moves from Max-only to Super; $50M of bookings friction removed) to reach 100M DAUs by 2028.

5. Guidance Track Record

Three years, three big beats. The FY2023 Adjusted EBITDA margin beat (~17.6% vs. 4.5% initial midpoint) is almost hard to believe — but it reflects how drastically operating leverage kicked in once paid-sub penetration rose, and it established the credibility that carried management through the next two years. The FY2024 bookings beat was ~$70M above the initial guide midpoint; FY2025 was ~$70M again.

Management also raised full-year guidance within year in every quarter across FY2023–FY2025. The cadence was: set conservative at Q4, raise at Q1, raise at Q2, raise at Q3, beat again at Q4. Every cycle.

Credibility score: 8.5/10. Three years of consecutive full-year beats on both the top line and margin, with transparent in-year raises, is unusually clean. The deduction is entirely on communication: the FY2026 downshift was proactive, but shareholders learned about the "AI-first" memo from LinkedIn rather than the IR letters, and the contractor wind-down — a reputationally significant workforce decision — was never acknowledged to investors in writing. Delivery is top-decile; narrative discipline has cracks.

6. What the Story Is Now

The old story was simple: a freemium language-learning app with a beloved mascot, a gamification engine, organic word-of-mouth marketing, and enough engineering culture to be a profitable business at scale. Every piece of that is still true and has been de-risked. Revenue over $1B, nearly 50M DAUs, 12M+ paid subscribers, 29.5% adjusted EBITDA margin, ~$305M of adjusted EBITDA — these numbers were a fantasy in 2021.

The new story is bigger and stretchier. Management is saying Duolingo is no longer just a language app but a multi-subject learning platform (languages, Math, Music, Chess) that plans to use frontier generative AI to become meaningfully more effective than a human tutor, and is willing to take a 450bp margin hit and 20 points of growth off the FY2026 guide to fund that transition. The target is 100M DAUs by 2028 — roughly 2x today. Video Call, the AI conversation tutor, goes from a premium-tier feature to free-tier bait. Content generation ("over 150 new courses via AI in Q1 2025") is supposed to be the wedge that turns language dominance into subject-wide dominance.

What to believe. Execution is real. The beats are real. AI variable costs did come in lower than feared (GM pressure of 100bps vs. the 300bps management warned about at the start of FY2025). The pivot to teaching efficacy in A/B testing is directionally right if you believe personal-tutor-quality AI is coming — and it gives Duolingo a bigger TAM if it works.

What to discount. The 100M DAU target is a management bet, not a plan with intermediate milestones. Apple concentration at 62% of revenue is a quiet structural tax that never gets addressed in prose. The "AI-first" memo incident shows Luis von Ahn can misjudge public communication — and the decision to never address it in an IR letter means the next communication misstep is likely to compound. Subject expansion beyond languages is still unmonetized ("we do not expect these new subjects to meaningfully contribute to monetization in the near-term" — Q3 2023; still true in Q4 2025). Founder dual-class voting control has not softened.

The honest read: this is a management team that has earned the benefit of the doubt on delivery but not on communication. Hold them to the 100M DAU-by-2028 number, watch the Q1 and Q2 FY2026 DAU prints for signs that free-tier Video Call actually moves top-of-funnel, and assume the next surprise will come from outside the shareholder letter — because that is where the last one came from.

What's Next

The calendar is carried by a single print. Duolingo reports Q1 FY26 on Monday, May 4, 2026, after the close — the first tangible read since February's guide-down reset bookings growth to ~11% and Adjusted EBITDA margin to 25%. Sell-side has already rebased: FY26 EPS consensus fell from $7.91 sixty days ago to $7.07 today, with five downward revisions against one upward in the last 30 days. The bar is lower, but DAU growth and the Q2 guide are what will move the tape.

What the market will most closely watch on May 4:

1. DAU growth YoY — must stay above ~30% to keep the "flywheel intact" read alive; a high-20s% print revives the April 2025 AI-first memo damage thesis and opens the downside target.
2. Q2 FY26 bookings guide — whether management holds, lifts, or lowers the ~11% FY anchor. A lower guide on Q2 is the single fastest route to a new low.
3. New CFO Gillian Munson's capital-allocation posture — any buyback expansion to offset $137M of annual SBC would directly answer the biggest bear point.
4. The long-term margin target — re-affirmation or formal retirement of the 30–35% Adj EBITDA target. Silence here is itself a data point.
5. Foregone-bookings update — directional read on whether the $50M free-tier investment is producing the DAU reacceleration management promised.

For / Against / My View

For

Price target: $175 · Timeline: 12–18 months

Primary catalyst: Q1 FY26 earnings on May 4, 2026 — a DAU reacceleration above 35% YoY validates the voluntary re-basing narrative against sell-side targets already collapsed to $85–$110 (JPM $92, Goldman $105, Barclays $110, DA Davidson $85). Disconfirming signal: S&M above 15% of revenue OR paid-sub penetration stalls below 9.2% for two consecutive quarters.

Against

Downside target: $65 · Timeline: 9–15 months

Primary trigger: a Q1 or Q2 FY26 DAU print that stays in the high-20s% YoY while operating margin slides below 10%, confirming the $50M "foregone bookings" investment is not producing DAU reacceleration. Covering signal: Q1 or Q2 FY26 DAU growth back above 35% YoY with S&M below 15% — the pair proves the founder "take the long view" framing was correct and the 100M-by-2028 target is intact.

The Tensions

1. The February 2026 guide: voluntary investment, or structural reveal?

Bull says the guide-down is a founder voluntarily trading one year of margin for a larger terminal DAU base — the same "take the long view" move von Ahn wrote at IPO. Bear says management itself just proved the margin curve was a dial, not a structure, which is permanent multiple damage regardless of how well the investment works. Both cite the same February 2026 guide: ~11% bookings growth and ~25% Adj EBITDA margin against the previously stated 30–35% long-term target. This resolves on the May 4 print — a DAU growth number above 35% YoY with steady S&M vindicates the Bull's "investment" framing; a high-20s% DAU print with margin drifting toward 10% confirms the Bear's "structural" read.

2. DAU/MAU at 39.6% and 30% DAU growth: widening flywheel, or last clean reading before deceleration?

Bull reads 30% DAU growth against 14% MAU growth as a habit-formation signature almost nothing at this scale reproduces. Bear reads the same 30% as a number already impaired by the April 2025 AI-first memo — the company itself admitted in the FY25 10-K that the memo contributed to "a deceleration in user growth" — and says the next two prints will show the deceleration compounding, not reversing. Both cite the FY25 DAU trajectory and the 10-K's own AI-memo disclosure. This resolves on the second derivative of the May 4 and early-August DAU prints: reacceleration vindicates the flywheel; two sequential quarters below 30% breaks it.

3. $1.05B of net cash against $137M of SBC and $12.6M of buybacks: optionality, or idle capital the board won't deploy?

Bull points to a fortress balance sheet at 22% of market cap as optionality at the exact moment the stock is on sale — the cash is a put under the thesis. Bear points to the same cash pile as a board that has chosen not to offset dilution at 80% off the high, with SBC running at 13% of revenue and buybacks covering under 10% of it. Both cite the identical balance-sheet and capital-allocation facts. This resolves on the May 4 call — any expansion of the buyback program or a direct capital-return commitment from the new CFO settles it for the Bull; a restatement of the existing posture settles it for the Bear.

My View

Close call, with a slight edge to the bear ahead of May 4. The Bear's first tension — that management themselves just walked the 30–35% margin target down, and the February letter language ("I know it may come as a surprise") implies this is larger than the "long view" framing admits — is the heaviest single fact in the name, and the Bull's own target methodology tacitly concedes it by anchoring on normalized FY27 rather than FY26. Against that, the Bull's balance-sheet and founder-alignment points are real but not yet monetized — the board has not acted on the dilution gap even at 80% off the high, and "optionality" without deployment is a story, not a result. I'd pass here rather than position either way: the name flips from "price the reset" to "price the flywheel" on a single DAU number two weeks out, and starting now is paying for conviction the print hasn't earned. The one condition that flips the view: a Q1 FY26 DAU growth print above 35% YoY with management reaffirming the 100M-by-2028 target and any buyback expansion — that resolves two of the three tensions in the Bull's favor and makes $175 a mechanical outcome off a bombed-out sell-side base.

Web Research

The Bottom Line from the Web

The filings show a business that grew revenue 38% in FY2025 to $1.04B, hit 50M DAU, and generated record free cash flow. The internet shows a stock that lost roughly 80% of its value from a May 2025 peak above $500 to about $100 today, a CFO who quit after six years, a securities class-action investigation now actively recruiting plaintiffs, and insiders who have sold into strength at every tier above $170. The single most important web finding: the disconnect between fundamentals (still strong, still growing) and sentiment (broken, litigation-adjacent, AI-disruption-flagged) — that gap is the entire DUOL trade today.

What Matters Most

1. 80%+ drawdown from May 2025 peak — largest stock story of 2026

2. Securities class-action investigation is active and recruiting plaintiffs

3. CFO Matt Skaruppa resigned after 6 years — replaced by a board member

4. Insider trading is ~100% sell-side — but one director just bought

5. Analyst downgrade cascade in late Feb / early Mar 2026

6. AI disruption is the structural bear case — Jefferies flagged "replicability"

7. CEO social-media backlash over AI-first memo + "Duo Death" campaign

8. $400M buyback authorized alongside weak guidance (Feb 26, 2026)

9. UK Home Office English-test bid surfaces (Apr 21–22, 2026)

10. Free advanced content rollout — strategy in action (Apr 22, 2026)

Recent News Timeline

Interpretation: the timeline splits cleanly in two. Up to May 2025, the news is all beat-and-raise. From November 2025 onward, every High significance event except the BofA upgrade is negative — three earnings/guidance shocks, a CFO exit, two analyst downgrades, and a securities investigation.

What the Specialists Asked

Insider Spotlight

Luis von Ahn — Co-Founder, CEO, Chairman

Guatemalan-American computer scientist; invented reCAPTCHA (sold to Google, 2009). MacArthur Fellow. Co-founded Duolingo 2011 at Carnegie Mellon. 14.7-year tenure. 2024 cash comp $767K; founder-equity-award market value exceeded $213M at year-end 2024, tied to 10 stock-price hurdles over the decade. Directly owns ~7.14% (Simply Wall St). High-visibility CEO (Reddit AMAs, Fast Company, CNBC) — which cuts both ways given the April 2025 AI-contractor incident.

Severin Hacker — Co-Founder, CTO, Director (10%+ owner)

Swiss-born; Duolingo's other co-founder. Dual role: CTO and board director. Owns >10% of the company (Section 16 filer). Sold 10,000 shares per week through Oct–Nov 2025 on a 10b5-1 plan, monetizing ~$15M+ into the stock's rally. Per secform4.com his post-sale stake is still ~3.1M shares. The sales were programmatic, not indicative of urgency, but the scale is large.

Matt Skaruppa — CFO (outgoing, Feb 23, 2026 handoff)

Six-year tenure, took company public. 2024 comp $5.58M (+44% YoY). His Feb 17, 2026 sale of 5,856 shares at $112.42 was disclosed as a tax-withholding sale linked to RSU vesting. Will remain in an advisory role after handoff.

Gillian Munson — Incoming CFO (effective Feb 23, 2026)

Former CFO of Vimeo, Iora Health, XO Group. Leadership roles at Union Square Ventures, Allen & Company, Symbol Technologies, Morgan Stanley. Served on Duolingo's board since 2019 as Chair of the Audit, Risk and Compliance Committee. Currently Audit Chair at Phreesia (PHR). Internal candidate — signals continuity, not rupture.

James H. Shelton — Independent Director

Notable because he is the only insider to buy stock in the open market in the post-crash period. Purchased 5,000 shares on March 3, 2026 at an average $99.76 (~$499K), increasing his direct holdings to 9,632 shares. Former US Deputy Secretary of Education (Obama administration); brings education-sector credibility.

The pattern: every sale was executed into the rally above $170. The single purchase was executed under $100. Shelton's buy is the cleanest insider signal — a direct director opening position after the crash — but one trade does not offset five executives selling into weakness.

Industry Context

AI disruption is the dominant narrative overhang

CNBC (Mar 1, 2026) compiled Jefferies' "replicability risk" screen of companies whose products AI chatbots could commoditize — DUOL was explicitly named. The November 2025 earnings cycle reinforced this: CNBC (Nov 11) reported "Wall Street rewards hyperscalers, punishes DoorDash and Duolingo" for AI investment, because the latter are spending on AI rather than selling AI. Morgan Stanley's Feb 27 downgrade framed the same point: AI cost pressure was not fully priced in at the prior Overweight view.

Management's counter-narrative (Nasdaq, Sep 18, 2025; BeyondSPX, Nov 29, 2025): 18× acceleration in course creation via AI (148 new courses in 12 months vs. 100 in the prior 12 years), Duolingo Max at 9% of subscribers with doubled bookings YoY, and Video Call costs falling 10× since launch. The bull case is that AI is an enabler for DUOL's platform advantage; the bear case is that AI erodes the use case entirely.

Consumer subscription valuations have normalized

SaaS Capital (Aug 2025) noted that SaaS valuations in 2025 returned to the "low normal" 2016–2017 band. DUOL's de-rating from high-growth premium to consumer-subscription-normal is consistent with that industry mean reversion, though the additional penalty (-80% vs. -30–40% for the cohort) is DUOL-specific.

Geographic tailwinds: Mandarin, Asia, UK contract

• CNBC (Jan 2025): Mandarin-learning on the app up 216% YoY — driven by pandemic-era reopening and China consumer-app interest.
• Q3 2025 call (Motley Fool Nov 6): Asia (led by China) is fastest-growing region for user acquisition; Duolingo Max being tested in China pending regulatory approvals.
• UK Home Office bid (Yahoo, Apr 22): potential government English-testing contract for the DET segment.

These are genuine green shoots that the filings framed cautiously but the web has covered in more specific terms.

All figures in USD. Web research current as of 2026-04-23. Sources are linked inline; the news timeline table is the canonical reference.