DUOL Q2 2026 - Post-Call Earnings Update

Event: Quartr Q2 2026 event Retrieved: 2026-08-06T11:25:00Z Comparison mode: maximal Scorecard contract: v1

Consensus scorecard

Management and Q&A

• The most consequential product disclosure came unprompted from CEO Luis von Ahn: "another possibility, truthfully, is that we may actually sunset Duolingo Max." He said an answer would come "in the next couple of quarters" and that the company is moving deliberately because "we're trying to figure out how to do this without losing much revenue."
• The reason Max is now in question is a cost collapse von Ahn quantified for the first time. Video Call originally cost "like, $0.30 per call"; it is "now under $0.01 per Video Call", a reduction he attributed "mainly to a move towards open source models" with "no loss in quality". Most new Super Duolingo subscribers now receive Video Call, with existing subscribers to follow "over the next few months".
• Asked by Mark Mahaney of Evercore whether cost is the only gate to giving Video Call to all users, von Ahn named a second constraint the release does not mention: "this is one of the main things that gets people to buy. If suddenly all users have it, there's less incentive to buy, and then we would have to find other reasons to get people to buy."
• CFO Gillian Munson disclosed a bonus plan explicitly excluded from guidance: it "will trigger if Q4 DAU growth is 25% or higher and would be paid out during Q1", at "roughly $10 million in cash, potentially higher if DAU growth is higher". She said it is not in the 2026 guidance because the threshold is currently uncertain.
• Munson gave point estimates beyond the ranges in the release: full-year bookings growth of approximately 11% within the 10%-12% range, revenue growth of roughly 16% within the 15%-18% range, and noted that "at constant FX rates from our last call, the bookings growth rate would be about half a point higher".
• Full-year gross margin guidance was nudged up on the call to "closer to 70%" from the 69% initially expected, which Munson attributed to driving "more AI content into our products, offset by AI cost savings".
• Munson said the adjusted EBITDA margin outlook is now 26.5% against the 25% outlined at the start of the year, with approximately $320m of adjusted EBITDA and "over $375 million of free cash flow this year".
• Third-quarter point estimates were given only on the call: bookings of approximately $307m (+9%), revenue of $302m (+11%), gross margin of 71%, and adjusted EBITDA of roughly $76m at a 25.2% margin.
• Asked by Wyatt Swanson of D.A. Davidson why stronger DAU growth did not lift the bookings guide more, von Ahn gave two reasons: the freemium model means "our users don't monetize immediately... it's going to take a while", and the company is deliberately holding to the box it set at the start of the year, with "the rest of the efforts... in increasing DAUs and in teaching better".
• Von Ahn detailed the free-trial change driving monetization without harming growth: Duolingo is shifting most free trials from seven days to one month. More users take the trial, more convert to payers, and DAU rises too because "as soon as they agree to go on the free trial... we turn off energy, we turn off the ads."
• A new lower-priced tier is in test. Von Ahn named it Super Lite: ad-supported, "about half the price" of Super Duolingo depending on geography, with twice as much energy rather than unlimited. He stressed it is early, a small fraction of subscribers, and "I don't know what's going to happen with Super Lite in the end."
• On the June Streak Revival campaign, von Ahn said "more than 15 million learners revived their streaks" and that these users "are also showing better retention than a typical re-engaged cohort".
• CURR, the company's primary retention metric, is at an all-time high and "has increased by about a percentage point in the last year". Von Ahn stressed the compounding effect: "tiny changes to CURR end up implying pretty large changes in daily active users over time."
• Von Ahn disclosed the pace of iteration: "we have a new version of the app every single week. Every version of the app has approximately 350 changes", and experiments per week "has increased faster than our headcount".
• On marketing, von Ahn said Duolingo gets "more than a billion impressions per quarter on our own social media accounts" and that "something like 2/3 of our total social media impressions come from influencers", working particularly well in China, Indonesia and India. He conceded a reversal on performance marketing: "There was a while where I was massively allergic to performance marketing", now used as a complement.
• Munson said top of funnel "got better" and that "in almost every region, the rate of growth there improved in top of the funnel in the quarter". Von Ahn added that DAU growth is "really broad-based... all regions are growing faster than they were before", with Asia fastest and U.S. growth up.
• Von Ahn conceded a strategic error on Math: "I thought that we could get the average person on the street to get addicted to learning Math. I was wrong." The strategy is now aimed at K-12, with the user base "mainly under 18", though selling to schools is "not our goal". Math and Music each have "single-digit millions of DAUs" against "60-some million" group DAUs, so von Ahn said "even a 50% growth here doesn't contribute all that much". Music updates are deferred "maybe a quarter or two".
• On advertising revenue, von Ahn said Duolingo "will remain a subscription business" for the foreseeable future but sees "a pretty large opportunity", noting that two years ago "our entire investment in ad revenue was half a person" against a professional team today, and that improvements should show "over the next few quarters".
• Asked by Ygal Arounian of Wedbush whether the company is pivoting toward monetization, von Ahn called the observation "perceptive", said there was "a little bit of a reset" at the start of the year against monetization that was counter to DAU growth, and that the company is now "focusing a little more on it than when compared to six months ago" while keeping the 100 million DAU target for 2028.
• Von Ahn disclosed a China regulatory constraint absent from the release: "in China, we simply cannot use the AI models, the kind of U.S. AI models. We have to use local models. That's by law... All of our usage of AI uses Chinese models." He said China is the second-largest market by DAU, is expected to become the largest "in a year or two", and "monetizes about as well as France".
• Asked by Arvind Ramnani of Truist to size AI spend, Munson gave figures not in the release: AI expenses in cost of goods sold are "in the tens of millions of dollars", with hosting the other major COGS item, and internal AI use "closer to the $10 million range".
• Von Ahn set out the open-weight thesis: open source options are "maybe three to six months" behind frontier labs for most tasks and "the quality's indistinguishable for many applications", so "if I had a magic wand, I would try to move everything to an open-weight model". He expects the model portfolio to weight "a little more towards open-weight models than it is today".
• On attention spans, von Ahn acknowledged a structural challenge — a Duolingo lesson is "maybe two minutes" against social media used "in, like, 10-second bursts" — and said the company is actively experimenting with session length but has "not yet released anything".
• Munson explained the implied fourth-quarter acceleration as unchanged phasing rather than new optimism: "the year is playing out the way we thought. We've always thought that acceleration is going to come over time. This is not a slow and steady, that's not the way this place is."

Guidance and KPI clarification

• A Q4 DAU bonus plan triggering at 25% or higher year-over-year growth, worth roughly $10m in cash and payable in Q1 2027, is explicitly excluded from 2026 guidance.

Updated neutral analysis

• The call is materially more informative than the release, and on balance more positive. The release showed a thin 1.0% revenue beat and net income down 26% year over year; the call reframed the quarter around 23% DAU growth that accelerated and came in ahead of plan, record retention, and a gross margin and EBITDA outlook both nudged up.
• The central tension the call exposes is deliberate: management is holding revenue and bookings inside a self-imposed box while spending the upside on user growth. Von Ahn was explicit that higher DAUs will convert to revenue but "it's going to take a while", and that the rest of the effort is going into DAU growth and teaching quality. An investor who wants monetization now is being told to wait until 2028.
• The Video Call unit cost falling from roughly $0.30 to under $0.01, a reduction of more than 96%, is the most economically significant disclosure on the call. It converts the company's most expensive feature into something distributable at scale and is the direct cause of the raised gross margin and adjusted EBITDA outlooks.
• That same cost collapse destabilizes the pricing architecture. Duolingo Max exists largely because Video Call was expensive; with the cost gone, von Ahn openly floated sunsetting the tier. He also identified the trap: Video Call is "one of the main things that gets people to buy", so giving it away removes a purchase incentive the company has not yet replaced.
• Super Lite is the other half of that repricing question — an ad-supported tier at roughly half the Super price. Management stressed repeatedly it is only a test, but Max potentially disappearing at the top while a cheaper tier is tested at the bottom describes a subscription ladder in flux, with the revenue-neutrality of the outcome asserted rather than demonstrated.
• The monetization levers management is willing to pull are those aligned with growth rather than against it. The seven-day to one-month free trial extension is the clearest example: it raises trial take-up, converts more payers, and lifts DAU simultaneously because energy limits and ads switch off. That is a genuine change from a history of monetizing by adding friction.
• Retention is the strongest number on the call. CURR at an all-time high and up about a point year over year compounds into DAU, is broad-based across regions and user types, and is attributed to product quality rather than a single feature or campaign.
• The Streak Revival campaign brought back more than 15 million learners with better-than-typical re-engagement retention, but it is a one-time June campaign. Some portion of the 23% DAU acceleration is therefore non-recurring, and management did not size that portion.
• The bonus plan is a useful, if indirect, signal of internal expectations: the board set a Q4 DAU threshold at 25% against 23% delivered in Q2, and management judged the outcome uncertain enough to exclude a roughly $10m cost from guidance. It reads as a genuine stretch target rather than a formality.
• Adjacency growth is being managed down in expectation. Math and Music at single-digit millions of DAUs cannot move a 60-million-DAU base, and von Ahn conceded outright that his original consumer-math thesis was wrong. The pivot to a K-12 user base is a narrower, more defensible position but a smaller one.
• The open-weight model strategy is a structural margin tailwind management is executing rather than merely describing, with the Video Call cost curve as evidence. It carries concentration and quality risk that von Ahn addressed only by asserting the quality gap is three to six months and indistinguishable for Duolingo's use cases.
• China is a growing dependency and a disclosed regulatory constraint. It is the second-largest DAU market, expected to become the largest within one to two years, monetizes comparably to France, and by law runs entirely on Chinese AI models. Von Ahn was candid that government conduct is "outside of our control".
• The advertising opportunity is real but early and explicitly secondary. Moving from half a person to a professional team is a low base, and management guided only to "some improvements... over the next few quarters" while affirming Duolingo remains a subscription business.

Market context and limitations

• The Quartr transcript for this event (134 paragraphs, complete and non-live) was read in full, including all Q&A from D.A. Davidson, Citizens JMP, JPMorgan, Needham & Company, Wolfe Research, Evercore, Wedbush, KeyBanc Capital Markets, and Truist Securities.
• The consensus scorecard is carried forward verbatim from the event's preliminary report (reports/preliminary/2026-08-05-duol-q2-2026.md) and was not recalculated. Its actuals come from Quartr press release 3671707 and shareholder letter 3671708, and its estimates from the matched API Ninjas row.
• The transcript contains an internal inconsistency in the CFO's adjusted EBITDA margin figures. Munson first said the target was increased "to 26.5% from the 25% we outlined at the start of the year", then two sentences later cited "a margin of roughly 25.5% I just mentioned". Later in Q&A she described the change as "going up about a point and a half" from 25%, which reconciles to 26.5%. The 26.5% figure is used in the guidance table on that basis; both statements are reported as made and neither is corrected in the quoted material.
• Von Ahn also gave contradictory statements on session length within one answer, saying the company is "trying to make the minimum session length longer" and then "If you make the minimum session length shorter, that's good". Neither is reproduced as a directional claim above; only the fact that session length is under active experimentation with nothing released is reported.
• All point estimates for full-year and third-quarter bookings, revenue, gross margin, adjusted EBITDA and free cash flow were given verbally on the call as modelling aids. The formal guidance remains the 10%-12% bookings and 15%-18% revenue ranges.
• Prior-guidance figures in the table are management's own characterizations on the call ("from the 25% we outlined at the start of the year", "as compared to the 69% we initially expected"); no prior guidance document was consulted for this update.
• Adjusted EBITDA, adjusted EBITDA margin, bookings, free cash flow, and CURR are company-defined measures. CURR is not reconciled to any reported figure and no absolute value was disclosed, only that it is at an all-time high and up about a percentage point year over year.
• Daily active user counts, retention figures, Streak Revival participation, Video Call unit costs, social media impressions, influencer share, and AI cost ranges are management's own stated figures and are not independently verifiable from the release.
• Von Ahn described group DAUs only as "60-some million"; no precise figure was given on the call and none is inferred here.
• The AI cost figures are ranges offered in response to a request for a "rough estimate" and were explicitly framed by the questioner as not requiring precision.
• No market reaction is assessed. The call was held after the U.S. close on 2026-08-05 and no price snapshot was retrieved for this update.
• The Form 10-Q (3671709) attached to this event was not read for this post-call update.

Source links

• Quartr Q2 2026 event transcript
• Quartr Q2 2026 earnings press release
• Quartr Q2 2026 shareholder letter
• DUOL Q2 2026 preliminary report