Duolingo (DUOL) 9 May 232023 Q1 Earnings call transcript

Good afternoon and welcome to Duolingo’s first quarter earnings webcast. My name is Debbie Belevan, Head of IR. Today after market close, we released our year-end -- our quarter-end shareholder letter with our Q1 results and commentary, which you can find on our IR website at investors.duolingo.com.

On today's call, we'll have Luis von Ahn, our Co-Founder and CEO; and Matt Skaruppa, our CFO. They'll begin with some brief remarks before opening the call to questions. note event please being And Instructions] that is this recorded. [Operator

and forward-looking reminder are future risks that a statements subject we'll material which Just performance, make regarding financial events to uncertainties. and

of obligation been as as set of of information filings we of new we based today, the with believe risks a are no Some These statements assumptions forth and our forward-looking events. these to SEC. on future the statements risk in to result update factors these have that or be have reasonable

all Additionally, we'll for on performance. considered measures intended present our you both non-GAAP measures substitute financial consider And be superior or GAAP are when today's measures and results. analyzing in to a encourage GAAP non-GAAP our call. from, to isolation These we not to

over I'll to that, with turn Luis. And it

and Thank user everyone. off kicked We year Hello, proud I'm Debbie, top welcome. growth, that line cash had you, to great we report profitability with free quarter. strong flow. another this and results,

In our user all our to nicely. XX% exceeded the first increased $XX.X with growing markets major quarter, expectations. year-over-year million, DAUs growth

user helps growth, subscribers, are mission make education or million up at end. universally deliver year-over-year MAUs and quarter on it the course, create paying the Our strong best helps to in But available. $X.X to it us world of increase also us our XX% which X% of

led us thanks subscriber bookings revenue and post and growth profitability and to continued this highest ever. our to And This financial quarter year-over-year, our user respectively. climbing XX% discipline, XX% saw

outlook year. we're you to Matt profitability result line for this through shortly. this and walk outperformance, a top going our of As our updated raising is guidance

to longer proud DAUs staying why And that learning MAU grew the thing than of compared before ratio about all-time numbers number of DAU XX% an improve, a nearly XX% hardest a year to I'm million. language days engagement with motivated. is high a to I've $XX our to streak our about ago. continue said reaching with many times That's X that the

our Our products innovating goal is they to mainly we and and back that are continually so fun day, do every them that improving to keep learners and effective. coming by

letter This generative on our competitive of our the our unique existing complement like advantages beloved in recent AI teaching. moat, our advances data brand shareholder quarter's well focuses and how way

As companies believed Computer chances accomplish a can humans also of applying the to computers incredible this been best together taking I passion. former in fortunate I've synergy long things, with and Science the the benefit advances professor, always feel that to good very among advantage the working of has my be AI. greater and rapid

last which recall, is subscription announced you'll GPT-X. higher-tier Max, our Duolingo As offering, new by powered we quarter,

few the are That to this this the that did we that product have We our are we proud companies consumer-facing technology. live of talent with of a speaks quickly so team. launched one

last us Also model our freemium data competitors amounts our better. to that quarter, make provides enormous at we everyone use with that and products business keeps of my grow to enables bay reminded organically, shareholder letter us

features on generative Max fractions great and iterate AI in Because how a This of rapidly. small our general is works use how test in we're of able users base, our particular. of user model to we'll large and example

the all very continue out. emphasize should the and see Max we're rolling We'll I stages tell you you As that update progress the making about over can I'm about early time. Duolingo AI but to excited in Max, possibilities with of we're I still

Matt. turn And with to it that, over I'll

was We loss the year we impressive million a results, $XX.X our $XX.X highest our quarterly on EBITDA of posted of had of loss net constant ago recap net a quarter. growth a which adjusted delivered margin. EBITDA we XX% to XX% compared in in To $X.X Luis. million, million first the XX.X% which was And adjusted Thanks, basis. year-over-year, bookings quarter, about currency a

the start year, quarterly on our XX%. about also about Based margin highest free the We year and flow we to cash our feel outlook. QX of strong full had very good

are we $XXX $XXX XXXX, to for of million EBITDA $XXX guidance issuing bookings, to million from are revenue at to XX%. strong growing $XXX in of total we $XXX an million our raising million total to saw to midpoint $XXX and of XX% we million to updating QX, margin growth, million guidance are growing midpoint. which our year-over-year range in XX% XX%, to revenue at in bookings, in revenue, and at an XXXX, year margin XX%. strong adjusted EBITDA XX% total QX $XXX line about the For Because continued full and the in top adjusted incremental And trends reflects XX% are million at bookings to we guiding the $XXX million we margin of with the

about the The at guidance combination adjusted discipline we raising of good strong operating margin top and line midpoint. why EBITDA expenses feel growth on our is continued

improvement. million was year In item showing from points QX shifted points of showing about the QX, show know we I S&M of to S&M improvement R&D roughly X.X G&A improvement a and about X.X of operating into of X of there leverage percentage of with improvement, would non-GAAP for year-over-year showing this $X.X but expect that line show as revenue, X about S&M the each point time of expense about points line QX. spend

EBITDA in EBITDA, XX% seasonality to guiding the of an are year adjusted to expect adjusted we we XX% As the for QX. for to the margin rest

of hire will our in costs typically margin will QX QX, then the that largest a expect And new in step-up and material lower QX, QX quarter in We because quarter don't Our QX. revenue quarter. the is expand be margin and strongest than between start. portion that's the which our

Luis about in Max, the rolling As are out. mentioned, it excited still days we Duolingo but early are of we

included not any or this bookings tier. in have amount guidance of for We yet material revenue higher new our

coming the progress on will in the months. We keep posted you

diluted we mentioned to call, close the employees. year as X% dilution And end last approximately the the expect quarter ended to on year shares fully we end using we Finally, from issued XX.X about price. equity million the with with outstanding

back to that, I'll Luis. with turn And it

to Thank you, to amazing quarter. commitment just And team helped thank and take another Matt. whose this our us opportunity excellent I collective excellence deliver to passion want

take as to happy we And they're it queue. long be now Debbie your good. the as to to manage questions would I'll turn back

right. Instructions] All first our comes Mark question Thanks, of Evercore. Luis. [Operator So Mahaney from

good. this Well, I hope question is

to... haven't we talked So

It better be, Matt.

of Math. We my mentioned topics, -- which is yet you haven't one talked favorite haven't

about So you're in on just seeing in could you terms talk interest the early that? what of

wane, of talk stronger. or about more bit just I little you China So wax market, it's been -- but waxing think for could and then you, the too? no waning a that's And kind becoming

and, that's So a if much terms how contributor at of to all, subs? of paid your your you MAUs in been

for thank Well, questions, Mark. you the

Of course, one my also Math favorite is of topics.

learn months As of and you entirely ago. and well. really still It's a should very early growing to it's course, still know, small it's organically, few the very this also we app doing It's very a an Math Math mention, is days -- launched growing team. I nicely. a

more say And still just using on features know basically there's there's more and don't that actually we that. soon, very we're need doing content have loving add content. advanced else So it. to if other be a I going to list anything working long in of I'm it, to than we're

that. So with we're very happy

very one nicely, China. also X% our In I with of we're to and I China, It's happy in probably but though, still with fastest-growing of our happy small people, China should market a of we're our Mark. well, X% but it It's we mean, is is revenue, us. our terms fastest-growing countries, for doing it. be to growing take, progress you, yes, give and very Thank or it's countries. of one remind seem very

comes of KeyBanc. Justin from question Patterson Next

Great. Two, if I can.

first big and a monetization in or the seen [indiscernible] companies one. of First, where lot don't a question. impede picture Luis, a product in the monetization impede some a talk that just in be of the engagement. building the just guardrails about necessarily monetization would consumer have Could dating, initiatives way around So bit, you We've in experience? gaming, you

in investors alluded of think how as participate, can over beyond thinking so future? Duolingo just remarks, moving since apps even the educational to set would the prepared broadening then you just about your math should more And over that of Whether that it something leveraging AI time. you're love that's

So even just building deeper, more immersive experiences existing and apps.

be learning to casual from mastery a competency. really experience moving So a what might of gaining up

questions, Great Justin.

a trying you to trying usually it's each or I usually lot we make better, quarter. improve In of to trying of the cetera. way well terms run tests. et app it's of it's just better A/B the monetization engaging guardrails, how let every goal. me The -- either how to monetize A/B we something has to A/B product. so more a running it's we for is tests trying And to make test, our hundreds we're or mean product describe develop It teach improve we

addition in the to also we guardrail A/B goal, for test metrics. run, every these we have Now

if do bigger That's The is like button would may that. to the or silly this to something example, a where make just be for an we goal of like of the test So monetization, A/B that example. would improve goal just monetization. subscribe be we improve something --

at it does look to also what We engagement. our

test, more experiments the at guardrail et that whether our are or And that do not up So look metrics. look on only learning We launch less time are spending from app. or more at we A/B we people cetera. people less mess

experiment. would that on launch we So that spending in the subscribe this button would less it case, people so the make app, are if making time not bigger,

we from remind for good -- what single able and engagement, been thing things most come our worked well monetization, out all engagement us, That's grow important worked having impact is you, an that, without the on we've believe up. -- for doing just years, And our this to which because internally, user user actually which user we're kinds growth to doing. of to and it's be out really it's

subscribers free years I Duolingo. they'll mean, to user, to paying a even next have convert a chance the we if it's continue them into several user, grow coming when we because over the

So about this that. is think how we

I'm AI, but scalable. about the we're human excited extremely launched has can the something goal personally without In you terms tutors very teach well that tutor and one-on-one been a one-on-one human AI. we expensive always of Duolingo, one-on-one human make Since because tutor, are as extremely not to as excited.

we've be And to goal. been AI to from beginning the the able using been do that's that. So

our use way we how right every we've teach. we that day. is And well exercise our exercises users at users to X and billion right give that try all Now we've in looked historically improve getting about in time. we're use point, that to the -- to particular, And which the by this solve used AI we solved at the at the users main exercises

right the exactly So we sure get try and all make to we from our use data exercise, users. you that the

We've that built called very a that. model used we've -- sophisticated Birdbrain can do

what's AI. got Now months, out, us of to came that last And since since writer come over about access developing having amazing a to really we've just it. quickly. few this that good it, early had have thing is we features And the ever started ever generative on we allows for pretty language it's can good It really it staff. like a amazing

us develop has allowed features. to So new that

X For developed example, called new offering of new features our Duolingo as higher-tier part we Max. subscription

be you in a is human to Explain Roleplay, one can called can other understandable where France mistake, language basically like croissant you to a which basically have the explains you or called One something. of ordering a And it be pretend you My gives -- explanation. is pretend role or whenever them something play Answer,

so gives you conversational practice. And that

So, other to us to That's we've human. to allow a to of far, try teach into AI areas. use so also, decided course, to going get to closer

making, So ABC our our be going investing better. app, Math invest -- app, example, we better; are to other for in for example, Duolingo

content cheaper. We're to to using also create be going AI faster and

is that part content where are with live, online, but we lot people the and there's where part is faster is AI, the that the just also that there's a cheaper. off-line of So interacting generate

and which engaging, case, teach costs. there's a massive just really apps we important to to our our have lower also better, in be So opportunity is more think make

there, jump to in Justin. just And Yes.

the our would weapon not-so-secret I is one think think I additional make that Luis point on I this. is

work I teach you technology is. he PhD this his You been really can thesis learn he dream a type can wrote humans mean have better it's this and of tell by excited of really on could his answer there way. in better how so to how I his to together. computers mean

a moment in this. us it's So that That's pretty he's We're time. just great. through lucky leading

And your from next of Schackart comes William question Ralph Blair.

which is and again DAU reacceleration in the on yet impressive. strong curious MAU really is just saw I'm to quarter, further

curious once really that call out Just if there's anything there growth you'd driving again. strong

question, different pricing I been you've And in know markets. international testing second then

users to seeing different trends and Just curious pricing are how what of plans? there you're sort responding

keep the of DAU MAU, the and getting the keeps In getting Duolingo more our the more case learning right, language we engaging. product, and you're app main product,

now. Our MAU keeps getting DAU XX% better ratio at to better, we're and

reason more running And that engaging. is for and the that because more it. make our more tests really That's keep A/B just we product

types prominent intuitive. feature our to We more their things of we that just better so make street And so get the social more we that make friends or our features users come back. do,

our we a And it is it does area. our growth for have just growth high-performing So area, very in general, grows us, what DAUs.

looking at going and to course, get because And in better is this nobody of our ratio company MAUs. to better MAU DAU is

we we are day is to to that learn somebody getting way language. the a grow because to right We it's come but DAUs MAUs, trying So report our every everything, and single believe them nice.

coming learn a a can't You language once by month.

keeps ratio expect would better. better. that and And to better So continue that getting I getting

Your next about is regional pricing. question

So same was everywhere price not -- we in a the the of knew everyone, IPO-ed to couple when that, of We, price. had most years just course, optimal we that ago, the world. remind

in we sense Over world, GDP the have every in in the point, most correlated with prices pretty in that of at years, most that this the last the couple world, for country. makes And the price regions in country. it's countries in tried

countries an U.S. that increase bookings. our wasn't Europe, has of increase because or prices it, the was nice. should happened increase prices already in with has the in -- the prices been So implied similar. course, wealthy life-changing There the sense. like I We took say but an remained in it the what pretty make was,

what went poorer down little did similar, got they is whereas bit, quite that remained went in to up Sometimes buy. And a the lot a but prices pretty countries, people you they significantly. more

poor countries subscriptions to are not payment got just even you are don't as remember, have sometimes But mature or these where digital people methods.

prices for these years that as in years, the So this poor an next monetizing we to we well of monetizing the as over have monetize more in understand to decrease But the step countries as things or few happen saw order the countries. us them. over we're also important in wealthy

Boone Andrew of question JMP Next Securities. comes from

true? that assume platform. the the on advanced of attracting seeing of Max. that One What in is to you adopting are a are would learners type terms more I Max learner Max? Is that actually

LOMs change for then How this And secondly, broadly. do about more there? mind just a think at think the competition through of broader top you comes you Duolingo? as question concerned What that? in on Are you all

Thank Andrew. you,

Okay.

GPT-X so I that should the super happy we terms Max, In say, of Max that day -- we're exactly launched announced. was we're

that happy So we're super fast. . out there put we that very

And is is like to start small like giving our screen our all features more reason users. Duolingo. the just of redesign on we Now only more or here it the put that's small a us Max. of start fraction and give do of Max year with We or to home that we it is product development That like users. it we and to That takes then what's or because family gave how didn't at complex development. with groups it anything, plan a do when optimize about there, particularly course, and users We product usually small to our it a our we how for of users that fraction then out we users. for happening

anything end Max But good, know up. we where demand a of also with There's like really say give, So happy features. can for lot what higher example, actual is just because is there that is the will we're demand can't metrics this at this for tier, I for or this don't for point, we that. which

things -- people hard whether or to It would know. learners, to or they're I of rather doing. advanced, not, we more more probably just attracting now committed it's They about know but that answer are question number it's buying It's the than And know don't to wealthier advanced is is your learners say is may they're is what interested. advanced honest but whether it. this hard more the very be more a

we early kill say also to when are Matt models if say our should me Max you not is I preparing. that because should Max were ourselves not I in too It your would we did not including include model. tell. because in

with thing Max. the that's So

this competition, language not just we're for There's something models particularly the moats and With worried is us. about. enough

great generic that whole language large they thing the very have example, worldwide they For And on are web. you train allow -- this models information. this

And it would what write to kind worldwide say. it's web something, ask worldwide the web the -- of like average you it's they're whenever what the say, so basically of would

they about What they're a with, the not learn for data we or have trained trained they're don't not example, how data people with have. language all or

We a of lot data. have

And language of so on AI -- our top do how kind the models for of is we we these stuff. use large we way our own

we moat are our any that a don't than language with people larger So it and data this is, competitors. XXx, us, know, of huge the of XXx all have learning I

one that's so And thing.

we even -- which We without we're actually the the we're any not any much large But tweaks, just larger we distribution we distribution anybody else. tweaks, have doing, language did also them. tweaking have still that models even if if than without apply to

that's big thing. So a

also aggressive characters engaging like owl that. our like and the so fact passive of brand our also, product our that being We and stuff engaging all other and like course, have is moats our

that answers Hopefully, this in something your of particularly question. that not just we're is about competition. terms So worried

of Needham. from Next question Ryan comes MacDonald

topic. piggybacking maybe quarter. Congrats awesome on off just the of that Luis,

it and as of it was that gained not think -- only sort why think is also of of was with sort of free. with so to powerful, sort the it. one adoption learners it of out, so of scaling got sort of of about you and mature And it the benefits you I evolution experiment sort ChatGPT Max As quickly much but

out steep more of So of to Super obviously you the a relative broadly features sort how as you with Max about Duolingo? given the think rolling core fairly Max, functionality sort is pricing, do with increase balance

We've where. excellent question. At going or -- see happy we're had knew Duolingo years, for Yes, to -- belong business. X time, point, the had tier. tier tier. have we're over something last features that's to We that, we the of Super an -- extra adding going over this always free standard features the what higher somewhere we that for belong we've was have the be good pay It's what kind an that

kind a that. we a a And for reason tier the So business. excuse, good GPT-X was good higher of have to add

now have so And these X we tiers.

be figuring each business, And next care us leads going we're that important revenue take where literally feature. free think going our for over important user about out the our to we it's going single good And because but everything, to course, most to do what to best everywhere. and it's we're to to growth years, us years, is we for put of the and mention, stuff is

them things we'll the we sense So that free more see yes, put if in make certain there. tier,

mention, important models to Now which the is to super not that one cheap. language providing mention is thing live particularly access is large I'll

et our GPT-X, pay have them use to OpenAI, in case, for we example, cetera. For we

we tiers So features can that actually that for pay in so we highest now, cost. paid for the keep these

time, that of for the to us certain querying expect tier. and that things, language the large may example, do is other allow significantly, go quite Over down to going models free though, for cost we

these kind to is of that an end going something time, where to just little ongoing on up are of going evolve it's to features it's tell over kind a in early basis. So and

I appreciate lame metrics quarter-over-quarter what I was missed the sense. large if sort it, if missed a the was specific sort you'd Makes line. That of fourth specific subs you first me But -- the feature on that anything you that really jump apologize, apologize And the on question. of but paid That sort nice I is paid in of I subs. this. and jump of color. my would region to that good my saw struck bad was quarter or terms side? there quarter that attribute in I

there's number of Well, a things.

just growing. our launched they keep subscriptions growing I have for there's our that. growing do and we it of just just mean since been for no and to better or -- one making that at just them growing we more single get So improving paid number just it. by the subscription. things. consistently ever more That's better And I reason features and converting we our subscription adding the a of mean, interesting features is And users. basically by

say a you the also merchandising, what we get at by subscribe. you get But advertise when to lot knowing to better to to to subscription,

all more we So us that that more tests A/B -- just gets and enough run to of subscriptions.

main I standard the answer. think That's thing the quarter. just that's that every happening keeps

So jump there's every in, would was been we've jump there I public. been quarter a say, a since --

up Since in a nice trend we've we've seen quarter. every Yes. basically public, conversion been

just So was of in think based what trends strength this our on recent continued said. Luis quarter I conversion kind

question comes at Mario Next Barclays. Lu from

is real The first a TV show? why Love one Language not is

have -- employees real is vote, the right if our would a a business, it is concentrated think, make I or show had vote that. more it but But us the it, of we Listen, main internally, now answer. lot to for half would say on people asked than we're a I

it. Got

vote my well. had as You

first DAUs. one is The on

how we that acceleration? should accelerated main DAUs media track. year? guys is first drivers metric and talked XQ the I this like about like you social guess important think any guys very rest a to know newest you compounding. of it's said quarter. TikTok, promotion, the But You for call out? this in I about anything It And for

Yes.

and the reason. compounding DAUs that. I main because mean, effect better. I getting it we it. our of becoming stickier. product our growing just So is track can keep That's reason the That's keeps it's know and this stickier mean We main that,

getting and are at mentioned better reasons, efficient you example, keep marketing. other marketing. there We Now more for

We have really found

stride terms marketing has really don't. team -- and its found that the our what work are the things that -- of levers in

for For in social is example, we in Twitter, do we us. we but I TikTok. just really is example, well good do media an TikTok have really found it's well really not Instagram. that organic mean

So is media. it's good it's very we just social that for that And found paid organic, we a not brand have stuff.

that works So really well.

found well. have really Asia have -- paid us. also really work is influencers paid influencers cheaper We usually of that And work in really well works this countries. performance Latin in small also -- we and influencers amount in certain well Particularly in work and markets paid, America, found for that some marketing

are So -- work we these has the keep all have things this combined, DAUs. And that of accelerating found our just well.

I expect to our what users of speak But is I think say growth to I'll think Obviously, That's users go keep -- forever. of has thing QX, that one terms thing for, of row let of can the many very a quarters accelerating it's on I'll don't growth. -- that. can't one this don't we been the now. I by that I'll know, user Matt nice In for At our people. accelerating quarters the we say, point, out some in run that how XXX have

we to on can't those point really been good quarters my feel go strong It if still doesn't is accelerate. and really forever, growth, even X going That have since user we be was we've about can it public.

trends year. far just to this and strong, want accelerate, things that we the a in stay like it's not requirement so So we that

this gross terms assume direct more one what apps if big then court go out portion If going direct-to-consumer look just the a of ruling Great. to margins where Apple of There's is I I potentially gross margins. Duolingo's on guess at is Android it and strategy new vs fees. will. in of you we store app to now consumer, kind expanding And can COGS, from case of it?

Yes.

mean, clear in stores the good historically, being and app part I widespread pretty our because that that's advantage is of part distribution, our incredibly -- app as the we've So been partners. stores. mentioned Luis earlier of view we

store our way are increase our Our on primary more over up and time. products our staying than our to margins subscription gone platform and longer app margins users because on year, reduces have And that's a for that gross subscription more fees.

As experimented the be -- expect [techs]. a focus, and other far been it been focus term. for near not to That's the with really different really in taking but we've I it's don't not past, as us the it something in a

the comes Piper Sandler. And of question Arvind Ramnani next from

Thanks, Debbie. All have the questions taken. been good

So you get my stinky questions.

Canadian fall rolling? place. of from in to acceptance are the British about test, universities? Where So of but drives it's it get some know the -- broader I dominoes government I dominoes big kind wanted like think English stand said optionality amount of of a decent I not of a the they it's if it does ask or story, the terms think part in of Duolingo the I would that Or big you sort will like that

about great Duolingo I all Duolingo very something you question. for we're is English Thank the very I'm Yes. the excited test mean about. Me, excited the test. personally, that English

to bit language What able to use are us a if own that the that to quite at Duolingo win talk XX." get I'm I help opposed want do "Oh, Duolingo as saying, . to years French a people where saying, people you school," people to high the love to French would as provides how know say, of it to they're we that, we asked app to people opposed quite point X a said, really you complementary is took about we optionality. learning of to because really score in As their It's would much when "Oh, language.

take which English would that's of that And the the kind to. of to where while. in of score a Duolingo that's But type to course, of a get we lot is That's of legitimacy. get -- going one want test ways

excited very keeps by we're It So it. growing.

So growing nicely. it's

that just our English of important not universities has as smooth very learning of example, kind language of The revenue. smooth governments it's app, should it, the is of I test accepting just test because growth the be to Now the these XX% predictable, our externalities, the mention Duolingo about it. this growth of growth English it's accepting like, as course, for say is going to very which Duolingo

is the So just to less be growth smooth. going

of But progress. We've the number with and growing we're we made institutions U.S. Canada, progress the keep very happy accepting and among progress Australia lot of U.K., We're with making there. the a with now institutions. good

Now governments, and government. by we the you government working Canadian on are U.K. mentioned acceptance and yes, Australian

even Of going that's they've to one. course, year. being the have last the call about which time. been -- out, for for proposals, They and U.K. know go it's the to hasn't just when take gone tests, a going for That's talking bigger don't out. it some We

will So go it outside at some It's completely out. our control. of point,

And thing -- dealing with that's the that about so will governments.

Yes. but your Math? in of sort I like just of sort Terrific. of then if anything I willing if of of usage, you of on in And of can of you don't some on Math, progress know share to -- of kind any it, share missed you're made kind terms metrics kind of kind terms kind

but metrics nicely. happy Yes. it's you. growth sharing is mean I what We're the with we very I are tell exact yet, can very not growing

you growing there's probably see can track ways that it's can look very and at external, You nicely. metrics to

BofA. of Schindler Nat from comes question Next

you're I and Luis, a learning think scientist computer -- expert. language I'm a but last here,

question recently. the search want how know. out You're because a going trying been this to happen. teach French? asked I'm person been ask to this lot going is to box to a I is to really really perfect I've figure me told this I've

you going how like teach box mean to ChatGPT search French? a a is You

Do have told a Yes, hard again any is would that again out it's pressed -- me going figure really I little how work. I'm teach getting languages. truly and And to ideas? that you hard to amazing. keep ChatGPT

Well, out querying just who by years you've am exists there. there I'll to out Books probably for there like able can you this. learn Similarly, books. have French sure learn some tell been reading I savant hundreds something by ChatGPT. been French of

with But small aptitude even there some enough person they average just of I can Similarly, just be language just -- box -- a usually, ask. not that You then search in have what a a number can can not don't student your reading by to bunch sit actually like, able language enough think your to actually and a books. by learn reading be people to average or really learn does motivation people, it does fact, know just And "Hi." books. a of

it's problem yourself thought, other there that believe books. to something learning this of because most the that And the motivated. I we've is sets years, from staying about us think always thousands the And always big is something keeping apart learn hardest been reason also thing the We've technology companies, we is anything called has like motivated. thought is for and educational

You could almost learn actually anything.

want quantum learn You to mechanics.

actually can. You

can and go book. read a You

vast Just choose majority people really of because turns the out not to it's boring.

So anything the thing about is learning staying motivated. hardest

what learn really just going like to you asking Math with keep mean we that you like things really through us have more with that's you I if guide we're so combine it. is apart mean thing, use sets no motivated to hard. are and this the to GPT-X, but yourself. Duolingo, that's And it's to you and it -- guide, are we very by I things also going

majority it's just won't can, it. of do You people vast the but --

it's... think I

are of the serious How language is, say I'd not learning world adults. that which brought up more I And base, billion my in bulk -- Duolingo. know are the actively before probably was of language. subscriber half which of XXX% your least at which are people guess group of who learning U.S., a all do I a size a the actively And foreign market? question. of your around current the X you

So you've I is out everybody goes. Because just where I'm to wondering, be where the when looking and -- -- penetration could how your possibly to cycle? accelerating trying and what MAUs, the at I'm figure this come if this know DAUs and do use way in what's

spending It's good tell things. you a about a really . billion most -- There's billion know a learning We is question we the we, to world we in $XX one things the foreign which is that. a know know. I'll about don't know are really year. language. They're the That that people -- X few ourselves, honest

similar of before which U.S., language a to the market. were country what in XX% They know -- said not market. we're learning if look were a U.S. also our the not the -- massively in very Duolingo. We you users like is you at growing the

market. growing we're also . we the know So

that because then so make it good use and so can engaged people it's feel feel Duolingo they're it and that. friction, with with good -- easy very get they we And we well, they so make it we it, they no about make essentially -- they're that a learning language,

would couple wish a latest mean, I to exactly lot There's million XX learning language. actively tell I more I I that So you. out do mean the But knew, -- there room. know is I people say a a we there's stop. have knew. billion think, where I it's we numbers MAUs. will If I of hard this

weren't just there if you even even But room. -- market, we again, growing more market. the if we're still the is growing So

is, So wish it's -- I answer to my knew. you I

more And questions. we have no

Luis. turn I'll it So back to

really questions. and well been dream Thank is It what that. to as the be to apps That's it. And able we've going for on, as this all working on we're teach you our that is working great to make can. continue humans

the this dream beg language But to please end that the have, most that's is you company lessons that I this to is can remind you thing I that important this do -- dream today. the your has. So single with