Duolingo (DUOL) 28 Feb 232022 Q4 Earnings call transcript

We hope you enjoyed that fun TikTok. It's one of the many examples of organic learner-generated content that goes viral every day without any involvement from us. And we love to see the creativity and enthusiasm of creators like this, which continues to drive our brand awareness and word-of-mouth growth. And now let's get on with today's call.

Good afternoon, and welcome to Duolingo's Fourth Quarter and Full Year 2022 Earnings Webcast. Instructions]. [Operator at close, XXXX year-end you results on IR market our letter can released our and QX and website we after shareholder our with find Today, investors.duolingo.com. commentary, which

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

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

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

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

now will And to it Luis. I over turn

of terms year and a record in Thank XX% achieving These XX% metrics that we XX% year-over-year growth. a It ever was to you, which bookings led which year-over-year with be XXXX fantastic our we everyone. report is with paying and to our million respectively. XXXX We ended best subscribers, X.X and XX%, XXXX with year. highest-ever DAUs, grew Debbie, pleased finished of ended subscribers. which revenue, is welcome, I'm DAUs number QX million and quarter, XX.X growth, and

indicator this quarter, MAU about a up past to engagement, before. to improved user ratio, of Our year which from XX% is an DAU XX%

XX% user you, seen also learners from mechanic last with long DAUs. To the days has encouraging now a the app. engagement of immensely. or up of consecutive grow tool streak have remind grow and days streak to of helping XX% have is number us number that an learns with a for year. efficient is X our our a We be streaks proven This of DAUs which greater,

What's from now year. that million X Duolingo of means we more streak that X longer. million X for took impressive DAUs that impressive is even or have million. years a only with reach to This every X our with of X That the people and us improvements. effect example streak million one get DAUs the than over day it a compounding X more X It's to XX single last year product months is year-long also million to use over of

growth every the user was strong world. of region Our in

The who We have to ages, also trends. has is business travel. a of variety that statuses work, motivations result socioeconomic school have including and all of with diversified all a largely been users geographic uncorrelated nearly and and learn, of macro

monetization Our in power have also of investments shown compounding. the

been all increase past app years, to conversion expect should app what multiple us on the to paid have the new users and across conversion who are subscription increase more seen Over our readily cohorts. nearly the have we've Users who for converted now free their seen year. rates from like also longer this periods, discuss you like I'd an have years, to from pay. And to in X

XXXX. expect In to a to similar we lot of ways, look XXXX

a currently, you growth subscribers massive want market organic strong that and have we users. that and to $XX users only penetrated, billion nicely. our side. addressable barely monthly we to We up bookings about active our expect of remind make We And grow X% have still our on I have

base. further our to user monetize So we have plenty of room current

any again carte increase bookings that bookings from to XXXX. in But not math did want an opportunity in-app also attractive purchases out I apps expect this do we year. to have our from point We ala you or like should

As to the in continue on we've and designers to where growth, managers had will engineers, invest, majority build our success product of efficacy the we'll and monetization.

focused reduce generative teach the improve That in we XXXX: experiments users on we'll how help subscription pay. teach convert how we higher-tier we is costs; engagement, is as means which we're additional the increase will continue and shareholder areas and to And a I English free and that the well DAUs optimize first learners. power second AI, how creation Max and to to letter, in run our X improving as us Duolingo monetize discussed content

early to are like strong I'd higher And We did and well they we as on we've that we profitability. want in so XXXX, the delivering I Coming these to the stages. but grow initiatives, a are profitability. we positioned rapidly returns still bookings don't investments, touch to while emphasize also year. compound about to plan past to last continue seen are excited R&D off hire plan many as from also we our people to users

years. gone we nuts hiring. in year, we've and marketing sales R&D, show grown rate costs, we've twice across to at in count all you, never past leverage remind with plan the to Just X This revenue head And the about fact, and of G&A. operating

up XX% Given XXXX. XX% the guiding we're from a those adjusted about X% EBITDA strengths, margin to to year, for in

to our it to financial that, talk over Matt about I'll outlook. with And more turn

XX% which margin, margin quarterly In to of basis. cash of our Thanks, We XX% impressive of loss was quickly about year-over-year, million, $X revenue and a X% We year of constant year We adjusted QX posted full in highest which constant quarter. ago loss was a a growth was adjusted had growth net our EBITDA bookings a a currency XX%. had we highlights which a results. EBITDA XXXX and basis. Luis. quarter, compared of million fourth free on $XX.X about the the on XX% a XX% net the saw we year-over-year, million recap $XX.X flow To delivered currency

which adjusted growth, was XX% XX% about year-over-year, bookings was currency. delivered constant a last to net compared positive a of of $XX.X flow year, of we cash million margin an growth last in about to XX% $X loss EBITDA net $XX.X saw free about of on We basis. saw adjusted loss $XX.X currency compared For constant the XX.X%. XX% loss which full revenue We had a million million a year, and had We EBITDA we year. million

to continue adjusted XXXX, and an expect strong we In bookings margin progress do XX%. making we while goal of growth, to profitability delivering towards long-term so of our EBITDA XX% will

XX%, guiding year XX% to of incremental margin million margin For a XXXX, translates to about the an and we $XXX million revenue; to the total full EBITDA at in bookings; $XXX $XXX to $XXX in million adjusted to million midpoint. are which XX%

in And and million X% are million an of million in XX%. EBITDA margin to QX total adjusted guiding we for XXXX, $XXX $XXX to million bookings, to to $XXX.X revenue $XXX.X

Our guidance prevailing rates. assumes current foreign exchange

every tailwind, growth. to be year-over-year full bookings. in outside As comes U.S., or a exchange At the in a $X QX reminder, prevailing so we've or on a value our is X% headwind roughly expected about increase total basket decrease from the dollar used has bookings rates million XXXX the revenue exchange of to X-percentage-point our versus headwind our the half of of year currencies respectively, guidance, foreign

same terms total expect year, growth QX of QX. year-over-year remainder of quarterly we our our for as about the for in to In bookings cadence the bookings the of be

and terms be biggest to We rate year-over-year be we in higher, quarter expect bookings. dollar expect QX to our slightly QX of growth

Given this our it year, a growth rate. outperformance will lower in QX slightly be of

plans per part, based This lowering per with speak. we we After lowered globally. capita. to and on sub. our places with was in we're more prices are XXXX, country's believe strong from pricing conversion GDP overall in have annual get platform, opportunity in some pricing each several fact, the to raise But intentional And that countries shift our for LTV in prices drove initiative line -- down the mix subscription prices pricing and we as to revenue increase our most experimenting we monthly the prices of paid in to with done, and trends,

any it, have provide since when an but of material just our included amount with we'll Max only guidance you revenue Duolingo we started bookings We results. or we've testing in our QX not report from update

are adjusted I'd year spending managing that emphasize making We'll based throughout the XX% we our our for ensure this revenue XX%. are to year year. achieve in trends like be so we decisions we that the of focused that margin on full on to and range margin business EBITDA to the gross full

the will be to seasonality will QX, given our Also QX you, remind be QX and and QX lower our slightly margins higher. of than revenue,

to compared run of QX, leverage in OpEx. expect we will see about X our non-GAAP last in Starting expect the meaningful in total XXXX, fourth with And total business in to expenses. we manage prudently of to achieve to X the percentage continue and points leverage discipline to quarter operating year. OpEx We non-GAAP

lesser the and R&D. periods, mostly and sales G&A to is improvement both a For extent, in marketing and,

XX the X% in outstanding year ended issued we fully X% we employees, to which price. shares end We with approximately year-end the from less dilution to using is the the with year dilution expect 'XX. In equity closing XXXX, diluted about than roughly million had

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

Matt. you, Thank

very continue we of I'd Duos, learners. and delight to to into thank our Before team to new talented our who innovate like get find Q&A, ways

here to And going Debbie now we to questions. back and to take happy manage queue, we it be turn to your I'm would the go.

question Instructions]. from comes Schackart [Operator William first Okay. Thanks, Luis. Ralph And Blair. the of

letter, shareholder XXXX. you think a I to called the it end First question, "Super"

that I to users. incorporate think some that year may you campaign run attract of the new beginning the of

I really follow-up. did Year's curious, performance year-end? Just And And then strong expectations? New your a what drove in have the how campaign to compare the

That's you. a Thank great question.

great reasons. had So a X we yes, there's reason main The year-end. -- great main QX,

our The getting compounding users. better better mean been first improvements. products product more and just one stickier our we've and at just I making is attracting

high-performing We that very growth thing. does one have that. That's a team

where We shares. shares, over to days. -- really or stats -- And topics lot marketing we year-end of their people And summary whole they're did our Twitter, hit stuff that got example is good well. all what then a on users was one millions campaigns, of Duolingo December they just XX the review December year. got so And XX, several throughout top A XX of or the and just campaign and did share expected share another of millions that, some having they're around a towards we December trending by well, encouraged of to like our it.

like stuff helped outperformance. that So do us really the

Year's we New just start every the usually way of December for reference, In it all goes on and Year's XX year, terms campaign started -- we a January. New campaign, to

little So quarter, performed campaign, learnings the take and the better. from reason than campaign, about it's for And fact of the they do of to previous a year, outperformance just the that the we we we better with for has just X every days-ish this expected. that get and year

A/B more So few year, to effective, make we take year. next and we those every that run the a campaign to tests for try learnings

So us. that out worked pretty for well

quick Great. Maybe a follow-up. just

Just is think in what's for terms you I driving to tough DAUs, and color pretty do. maybe consecutive quarters, acceleration reaccelerated some which But X strong just growth. the of

there of or in sort the than about remarks, that's broad-based. regions growth talked prepared But think stronger I of X sort you X were others? any

it's too. Yes, great a question,

the really of in it growing every in world. regions, So in is terms single region

want stronger the is to If U.S. you know particularly regions, strong.

strong, is Asia terms is in growing Western particularly particularly DAUs. of Europe and strong

out and terms just better with better reasons. time. It continued is and the improvements product better The strikes then why it's at X just mean stickier, of And it I that product and marketing. same in just figuring a is, over chord again we get what compound is

So our campaigns do. some TikTok we media and of and influencer campaigns our Twitter also like our campaigns that work things like social the

we even the And do. stuff, don't some of

of don't organic entirely just the that goes but this there almost that you every call. beginning It's people day, every viral These other Duolingo. every know example, day video if is make. single videos are in a saw organic day. single I almost mentions is it an For at this that single that video fact world, the

So all -- keeps of that just growing users. our keeps getting

KeyBanc. the comes of Justin And from Thanks, question Patterson next Ralph.

to optimize there recently, blog just the using Luis, on data a post DAUs. really science about interesting Great. was Duolingo

rate the that drive how cohorts. you've individual thinking past current to with really you're years. product few was the using think together. Would to And top-down about lot to more that metric, far user current the a how now I tailoring in of link that you've statistical pretty models team then gone hear just been love optimized retention you're these the metric growth, about DAU

a to of having Z, sure So like, more bottoms-up things attention spans of much say, having Gen approach advantage healthy engagement. and shorter they're take making still

proud very glad We're post. our I'm read growth model. blog you of that

were more it. And important significantly what mentioned over improve the is, is post, to the for days, is sophisticated what is current -- which we've CURR, they we right rate In at and what years, and or single most making analyzing the other the cohorts. is number That retention been last last say the here it rate. has one come at called that's our here if they're XX% probability user that the and current over today retention years, gotten last company call going this changes that we Over user people something just over they that were now, the we've increased I X would back time, tomorrow, user of rate. around least this base several metric certain --

increased to about by CURR our and the X, Over we've lot. X% X of that increased DAUs XX% years, about a increase XX%, every from last number

has keep more model really -- so we and more helped sophisticated. And that us, and been getting

we're when current model, starting users, this would treat sophisticated, a was first, before getting et -- we kind our started we cohort. and users, you're of to new at cetera. right, all in just model more are one Then we treat this just -- -- You growth users

getting and a that think -- sessions don't than shorter shorter we we you span. are have thing shorter our specific like for do. and making we a because, are even I But that. attention have more Gen said, Z relatedly, Now shorter they we're sophisticated we

Andrew The JMP. Great. from of question comes next Boone

half from quarter. as beyond X-year about with something you Is it there a saw XXXX? there? think this of year? reason Great. And talk should profitability, Can contribution And then this what just back we Can I that about last learnings Matt on purchases? about or kind think of talked this margin down about are we [indiscernible] off steps the doubled real we how any secondly, successes year. think that, of to in that any start for in-app plan kind upcoming that then

question, I and the the IAP can one. take take can you then Matt, next

behind just have for and prominent, in-app and this that more high-performing the successful making we in yes, purchases, team more more So a very in-app purchases more is app.

been they've saw, rapidly. growing you very As

rapidly year. to them growing expect We continue this

by the main the by learnings, is we of terms in And purchases, way, selling way do in-app gems.

them. or to buy kind in-app there's So can the gems you by in also can that And use either then currency power-ups this on you called can you stuff, app of earn app. doing gems buy stuff the other

power-ups that app use main the timer we help the for is the of that One in boosts certain timed lessons.

are get most type spend -- lot there not boosts time idea purchases. to out boosts on fast, run timer timed, these timed. of going to those us is are to see or if with gems. can a get you And timer of Duolingo the people on buy And in-app these have we lessons that you So do. stuff the you're are This not can you lessons very that some that is some -- you're do lessons but practice of

be we're gamification general, like. content purchases, for we're but people be selling that in-app going our to in to within us, not really purchases, going mainly selling for So things

And so well. strategy far, has pretty that working been

on Great. profitability. And then

point increased This improvement So goal, basis guiding Andrew, the XXX XX%. our on XX% EBITDA which to in XXXX was of from to about good XXX-basis adjusted points by profitability XXXX year, to profitability that. basis, progress an long-term towards we're that XXX over

longer showing that guiding year, we year We term, be goal. what want right progress so our of we margin. between a goal XX% to look and And EBITDA out think And adjusted to the towards at into to changed. as this kind XX% we're we're the hasn't last

the Thanks. question MacDonald of next from Ryan Needham. And comes then

seems citizen is year. the online And you close Luis, the the fully China on to a degree. that start this person what university. kind an of how strong a Duolingo with was it in for on are test. international curious Congrats impacts start in on looking your for XXXX English me expectations to just no let you've an from Chinese There ban on degree -- recognizing taking of degrees, or in this recently more a to think from government essentially longer They're online going like or test shift far if Chinese any thus overseas. the in here. and inflection of that sort applications English learners Duolingo a sort of I'm sign-ups seen and point, go creating you a

So well, thank the for question. you

large for all We have have of based but I we reasons. China I really not relatively mean, think is seen takers one, on Duolingo a any impact over test test English many this, there's the the and chunk world. yes, takers,

are reasons is for tests varied. very thing other the the taking The

Some applying it. admissions. of There's are all kind undergraduate And also of kinds healthy. them to There's transnational the sort policy admissions. one pocket international one general, of maybe has graduate implications pockets. pockets education. who That's in There's different But are is some people for thing. of pretty just

are -- test, with the coming is usually thing other this that The this not these had it we we into leader tests. all online in We're and market. is an market off-line has a market are still

tests in so still volume significantly that given a when don't this, something that kind yet. we're us the is we're under We're this percentage it growth something affect have, we we that changes pretty So there's the coming growing of still because with in -- of XX%. whatever fast, phase it too are of which doesn't is bit quite full seeing of we not much and

follow-up. it. as Appreciate And helpful. a That's maybe then

the lot the content You very creation. use I'm generative around and curious about talked AI and a of that,

think you how do AI time increased English non-English you this showing Matt, compress the content you're you within you that are margin assuming sort in basic creation from the creation -- an generative content or that that line? speakers content do for quickly more that? speakers. your advanced AI? focus profit And mentioned can leverage How year any side, And for of manage You English on then any quality on the are leverage assuming for generative from

Thank Yes. for you the question.

at AI generative put So context, are our et products mean cetera. built very AI, always based from just on of lot to we've leaned things into And beginning. excited into we're about AI. Duolingo, I the a

of the lean going well. years, become significantly has last we're better, couple to Over definitely AI into as this and generative

up That's me how lot things content one creation. that a it you are mentioned us speed I bit we for a but It's help the can early estimate -- You up, a and to where give speed already costs. a you places it's think and much bit. of precise speeding We're lowering it can mean quite things of also we certain already happening.

One learning much that content. able English a where the we're of to using be more places is create to

So you'll that. lot a see us of more create

all In reviewed humans of just generate. it now, go quicker this to almost time review case the to still lot terms assurance, it to going quality have over But to of By was has a than by it's generated humans. and by that humans, entirely it. we're used to the frame, be be

terms a get going massive and I to assume increase same in that, So of for cost. in frame, we're time the it's

this So we're yes, about. excited is very something

And going to should so But other tier help subscription, create the the to just stuff online use that. be that content type able we're one I better and we're them, new we're to is creation is do. practice Duolingo our AI, higher also mention Max. to going of that to of it that's our thing putting experiences of going us kind And conversational generative definitely into users, about more

excited something we're is very this So about.

and on very we're with company, hard a AI, going we've partnered As working this. be to open

the We is into things Ryan, mean bet. And direct in of not AI guide to answered we ton yet. probably of It's But have direct generative the way as Yes. time, over included line into. guide. I new. -- we sight so a only don't have Luis that's said, for of Luis the line savings question the sight

of comes James. the next Aaron question And from Raymond Ryan. Kessler Thanks,

questions. of Maybe Matt. couple A for

paid And kind Can On QX. the it that? maybe any part about conversion then we just strong should then how about think maybe thinking that? there family especially rate, relates on any in very direction in tutoring a XXXX? to paid that of maybe as conversion update you're plan Was was And you a as -- on talk of well. test, given that well? as integrating about Is it talking human about well, how of as that Duolingo thoughts Max, kind

Luis let answer part. I'll last Yes. Sure. that

compound throughout quarter penetration, the Luis we mentioned, has metric. the For had that to expected of think time coming year-over-year most so to we before, I of was go described or last any make in X% And of plan product or up, them that our It There that's that a from the family item each payer all year, over increase that things X% wasn't free-to-pay that. all conversion this like particular in over improvements compounding one general. time. all

given every the I X growing or just think X of mentioned think, again, forward, call quarter the X% I to as closer on, denominator QX year, this nicely. and retention MAU, in really, instead points is going trends, really QX be of growth growing denominator or that the is year, nicely, it that we because, number last actually so just really we year-over-year, would to expect adding that

point. want will it to And talk So the subscribers about in into then, X time. will over and be tutors? short you about turn it term, Luis, But ultimately Max

all. right it We last over we and tried human the We that is human human are we is or launched think I was tutoring. We've tried that to we it. mean with Duolingo the first, I involved. the not year, second about tutoring human in. on no decision Then tried tutoring going about waffled This get at the the question it. tutoring. over made At tutoring. we've -- question the was human decision putting humans Yes, with into first we've -- years, tried

we are human a reasons nothing like number not There's against a -- tutors. that's of excel for are have end, and it's anything we just our It's at. technology We we that. what There's out. company, on

at getting look in they're if AI, human generative are and They're in Whereas there. particular you but getting AI, that, better. of and great, from just to year. particular it, lot better better any stuff every not technology, just the a addition in coming And tutors is

So language though we're computerized could XX told ago, even computerized with on we that, years were to synthesized in same ago, synthesized way, XX on going teach not voices. years bet Much the bet still kind that. they like we a of technology years, voices rather that AI We the sounded years, robotic. X, we that happy be feel have Today, excellent. X don't will I on in that sound very we'll XX know, with same than humans. We

from next question And Bank of comes Nat Schindler America. from

want just go a Yes. before, I the you about little have of some had deeper the actually really questions to changed. how users on

You prior year there those? subscribers the Has than Is were dramatic year last substantially about and a been in this real a demographics anything over this in of year? both change there users there versus different saw the paid year. users acceleration

a It's our great generally, been, user very it's question. is has wide. base There's -- If there small.

in country country in world Every single world. the single from every users is the have We represented.

status We from to have from long as read. ages all as single socioeconomic how users know every they

So it's to of from X kind about XXX. years

users all wide the it's Just base. and -- place, really have over just we a user very from

So it too changed much. hasn't

TikTok, social has and that younger, thing ever one prominence we've to do like media but on so a our it with -- gotten slightly, that's has of lot noticed slightly it. our The probably is

really haven't change. a too we noticed much of So

and users those as both paid? towards changes in has going that And younger, free slightly happened

is don't Maybe... I don't that free paid. younger. I slightly know getting know answer it's is I -- for honest The know. users

the have paid. to back go double I'd Nat, Yes. and check, on

I paid think seen that the changed. haven't conversion what on really is we've trends

Some have accelerated.

overall of year. course last just the across over For kind conversion example, of of the cohorts has board increased

converted the in platform first more So folks the of new XXXX. couple to than are months, who rapidly they in

have that a who had better. Folks for always convert the got time, on the than the the little app a in We've bit or been That platform they use converted more people stayed long of same past. period more. bit seen that faster a little they

it base. of yes, broad So a kind was

broad-based a around the on grew to our trend users way, like in world free same pretty paid Just the conversion. the

you very again, very much. Thank impressive. And once

Mark next of Mahaney And Thanks, Nat. question from comes Evercore.

questions. Two

line of or You been the been the is But the how the brand-new number users. is for had users would of the it's returning this I that, release always that as roughly case about in assumed just they'll the case they have has a something language about I number while. that's point when will how come realize back then country. a it, people learning to different happened? either new or importance new they're it another Maybe in be that Because and there'll traveling to use a your get opportunity. same

be? really in So not? being What's quickly, real that. though, just to you what the just that? want it? is you've math interest Is still trend a And kind seen that is it new adoption, of Math, you that out or tweaking Are so product, from it talk Duo where Just that? about where far of how much

new to go been even a case. away very while there's returning I'm not see away and the that. never people may same user a goes But first a right. That is X years since us such and just for a was back. to little assume you're going question, then comes because is because that as of new It's your Duolingo uses is while friction. early like It's not that been it's it rare Usually, come Duolingo, And for It's users a back. thing. this rare. the users. the For has are that been at

come something, rare. think our they take is So and it is to churn months been as back not then a That a to more is we months It's having for for there and we may see they that that right reason of that few they a the stay users come. gone mainly break forever. couple coming, of that. put emphasize or They something. or And

that. like something So

that. math, with terms It's In very growing the with happy happy of rapidly. progress. We're we're

the exactly to you be -- a Just active it now, million users, itself. -- get kind there, is the Once pretty users, our going of to we're math X of of are decent organically call a by goal product have get it is okay, great. million. number right to grow to to remind we it we plan, Once say, for to we daily this doesn't grow

but It good growing results. market. there have yet, with time not and We monetizing. is so We're it product to the is we're start pretty happy nicely,

In very we're the well adding that to working lot of We're features of math tweaking we app work a on the main the product. terms on the guests, hard know tweaking Duolingo app.

the of adding improvements better thinking like So to We're notifications. leaderboards, things Street.

Duolingo the content app we're code of get to we've adding over to to We're on it and higher a it. to add the starting the we the So learned last to for, XX years lot stuff also retention. -- that that more have have math

of with right the content now of we that the the thing, run have it very use content. -- math is of they you actually very out run users some turns because heavy of people out One things are that if and out happens kind whole app then

more we're So adding to content it.

So just better. us it of the product you'll continue And make year growing. will see the next that, over better and because

Eric Goldman Next question from Sachs. comes Sheridan of

with has continuing how momentum be a Travel macroeconomic potentially talk to business you on and travel the implied little would usage, the to been result first the as shift funnel seen question about for return, and the be catalyst following a tailwind travel goods you've cross-border a bit bottom Maybe in tailwind something could Mark's forward and XXXX going funnel both was of what in the obviously been as has that's environment. continued for that and as a question. that over Can of nice services up over a a big of Great. tailwind, to maybe beyond. top first build the big

on content in to or lot TikTok. about and thinking platforms that ahead? think And away building about excited distribution Can virality we would about a maybe TikTok you be, or marketing dynamic then side virality on platform the a bit you side, in about building experiment the from the even should where years are you how be and the little talked talk second

Great question, Eric.

Let's see.

the years of height see slightly Since We our is a COVID more then, travel. for don't In terms been That metrics in dip of fewer -- travel, temporary more ago. are see and it's about we Duolingo a did the like fewer. people in same. a couple very of using that

is interesting. just we And -- travel

For be worldwide the a to perspective, that wealthy. people like pretty lot from are, of us honest,

like using or to a they're a our using motivation something. learn hobby a of think But for is or English travel Duolingo. I users, big it lot as for

my do So see things just like, "Oh, -- up, why we these people exploding." that. we We they're just Duolingo therefore, that's Duolingo. sign up, ask much of when using opened don't haven't is has god, seen That travel and

the tell we tell head, my not top And assume something it to. some something for that I But us off know it's travel. look don't us I'm using using that they're it's of much too XX% number the like to it to just travel. going fraction they're for

In terms second virality. of your question about

U.S., the to Not our been with other stuff. media we far, we've the and as few experimented so in TikTok social We And way. mainly platforms just TikTok we don't this localized Spanish Portuguese have for actually make word just a this to for us. So, TikTok really by pay content. get in languages. good out, and good organic, is our really has

how does things we're by actually languages. thing of They're do excited really the with about fast. well. amazing So in the social is is the it's -- Brazil about way, just this accounts one that in Anything TikTok And Brazil. amazing you actually growing other

they're XX soon Portuguese, Wednesday. from the dressed Duo of just in TikTok nuts. got, we put as so had many, go as Netflix Wednesday as show, don't kind or We it I million And you And amazing. if how something. views see those we And just went up know videos,

and about pretty TikTok excited we're actually doing platforms So We're Twitter. Twitter. well also in like

This is lot about doing And a able to get to things. too, now things. really and share we they the in are more or inside get things team just good we're more a it. adding inside the not a It's more that general, you thing media when we're a another product. virality ask is Now team, product people lot anything, social share behind to have people to

share, One way sentences now more of the easily. a main much things to people that there's share

are share people weird things main that see Duolingo. they the of on One sentences

So And husband," -- I of German of has one of the a has or the like something or for of I some something something just one "right course body bed, of people. people know sentence share. gets the have them. sentences. the a can weird don't if really lot weird to Like next the that lay very courses, like, reason like the that think stuff another,

on getting we're working it's working out So and people to like share that, pretty well.

I sentence about wrote the about -- okay. question that that want to say husband. probably Matt the But,

Mario Barclays. Lu Next question of from comes

Great.

first store is fees. app the on one So

thought the apps store. on your Just bypassing of trend recent curious the app to

and mobile Spotify, more done more method. games have like a like direct-to-consumer Fortnite example, a while. For traditional implementing this apps payment like are recently, for games But Netflix

on that have thoughts your you end. your curious Just on any guys initiatives

We're working not that. on currently

the stores. if love wouldn't it's it decrease we fees. they love For say hard us, we to their Obviously, that app

course, quite really that it friction. thing off with amazing stores harder And distribution a when the we the it's you And product. They're go But your would. also, we to app App is well distribute Of an of do has Stores. mechanism. bit

have about is and already card have the of that gets there. once and stores tap thing One to people to just amazing it they app in the credit the there,

too So lower it's always that users. eventually not that you would us but fees, working your we friction about we're add more something something so make able be to may clear to much it's we're currently money. do that to may thinking and but It's your much. that much it, on But you not

quite their a last from we significant fees get lowering bump And did remember, app store Mario, year. Google

So...

do to between about how Any for changes And And the at to parity. of then just Helpful. past it. being you quarters, Duolingo second the the app just one change you updates itself. Got single track balance learning optimizing of I kind kind in on to few that? general, talked believe having masses? just lab negative type that oral I But your So minority. thoughts to casual wanted removal be learning the prompts of got things like and versus hear a I attracts more like or kind could just some feedback. believe of the app a feedbacks, user answers, think like I there.

Mario, you Reddit. read

do. I

Yes.

great a app. about to the changes it's question So the

changes. always on why We're So mean we making much this better. is R&D. significant better. so spent I we pretty This app work on. making some literally always we what We're app the the is made

this about made year, to from we the giving what large change a this to our going last to linear lot learn -- home next Over path. of choice screen,

easy The people they not really a -- to main learn. fraction coming to by way, encouraged we a of giving increase we're linear Whereas we that the learn did on fraction path, also users they streak, their much of reason them, linear actually getting was noticed the that just we're actually the so we was to choice, extend going with when to larger streak something. to what lessons teach them significant in were We path, was we've better. with now that were significantly come whenever

reason main better. also better has But to to the simplify product, our that to teach us. really that teach actually was So actually and we which did was helped

And try teach. improving to The that there's is, to engaging screw have we're monetization teams And way happens is other We're well retention, get run time at and people look people we're app making the Whenever we're what subscribe, so of monetize. to how how behind that up enforce what do things We're on well these cetera, we improving they we -- we one the kind more is a of et when that cannot we team, how trying they test how we each think improving. large team, spend they're example, group subscribe. improving metrics. to it, have is much metrics. to we is we it of X what the for changes, basically, -- more of get each people how

get decrease but our So cannot you yes, you DAUs. subscribe, to can more people number of

cannot learning, learning outcome. people time or the decrease you cannot the decrease spend You

to pretty change to the teach or make the grow to it. that we monetization the cannot cannot made of teach up that works can can people we okay, screw And is better, screw that, actually but how Similarly, change make get group found we the faster that That's the or make better, more that people them, changes us balance people cetera. engagement. team dedicated trying most when you in user or up tell trace by of -- you the we app, well. et And down. was group money, us by make it is you That every you

so you how that it. some Hopefully, color. do And gives that's we

comes question Ramnani next from Our from Arvind Piper Sandler.

me management the is, that. first of streak I ask the in has guess let question my who But terms now team, among longest usage?

to but me, the know don't it's answer bet I I'm to I the question, the don't the question. answer But willing know to but

sure about pretty We be can that.

Yes, yes.

of with other beyond whole of the revenue really track kind a user there's with I -- bookings. what good a and that But you So and share engagement lot growth us. metric you'll know metrics

a some growth the to this So can you able of more are the for you're giving of year? kind you of the not share on you -- conceptually, numbers, shape share hard underneath to able if conceptually confidence bit in that's metrics

we share was product our as if each of I In mean and what DAUs, make, of metrics, I money X -- the usually we track do are there's Are metrics. I lot teach. monetization And numbers how using just a share we before, metrics. ton Yes. things know have metrics. we is is the we have There et they general, entire can of is base. there we saying which much more don't but which we precise that In the cetera? the certain metrics, is basically. there's then about user them, metrics we well stuff growth a Do of more? groups one how call of don't that share, we There

I current at I MAU I retention at at, kind users. least users. look resurrected of of the spent number of the I ratio. resurrection at resurrected. things rates, look terms the look so retention I of people users. are of DAU look of to time growth, In look at that look I of being look retention learning. at some new I at at the that

the So type of that's stuff.

users is that look of of types on stuff users of IAP free at are to paying an we In purchase have number users. look or monetization, conversion fraction LTV, or a we of paid, things IAP. do something at a the lot, a from Daily an retention like bookings

users the main of give difficulty In content learning, but a is the number of things without all at to things, we of we're away. look able we at out that scaring look one

we my lot top the and of of a head. them of there's metrics track, So that know even don't off I all

to you And helpful. That's the of it's in are to user inputs of coming even know for numbers. growth box. the that project for don't sense, into growth, some what a a sorry? growth, like just type of -- goes kind projections of But you what into then able predicting go Yes. sitting the when But conceptually, black up bookings I externally booking growth, kind of specific mean share growth? are bookings bookings kind we bit

take kidding. a well. just has up works that. a go We ruler Matt that. than pretty better and I'm like much answer It

subscription you renewals, main thing then side. think and new Well, with growth is growth, maybe -- subscribers to there's the adds, and there's so mean bookings bookings that X I just in can not. add about dimensions. obviously, IAP. on There's the And and the new

spend. want the a to on we how into blog new have a predict model, our post model, about model, and growth that's really combination input how complicated marketing read that our An is users, organic markup users, new On if users. and basically talked of we you a Luis growth

new for predict business, doing think have we basically time users the about. new we're will adding what we renewal element into to good And this on our pretty cohorts, subscription over they confident then at cohort we feel for which just of we're bookings. point. new convert of some We flows into. So And the cohorts then data That into that really users. see math

that your cohorts, got Add are sense the adds in bookings. in for that. IAP, together, detailed less a really you simple, little a bit get which again So you make less then and Arvind. you that bookings. And It's add subscription

we questions. all for Well, have that's Great.

Luis. turn to going I'm So over back it to

beg Thank And your I you do lesson. for please, all the questions. please, excellent you, Duolingo

Goodbye.