How to use data to make a hit TV show | Sebastian Wernicke

Translator: Morten Villadsen Reviewer: Simon Djernæs

Roy Price is a man that most of you have probably never heard about,

Roy Price er en mand som de fleste af jer nok aldrig har hørt om,

even though he may have been responsible

selvom han måske er ansvarlig

for 22 somewhat mediocre minutes of your life on April 19, 2013.

for 22 mere eller mindre middelmådige minutter af jeres liv den 19. april 2013.

He may have also been responsible for 22 very entertaining minutes,

Han kan også have været ansvarlig for 22 meget underholdende minutter,

but not very many of you.

men ikke for særligt mange af jer.

And all of that goes back to a decision

Og alt det stammer fra en beslutning

that Roy had to make about three years ago.

som Roy var nød til at tage omkring tre år siden.

So you see, Roy Price is a senior executive with Amazon Studios.

Ser i, Roy Price er en overordnet leder hos Amazon Studios.

That's the TV production company of Amazon.

Det er Amazons TV produktionsselskab.

He's 47 years old, slim, spiky hair,

Han er 47 år gammel, slank, strithår,

describes himself on Twitter as "movies, TV, technology, tacos."

beskriver sig selv på Twitter som "film, TV, teknologi, taco."

And Roy Price has a very responsible job, because it's his responsibility

Og Roy Price har et meget ansvarsfuldt job, fordi det er hans ansvar

to pick the shows, the original content that Amazon is going to make.

at vælge de serier, det originale indhold som Amazon skal producere.

And of course that's a highly competitive space.

Og det er selvfølgelig et meget konkurrerende miljø.

I mean, there are so many TV shows already out there,

Det er jo allerede så mange TV-serier,

that Roy can't just choose any show.

så Roy kan ikke bare vælge en hvilket som helst serie.

He has to find shows that are really, really great.

Han bliver nød til at finde serier som er rigtig, rigtig gode.

So in other words, he has to find shows

Med andre ord, skal han finde serier

that are on the very right end of this curve here.

som er helt til højre på denne kurve her.

So this curve here is the rating distribution

Denne kurve her er fordeling af bedømmelser

of about 2,500 TV shows on the website IMDB,

af omkring 2.500 TV serier på hjemmesiden IMDB,

and the rating goes from one to 10,

og bedømmelserne går fra et til 10,

and the height here shows you how many shows get that rating.

og højden her viser hvor mange serier der får den bedømmelse.

So if your show gets a rating of nine points or higher, that's a winner.

Så hvis din serie får en bedømmelse på ni point eller højere er det en vinder.

Then you have a top two percent show.

Så har du en top to procent serie.

That's shows like "Breaking Bad," "Game of Thrones," "The Wire,"

Det er serier som "Breaking Bad," "Game of Thrones," "The Wire,"

so all of these shows that are addictive,

Altså alle de her serier som er vanedannende,

whereafter you've watched a season, your brain is basically like,

hvor når du har set en sæson, siger din hjerne bare,

"Where can I get more of these episodes?"

"Hvor kan jeg få flere af de her afsnit?"

That kind of show.

Den slags serier.

On the left side, just for clarity, here on that end,

På venstre side, for at gøre det klart, i den her side,

you have a show called "Toddlers and Tiaras" --

er der serier som hedder "Toddlers and Tiaras" -

(Laughter)

(Latter)

-- which should tell you enough

- hvilket burde sige det hele

about what's going on on that end of the curve.

omkring hvad der sker i den ende af kurven.

Now, Roy Price is not worried about getting on the left end of the curve,

Roy Price er ikke bekymret for at ende på den venstre side af kurven,

because I think you would have to have some serious brainpower

fordi jeg tror at du virkelig skal have noget tankevirksomhed

to undercut "Toddlers and Tiaras."

for at komme under "Toddlers and Tiaras."

So what he's worried about is this middle bulge here,

Så det han er bekymret omkring er bulen her i midten,

the bulge of average TV,

bulen med gennemsnitlig TV,

you know, those shows that aren't really good or really bad,

I ved, den slags serier som ikke er rigtig gode eller rigtig dårlige

they don't really get you excited.

- de begejstrer ikke rigtig.

So he needs to make sure that he's really on the right end of this.

Så han bliver nød til at sikre sig, at han er på den rigtige side af denne.

So the pressure is on,

Så der er pres på,

and of course it's also the first time

og det er selvfølgelig også første gang,

that Amazon is even doing something like this,

at Amazon overhovedet laver sådan noget,

so Roy Price does not want to take any chances.

så Roy Price vil ikke tage nogle chancer.

He wants to engineer success.

Han vil skabe succes.

He needs a guaranteed success,

Han har brug for en garanteret succes,

and so what he does is, he holds a competition.

og derfor afholder han en konkurrence.

So he takes a bunch of ideas for TV shows,

Han tager en bunke idéer til TV-serier

and from those ideas, through an evaluation,

og gennem en evaluering af de idéer

they select eight candidates for TV shows,

bliver otte kandidater til TV-serier valgt

and then he just makes the first episode of each one of these shows

og så laver han kun den første episode til hver serie

and puts them online for free for everyone to watch.

og lægger dem online så alle kan se dem gratis.

And so when Amazon is giving out free stuff,

Og når Amazon giver dig noget gratis

you're going to take it, right?

så tager du imod det, ikke?

So millions of viewers are watching those episodes.

Så millioner af seere ser disse episoder.

What they don't realize is that, while they're watching their shows,

Hvad de ikke indser er at mens de ser deres serie

actually, they are being watched.

bliver de faktisk selv iagttaget.

They are being watched by Roy Price and his team,

De bliver iagttaget af Roy Price og hans hold

who record everything.

som optager alting,

They record when somebody presses play, when somebody presses pause,

De optager når nogen trykker play. når nogen trykker pause,

what parts they skip, what parts they watch again.

hvilke dele de springer over, hvilke dele de ser igen.

So they collect millions of data points,

Så de samler millioner af data point,

because they want to have those data points

fordi de vil have de data point

to then decide which show they should make.

til når de skal beslutte hvilken serie de vil lave.

And sure enough, so they collect all the data,

Som sagt så gjort, de samlede al data,

they do all the data crunching, and an answer emerges,

de analyserede denne og et svar dukkede op,

and the answer is,

og det svar er,

"Amazon should do a sitcom about four Republican US Senators."

"Amazon bør lave en komedieserie om fire republikanske senatorer."

They did that show.

De lavede den serie.

So does anyone know the name of the show?

Er der nogen der kender navnet på den serie?

(Audience: "Alpha House.")

(Publikum: "Alpha House.")

Yes, "Alpha House,"

Ja, "Alpha House,"

but it seems like not too many of you here remember that show, actually,

men det virker som om der ikke er så mange af jer som faktisk husker den serie,

because it didn't turn out that great.

fordi den endte med ikke at blive så god.

It's actually just an average show,

Det er faktisk kun en gennemsnitlig serie

actually -- literally, in fact, because the average of this curve here is at 7.4,

endda bogstavelig talt, fordi gennemsnittet på denne kurve er 7,4

and "Alpha House" lands at 7.5,

og "Alpha House" er på 7,5,

so a slightly above average show,

så en lige over gennemsnittet serie,

but certainly not what Roy Price and his team were aiming for.

men helt sikkert ikke hvad Roy Price og hans hold gik efter.

Meanwhile, however, at about the same time,

I mellemtiden,

at another company,

i et andet firma,

another executive did manage to land a top show using data analysis,

lykkedes det en anden leder at skaffe en top serie ved hjælp af data analysering,

and his name is Ted,

og hans navn er Ted,

Ted Sarandos, who is the Chief Content Officer of Netflix,

Ted Sarandos, som er Chief Content Officer hos Netflix.

and just like Roy, he's on a constant mission

og ligesom Roy, er han på en konstant mission

to find that great TV show,

for at finde en god TV-serie,

and he uses data as well to do that,

og til det bruger han også data,

except he does it a little bit differently.

bortset fra, at han gør det lidt anderledes.

So instead of holding a competition, what he did -- and his team of course --

I stedet for at afholde en konkurrence, han - og hans hold selvfølgelig -

was they looked at all the data they already had about Netflix viewers,

så på al den data de allerede havde om Netflix seere.

you know, the ratings they give their shows,

I ved, de bedømmelser de giver deres serier,

the viewing histories, what shows people like, and so on.

afspilningshistorik, hvilke serier folk kan lide, osv.

And then they use that data to discover

Og så bruger de den data til at opdage

all of these little bits and pieces about the audience:

alle disse små dele af informationer om seererne:

what kinds of shows they like,

hvilken slags serier,

what kind of producers, what kind of actors.

producere, skuespillere de kan lide.

And once they had all of these pieces together,

Og når de havde samlet alle disse dele,

they took a leap of faith,

tog de en chance,

and they decided to license

og de besluttede at lave

not a sitcom about four Senators

- ikke en komedieserie om fire senatorer -

but a drama series about a single Senator.

men en drama serie om en enkelt senator.

You guys know the show?

Kender i denne serie?

(Laughter)

(Latter)

Yes, "House of Cards," and Netflix of course, nailed it with that show,

Ja, "House of Cards," og Netflix selvfølgelig ramte rigtig med den serie,

at least for the first two seasons.

i hvert fald de første to sæsoner,

(Laughter) (Applause)

(Latter) (Klapsalver)

"House of Cards" gets a 9.1 rating on this curve,

"House of Cards" får 9,1 på denne kurve,

so it's exactly where they wanted it to be.

så det er præcis der hvor de gerne ville have den skulle være.

Now, the question of course is, what happened here?

Så er spørgsmålet selvfølgelig hvad skete der her?

So you have two very competitive, data-savvy companies.

Du har to konkurrencedygtige, data-kløgtige firmaer.

They connect all of these millions of data points,

De samler alle de her millioner af data point,

and then it works beautifully for one of them,

Og det fungerer fantastisk for en af dem

and it doesn't work for the other one.

og det virker ikke for den anden.

So why?

Hvorfor?

Because logic kind of tells you that this should be working all the time.

Fordi logisk set burde det egentlig fungerer hver gang.

I mean, if you're collecting millions of data points

Hvis du samler millioner af data point

on a decision you're going to make,

til en beslutning du skal tage,

then you should be able to make a pretty good decision.

så burde du kunne tage en ret god beslutning.

You have 200 years of statistics to rely on.

Du har 200 års statistik at falde tilbage på.

You're amplifying it with very powerful computers.

Det forstærker du med kraftfulde computere.

The least you could expect is good TV, right?

Det mindste du kan forvente er vel god TV, ikke?

And if data analysis does not work that way,

Og hvis data analyse ikke virker på den måde,

then it actually gets a little scary,

så er det faktisk lidt skræmmende,

because we live in a time where we're turning to data more and more

fordi vi lever i en tid hvor vi bruger data mere og mere

to make very serious decisions that go far beyond TV.

til at tage seriøse valg langt udover TV.

Does anyone here know the company Multi-Health Systems?

Er der nogen her der kender firmaet Multi-Health Systems?

No one. OK, that's good actually.

Ingen. Okay, det er faktisk godt.

OK, so Multi-Health Systems is a software company,

Okay, så Multi-Health Systems er et software firma,

and I hope that nobody here in this room

og jeg håber ikke at nogen i dette lokale

ever comes into contact with that software,

nogensinde kommer i kontakt med den software,

because if you do, it means you're in prison.

fordi så betyder det at I er i fængsel.

(Laughter)

(Latter)

If someone here in the US is in prison, and they apply for parole,

Hvis nogen her i USA er i fængsel og ansøger om prøveløsladelse,

then it's very likely that data analysis software from that company

så er det meget sandsynligt, at data analyse software fra det firma

will be used in determining whether to grant that parole.

vil blive brugt til at bestemme om der skal gives prøveløsladelse.

So it's the same principle as Amazon and Netflix,

Så det er det samme princip som Amazon og Netflix,

but now instead of deciding whether a TV show is going to be good or bad,

men i stedet for at bestemme om en TV-serie bliver god eller dårlig,

you're deciding whether a person is going to be good or bad.

bestemmer man om en fange bliver god eller dårlig.

And mediocre TV, 22 minutes, that can be pretty bad,

Og middelmådig TV, 22 minutter, det kan være ret dårligt,

but more years in prison, I guess, even worse.

men flere år i fængsel, går jeg udfra, er endnu værre.

And unfortunately, there is actually some evidence that this data analysis,

Og uheldigvis er der faktisk bevis for. at denne data analyse,

despite having lots of data, does not always produce optimum results.

på trods af masser af date, ikke altid producerer optimale resultater.

And that's not because a company like Multi-Health Systems

Og det er ikke fordi et firma som Multi-Health Systems

doesn't know what to do with data.

ikke bruger den data korrekt.

Even the most data-savvy companies get it wrong.

Selv de mest data-kyndige firmaer fejler.

Yes, even Google gets it wrong sometimes.

Ja, selv Google fejler noglegange.

In 2009, Google announced that they were able, with data analysis,

I 2009 annoncerede Google, at de var i stand til, med data analyse,

to predict outbreaks of influenza, the nasty kind of flu,

at forudse udbrud ad influenza, den slemme slags,

by doing data analysis on their Google searches.

ved at lave data analyse på deres Google søgninger.

And it worked beautifully, and it made a big splash in the news,

Og det virkede fantastisk, og det blev en stor historie i nyhederne,

including the pinnacle of scientific success:

inklusiv højdepunktet i videnskabelig succes:

a publication in the journal "Nature."

en udgivelse i tidskriftet "Nature."

It worked beautifully for year after year after year,

Det virkede fantastisk år efter år efter år,

until one year it failed.

indtil det et år fejlede.

And nobody could even tell exactly why.

Og ingen kunne forklare præcis hvorfor.

It just didn't work that year,

Det virkede bare ikke det år,

and of course that again made big news,

og det blev selvfølgelig også en nyhed,

including now a retraction

med efterfølgende tilbagetrækning

of a publication from the journal "Nature."

af udgivelsen i tidskriftet "Nature."

So even the most data-savvy companies, Amazon and Google,

Så selv de mest data-kyndige firmaer, Amazon og Google,

they sometimes get it wrong.

fejler sommetider.

And despite all those failures,

Og på trods af disse fiaskoer,

data is moving rapidly into real-life decision-making --

rykker data hastigt ind det virkelig livs beslutningstagen -

into the workplace,

ind på arbejdspladsen,

law enforcement,

retshåndhævelsen,

medicine.

lægevidenskab.

So we should better make sure that data is helping.

Så vi må hellere være sikker på. at data hjælper.

Now, personally I've seen a lot of this struggle with data myself,

Personligt har jeg set megen af denne kamp med data selv,

because I work in computational genetics,

da jeg arbejder med databehandletgenetik,

which is also a field where lots of very smart people

hvilket også er et felt hvor mange meget smart folk

are using unimaginable amounts of data to make pretty serious decisions

bruger utrolige mængder data til at træffe nogle ret seriøse beslutninger,

like deciding on a cancer therapy or developing a drug.

såsom valg vedrørende kræft terapi eller udvikle et medikament.

And over the years, I've noticed a sort of pattern

Og i årenes løb har jeg lagt mærke til et mønster

or kind of rule, if you will, about the difference

eller en slags regel, om du vil, om forskellen

between successful decision-making with data

mellem succesfuld beslutningstagen med data

and unsuccessful decision-making,

og fejlende beslutningstagen,

and I find this a pattern worth sharing, and it goes something like this.

og jeg fandt dette mønster værd at dele, og det er sådan her.

So whenever you're solving a complex problem,

Hver gang du løser et kompleks problem,

you're doing essentially two things.

gør du i virkeligheden to ting.

The first one is, you take that problem apart into its bits and pieces

Den første er at du deler problemet op i mindre stykker

so that you can deeply analyze those bits and pieces,

så du kan analysere disse stykker i dybden

and then of course you do the second part.

og så laver du selvfølgelig den anden del.

You put all of these bits and pieces back together again

Du samler alle disse stykker igen for at komme

to come to your conclusion.

frem til din konklusion.

And sometimes you have to do it over again,

Og noglegange skal du begynde forfra,

but it's always those two things:

men det er altid disse to ting:

taking apart and putting back together again.

skille det ad og samle det igen.

And now the crucial thing is

Og den vigtige del er,

that data and data analysis

at data og data analyse

is only good for the first part.

er kun brugbart til den første del.

Data and data analysis, no matter how powerful,

Data og data analyse, uanset hvor stærk,

can only help you taking a problem apart and understanding its pieces.

kan kun hjælpe med at skille problemet ad og forstå stykkerne.

It's not suited to put those pieces back together again

Det er ikke lavet til at samle stykkerne igen

and then to come to a conclusion.

og komme frem til en konklusion.

There's another tool that can do that, and we all have it,

Der er et andet værktøj der kan gøre det og alle har det,

and that tool is the brain.

og det værktøj er hjernen.

If there's one thing a brain is good at,

Hvis det er en ting hjernen er god til,

it's taking bits and pieces back together again,

så er det at samle stykker sammen igen,

even when you have incomplete information,

selv hvis du ikke har al informationen,

and coming to a good conclusion,

og nå en god konklusion,

especially if it's the brain of an expert.

især hvis det er en eksperts hjerne.

And that's why I believe that Netflix was so successful,

Og det er derfor jeg tror, at Netflix var så succesfulde

because they used data and brains where they belong in the process.

fordi de brugte data og hjerne der hvor de passede i processen.

They use data to first understand lots of pieces about their audience

De brugte først data til at forstå en masse stykker af deres seere

that they otherwise wouldn't have been able to understand at that depth,

som de ellers ikke kunne have forstået på sådan et niveau,

but then the decision to take all these bits and pieces

men beslutningen om at tage alle disse stykker

and put them back together again and make a show like "House of Cards,"

og samle dem igen og lave en serie som "House of Cards,"

that was nowhere in the data.

var ikke i nærheden af data.

Ted Sarandos and his team made that decision to license that show,

Ted Sandaros og hans hold tog beslutningen om at lave den serie,

which also meant, by the way, that they were taking

hvilket i øvrigt også betød, at de løb

a pretty big personal risk with that decision.

en ret stor personlig risiko med den beslutning.

And Amazon, on the other hand, they did it the wrong way around.

Og Amazon, på den anden side, de gjorde det på den forkerte led.

They used data all the way to drive their decision-making,

De brugte data hele vejen i deres beslutningstagen,

first when they held their competition of TV ideas,

først da de afholdte konkurrencen med TV idéer

then when they selected "Alpha House" to make as a show.

senere da de valgte at lave "Alpha House" til en serie.

Which of course was a very safe decision for them,

Hvilket selvfølgelig var en sikker beslutning for dem

because they could always point at the data, saying,

fordi de altid kunne pege på dataen og sige:

"This is what the data tells us."

"Dette er hvad dataen fortalte os."

But it didn't lead to the exceptional results that they were hoping for.

Men det første ikke til det exceptionelle resultat som de håbede på.

So data is of course a massively useful tool to make better decisions,

Data er selvfølgelig et enormt brugbart værktøj til at tage bedre valg

but I believe that things go wrong

men jeg tror, at det går galt

when data is starting to drive those decisions.

når data begynder at styre de valg.

No matter how powerful, data is just a tool,

Uanset hvor stærk data er, er det kun et værktøj

and to keep that in mind, I find this device here quite useful.

og for at huske på dette, finder jeg denne her meget brugbar.

Many of you will ...

Mange af jer vil ...

(Laughter)

(Latter)

Before there was data,

Før der var data,

this was the decision-making device to use.

var dette beslutningstager-enheden man brugte.

(Laughter)

(Latter)

Many of you will know this.

Mange af kender denne.

This toy here is called the Magic 8 Ball,

Dette legetøj her hedder en magisk 8-ball

and it's really amazing,

og den er helt utrolig,

because if you have a decision to make, a yes or no question,

fordi hvis du har et valg du skal tage. et ja/nej spørgsmål,

all you have to do is you shake the ball, and then you get an answer --

det eneste du skal gøre er at ryste kuglen og så får du et svar -

"Most Likely" -- right here in this window in real time.

"Højst sandsynlig" - lige her i dette vindue i realtid.

I'll have it out later for tech demos.

Jeg vil demonstrere derude senere.

(Laughter)

(Latter)

Now, the thing is, of course -- so I've made some decisions in my life

Sagen er, at jeg har taget nogle valg i mit liv

where, in hindsight, I should have just listened to the ball.

hvor jeg skulle have lyttet til kuglen, set i bakspejlet.

But, you know, of course, if you have the data available,

Men, I ved, hvis man har data til rådighed

you want to replace this with something much more sophisticated,

vil man erstatte den med noget langt mere sofistikeret,

like data analysis to come to a better decision.

såsom data analyse for at komme frem til en bedre beslutning.

But that does not change the basic setup.

Men det ændrer ikke den basale opsætning.

So the ball may get smarter and smarter and smarter,

Kuglen bliver måske klogere og klogere og klogere,

but I believe it's still on us to make the decisions

men jeg mener stadig, at vi skal tage beslutningerne

if we want to achieve something extraordinary,

hvis vi vil opnå noget ekstraordinært,

on the right end of the curve.

i den højre ende af kurven.

And I find that a very encouraging message, in fact,

Og det mener jeg faktisk er en meget opmuntrende besked,

that even in the face of huge amounts of data,

at selv overfor masser af data

it still pays off to make decisions,

betaler det sig stadig at tage beslutninger,

to be an expert in what you're doing

at være en ekspert til det du laver

and take risks.

og take chancer.

Because in the end, it's not data,

For når alt kommer til alt, er det ikke data

it's risks that will land you on the right end of the curve.

det er chancer der vil føre dig til den højre ende af kurven.

Thank you.

Tak.

(Applause)

(Klapsalver)