Hacker News Reader

Comment by EMM_386 | Hacker News Reader

Interesting to see the numbers. For example, Red Notice is #1 with 149 million hours viewed, and when you look at the Wikipedia entry for that film, it says:

> Due to the film's $200 million production budget and Netflix's global film chief Scott Stuber saying that the company's "big-budget tentpole films [need] to draw an audience of more than 70 million viewers within the first 28 days of availability," The New York Observer estimated Red Notice would need to total around 200 million hours (about 83 million household viewers) to be deemed a success

I guess by this definition, it can be deemed a success.

View on HN

>Red Notice is #1 with 149 million hours viewed

This number is so large that it makes me question the accuracy. The movie is 115 minutes so this is 77.6m views if we assume that everyone who streamed it watched the entire thing and only watched it once. It looks like the most recent numbers say Netflix has 209 million subscribers. That would mean 37% of the global Netflix subscriber base watched this movie in the first few days after release. That is literally unbelievable.

It isn't directly comparably to theatrical box office numbers, but the total viewers would likely translate to at least the 2nd biggest opening ever for a theatrical release. That would be a ridiculous accomplishment for an original movie. There is not a single non-franchise movie in the list of top 50 biggest openings ever[1].

And before anyone responds, yes I understand that the assumptions I made would impact these conclusions, but the assumptions are working in opposite directions, at least partially cancel each other out, and are mostly just used to show approximate magnitude.

[1] - https://en.wikipedia.org/wiki/List_of_highest-grossing_openi...

> Netflix has 209 million subscribers

Aren't Netflix subscriptions notorious for having multiple users on the same account? I've been using an account shared with 4 other people, for instance. The true number of users could be 2-3x that number.

For movie theaters its typically one ticket per person. So easy to get watch time or earnings, etc.

Not saying Netflix is doing this but if the general consensus is that the watch time is too high, perhaps they're not doing simple hours streamed, but trying to guess how many people are in the room watching it too?

So on a phone/ipad it might be one viewer, but if its watched on a TV maybe they're multiplying it by 2 (or some other number 1.5, 2.5, etc) because on average maybe two people watch so its technically two hours watched per hour streamed? That could for sure explain higher then expected numbers while still not faking viewership numbers.

>>These lists rank titles based on weekly hours viewed: the total number of hours that our members around the world watched

Guess it depends on whether "member" means only the paying subscriber, or if "member" counts as anyone in the family who might be in the room at the time of it playing.

Sure and some people turned the movie off after 10 minutes. We have no idea the true numbers so I am just assuming these issues cancel each other out for the sake of some back of the napkin math.

This is not correct:

- if there are multiple users per netflix account, you underestimate the total number of netflix users.

- if the total hour-count includes users that have only watched 10% of the movie, you over estimate the number of users that have actually watched the entire movie. (by counting in the ones that haven't)

In both cases you end up over estimating the fraction of netflix users that have watched the entire movie.

If you assume that there are two users per netflix accounts on average, and that 10 users out of 11 will drop the movie after 10% then your estimated percentage will be 4x higher than the real percentage. So, its probably more around 9-10% of netflix users that have seen the movie.

>- if the total hour-count includes users that have only watch 10% of the movie, your estimate over estimates the number users that have actually watched the entire movie. (by counting in the ones that haven't)

The number Netflix gave us was total time watched. If the average watch time was less than the runtime, that means more people watched the movie for at least some period. That is the number I was highlighting, total viewers and not the number of viewers who watched the entire movie. The existing metrics for theatrical releases is revenue and tickets sold, not people still in the theater when the credits roll.

EDIT: To use your numbers, if 10/11 dropped after 10% and 1/11 watched the whole thing, that would mean the average time watched was around 25 minutes. That yields a total of 364m streams. At 2 users per account, that is 182m accounts or 87% of Netflix subscribers watched at least part of the movie. That is obviously a huge increase to that 37% from my prior estimate.

> If the average watch time was less than the runtime, that means more people watched the movie for at least some period.

I'm suggesting some of this total watchtime is not meaningful (<1 min watch time) and may still contribute to a significant proportion of the total watch time.

You assume all the watch time is meaningful.

You also assume there are no users with multiple accounts.

I think these two assumptions are not very reasonable, and can explain why you find your own estimate "literally unbelievable".

>I'm suggesting some of this total watchtime is not meaningful (<1 min watch time) and may still contribute to a significant proportion of the total watch time.

How does this contribution a significant proportion to the total watch time? Even the last example that had 90% of users not finishing the movie required 87% of subscribers to at least watch some of it. Do you think the number of subscribers who watched even a second of this movie is higher than 87%?

>You also assume there are no users with multiple accounts.

Do you mean a person with multiple different Netflix accounts or multiple people sharing a single Netflix account? I have no idea why someone would do the former and I would be shocked if there was any sizable number of people who do that. If you mean the latter, my previous comment used the assumption that each subscriber equals two users like you originally suggested.

>I think these two assumptions are not very reasonable, and can explain why you find your own estimate "literally unbelievable".

Ok, why don't you throw together some numbers on the estimated percentage of Netflix subscribers who watched any part of this movie?

If there is a lot of < 1 min watch time contributing to the total global watch time it means the share of accounts that started the movie at all is even higher. This is what is surprising the grandparent.

Given that Netflix knows exactly what was streamed from their servers and to which viewers, including if there are multiple simultaneous playbacks in a household/login, I suspect they are much more accurate than back of the napkin math.

Turning a movie after 10 minutes is fairly uncommon though. People usually commit to watching the whole thing once they've picked a movie, even if they end up playing on their phones. Not saying that it does not happen, just that it's fairly rare.

Many Hollywood-type movies are pretty formulaic and start with the backstory of the hero, what is their daily life, before the first inflection point at eround 20-25%, the big event that forces the start of the hero's journey.

I would imagine many people stop during this first act or during the last one when you can tell how everything is going to be tied together.

But that is the whole point of a 'hollywood type' movie. It's pure escapism for 90+ minutes, even if you can guess early on how it's going to play out.

For people who don't like these type of movies, there are plenty of other 'films' (not movies) out there, that cater for the times when something popcorny isn't what you want.

With Red Notice, it's clear from the stills/trailer that it's a simple actioner, you are going to know if that's your thing even before you press play.

Do you have statistics on that or you are speculating from your own experience?

If the second I would say I’m very likely to turn off movies at 10 minutes in. That is the point where I realise that a movie is not what I imagined it to be, or that I’m not feeling like watching something after all. That being said maybe I’m the odd one out so I can’t say what masses of people actually do.

This is not true in my experience. Movie fanatics watch everything in full, other people turn stuff off.

Not true for netflix movies though. Most of them are really unwatchable.

The biggest complaint from Netflix subscribers is that they run out of things to watch. Netflix literally popularized the concept of “binging” I do not find it unbelievable at all that new high-profile content gets viewed by a third of subscribers.

Here’s a statistic showing the average viewing time of UK Netflix subscribers to be 8 hours a week https://www.statista.com/statistics/979048/netflix-weekly-co...

Do you really find it so hard to believe that a third would spend 2 out of those hours watching a new movie?

I assume netflix put it top of the list for everyone in their suggestions, I've never heard of it except for in 1 place, a thread where people with netflix say what film they just watched, in which it's very popular.

They have the data, they know who's popular, what genre is popular (brain dead action comedies), they specifically designed a film to appeal to the most netflix subscribers, and could easily get 1/3rd of their subscribers to watch it.

> That would mean 37% of the global Netflix subscriber base watched this movie in the first few days after release. That is literally unbelievable.

That was basically the estimate for Squid Game that I recall. Estimated about 1/3 of Netflix subscribers viewed it in the first month.

>That would mean 37% of the global Netflix subscriber base watched this movie in the first few days after release. That is literally unbelievable.

why? and some people will watch it multiple times, also if hours watched is just the straight calculation of all the time spent watching the people who go back and watch a scene over again would also count.

I would expect it to be a straight calculation of all time watched, easier and probably the best calculator of value for Netflix.

>but the total viewers would likely translate to at least the 2nd biggest opening ever for a theatrical release.

I would expect that streaming movies get a larger amount of the available audience to look at them if they look reasonably enjoyable than movies in theatres.

in point of fact I watched it and I would never have wasted my time to go to a theater to see it - but I was tired and felt this will be a brainless thing with some funny Ryan Reynolds lines and buddy hijinks with the Rock competently executed, time to chill out and eat some snacks.

> This number is so large that it makes me question the accuracy.

Given that Netflix will have verrrry precise insights into how much of a stream was delivered to a given viewing instance, I'm inclined to accept it.

It would include people who've put on Red Notice, forgotten about it, tried to watch it again, got distracted, managed to finally watch it, then their partner / kids, also watched it, all on the same account.

They can't know for sure when two people are watching at the same time on the same device (TV). And if they apply a factor, they can't know that my wife always falls asleep before the halfway mark.

Like I said, I'm assuming they're just going off "X milliseconds of content streamed".

> It isn't directly comparably to theatrical box office numbers, but the total viewers would likely translate to at least the 2nd biggest opening ever for a theatrical release.

“Biggest opening for a theatrical release” is a matter of revenue, not number of viewers. Every single existing Netflix user watching it corresponds to as little as $0 in marginal revenue, which is very far from #2.

>“Biggest opening for a theatrical release” is a matter of revenue, not number of viewers.

Number of tickets sold and total box office receipts are two measures of the same thing. Yes, with Netflix there is no marginal cost to watch a movie, but just the idea that in the first week after release more people saw this movie than any Star Wars movie is baffling regardless of the specific revenue numbers.

> more people saw this movie than any Star Wars movie is baffling

Is it? It's much easier than going to the movies. You could literally watch it on the toilet.

I can imagine that if, for $20, you were magically and instantly transported to a personal showing (in which you didnt have to wear pants and your cat was there), movie theaters would have a lot more users.

A better number might be star wars theater+home video, but even then that'll be depressed because it's not "free" with a subscription.

Ok, for a more apples to apples comparison, let's look at the Wonder Women movies.

The first one opens to $103m[1] which equates to 11.5m viewers. The new one was released simultaneously both in theaters and streaming. It had an opening of $17m or 1.8m viewers plus 38m hours streaming and roughly 14.9m viewings. Let's subtract the sequel's theatrical run from both totals and we have 9.7m theatrical vs 14.9m streaming which equates to a theatrical performance being 65% of the streaming performance (assuming they should both have similar openings which is another arbitrary assumption done to simplify everything and because there are multiple factors pushing in opposite directions).

Let's dock the previous Red Notice numbers of that 35% inflation and that gives us 50m streams. That would knock it out of the number 2 spot, but it would likely still be top 10. Once again, this is an original movie and no non-franchise movie is in the top 50 biggest openings. I still don't buy these numbers.

[1] - https://www.boxofficemojo.com/release/rl578455041/

[2] - https://www.boxofficemojo.com/title/tt7126948/

[3] - https://variety.com/2021/digital/news/wonder-woman-1984-hbo-...

> both in theaters and streaming

Yeah, but that was on HBO max. That had, at the time, less than 1/5 the users of netflix. Wonder Woman was clearly a marquee title designed to get signups, but that strategy is always risky and not super effective as a one-off - registration funnels are very hard. So, it's not as much a 1:1 comparison as you thought it was.

14.9 wonder woman streams * 5 increased audience = 74 mil. That puts it well within the realm of netflix' numbers. I mean obv that's not perfect because not everyone wants to watch it.

Also, using wonder woman as an example isn't great, because the general buzz on its release was that frankly, it wasn't a good movie. Critical and audience reviews suggest it's a markedly "less popular" movie than red notice.

The numbers are well within the realm of possibility. Although, I wouldn't be surprised to find out netflix is being very favorable to themselves - in fact, I'd expect it.

>Yeah, but that was on HBO max. That had, at the time, less than 1/5 the users of netflix. Wonder Woman was clearly a marquee title designed to get signups, but that strategy is always risky and not super effective as a one-off - registration funnels are very hard. So, it's not as much a 1:1 comparison as you thought it was.

>14.9 wonder woman streams * 5 increased audience = 74 mil. That puts it well within the realm of netflix' numbers. I mean obv that's not perfect because not everyone wants to watch it.

WW84 was the biggest HBO Max exclusive to date. People did signup to watch that movie and therefore their subscribers were a self selected group who are not a representative sample that we can just multiple by 5. Also we already factored in the theatrical release which provided the opportunity for non-HBO subscribers to see it. And to be clear, I wasn't directly comparing WW84 viewer numbers to Red Notice. I was comparing WW to WW84 numbers so the difference in size between Netflix and HBO Max is less of a factor.

>Also, using wonder woman as an example isn't great, because the general buzz on its release was that frankly, it wasn't a good movie. Critical and audience reviews suggest it's a markedly "less popular" movie than red notice.

But it was also a sequel to a beloved movie and an iconic character. Plus the original had a lot of word of mouth success that caused people to predict that WW84 would open bigger. It was one of the most anticipated movies of the year.

Also you are just flat wrong when you say Red Notice received better critical reviews. Both Metacritic and Rotten Tomatoes have WW84 with mediocre scores and Red Notice with awful scores.

>The numbers are well within the realm of possibility. Although, I wouldn't be surprised to find out netflix is being very favorable to themselves - in fact, I'd expect it.

So it sounds like we agree that Netflix is probably bending the truth with these numbers in the same mold of how Facebook was lying about the number of people who watched their videos.[1]

[1] - https://fortune.com/2019/10/07/facebook-lawsuit-settlement-i...

> WW84 was the biggest HBO Max exclusive to date.

im not sure that's a comment in your favor ;) ww84 was pretty bad. i imagine it drove some signups, but not as many as hbo would have hoped for

> People did signup to watch that movie and therefore their subscribers were a self selected group who are not a representative sample that we can just multiple by 5.

Yeah, my bad. It's actually * 12.

"HBO Max accounts jump to 17.2 million"

At the time, netflix had 208 mil. So, it gets even more believable that netflix is accurately representing their numbers - their audience is so much larger than competitive streaming services, that they can in fact drive record breaking traffic.

>im not sure that's a comment in your favor ;) ww84 was pretty bad. i imagine it drove some signups, but not as many as hbo would have hoped for

Once again, I will point out it received a better critical response than Red Notice did despite you saying otherwise.

>Yeah, my bad. It's actually * 12...

You are comparing a domestic total for HBO Max to a global total for Netflix not to mention you ignored the rest of my comment explaining why this comparison isn't as important as you think.

Most of your incredulity seems to be rooted in Red Notice (probably) not being a good movie. But I really don't think that factors into Netflix viewership as much as you'd like. Netflix is treated more as "friday night tv", than "going to the cinema" and people are much more willing to sit through crap on Netflix than you seem willing to accept

I don't think Netflix and movie theatres are even remotely comparable. To see a movie open at theatres you need to travel to the theatre, queue for limited seating, buy a ticket, etc. most people wouldn't go more than once in a while. Netflix is a monthly fee and available 24/7 in your home. You can turn it on and half watch a movie while you're doing other things around the house. I think their numbers are perfectly accurate and am not at all surprised they outperform the box office.

I agree. I bet this is like the media reach metrics of print newspaper,l where it is assumed many people will read a single printed news paper.

It's not possible to know how many people watched a single stream. Therefore, why would Netflix be inclined to imply less people watch it than did with 1 person per stream.

Even if these numbers are correct, why is Netflix making any of this public?

Won't competitors see what works and attempt to copy the formula?

Is Netflix so confident in its content strategy that it doesn't fear competition?

It's to attract advertisers.

I imagine more and more Netflix properties will start to include product placement over time. I also expect a cheaper tier that's "ad-supported" sort of like what Hulu does.

Look at it from the flip-side. You, a classic theatre orientated studio, are looking for a distributer for your new film called 'Big Bangs & Car Chases', now you know that Netflix is a viable place to put your expensive film, if you don't want to go the theatre-first route.

Of course this all hinges on Netflix paying enough money for the rights that the lost cost of traditionally sold movie tickets is covered.

> $200 million production

I'm aware that the accounting methods presumably aren't completely comparable but considering Dune apparently cost about $165 mil to make I am perpetually amazed how expensive some films have ended up being.

Keep in mind that nearly half of that went to the three main cast members alone ($50M + $20M + $20M respectively)[0]. The Rock got more because he is both a star and producer of the movie.

[0] https://www.forbes.com/sites/marisadellatto/2021/10/21/red-n...

Tom Cruise numbers for the Rock, good God.

The Rock is a mediocre actor but a far bigger star than Cruise these days. He's been the highest or 2nd highest earning actor every year for the past five years.

Rock is a great entertainer, cut his teeth in WWE. Why are we comparing? Is that even objectively possible?

Personally, I like the diversity he brings in all aspects.

The Rock has good crossover appeal, from action movies to children's adventure-commedies. Together, action fans + family-friendly crowd, is a pretty massive potential audience.

I think Cruise has the former but doesn't really do any of the latter, so his total potential audience may not be as large.

The Rock will be POTUS in our lifetime

One step closer to President Camacho!

Though given some of the other choices, not sure The Rock is that far down the list really..

The cast of Dune probably wasn't exactly cheap (Chalamet alone must command a big fee these days now that he has the chops and the bona fides now), but yep.

Watched it. So much money for such a stupid movie.

Yup, that’s my first thought too. However hard I try, I can’t justify the money some of these actors get paid. 50m for Dwayne - he is an average actor at best, somewhat entertaining, but 50M?

Movie itself was meh. I can’t get the 2 hours back :(

It's not about how much $ Dwayne deserves, but how much of the % he deserves. I'd argue 1/4th of the money makes sense, since a lot of people click on movies just because he's in it.

I don’t understand how a single movie’s viewers translate into success for a company where users do not pay individually for the movie. Netflix still gets my money whether I watch this movie or not. How does that work?

If you watch something, there is a higher chance that you keep paying.

Although, I cancelled my Netflix subscription a few years ago because I felt like I was watching too much. Same reason I quit World of Warcraft back in the day :)

presumably, views correlate to subscriptions. If view hours drop, it follows that subscriptions (would) drop in the future, if nothing changes.

A big budget film or series eats up capital, so if it is to contribute to the total view-hours count, it must do so at a rate that is better than making several, smaller series at a lower budget.

$1 per hour. Wonder what data wizard came up with that metric.

The same person who probably handled the hollywood accounting that inflated the costs to $200 million.

Cost for 3 A list stars is $60m+. Dwayne Johnson is the highest paid actor and Gal Gadot is one of the highest paid actresses. Ryan Reynolds supposedly received $20m to be in the film.

Netflix knows how to waste money. Having one at that price tag would be typical with smaller salaries for the other roles.

That still doesn't change the definition of inflation. In a very oversimplified example, if Netflix costs $5 instead of $10/month those actors would be forced to accept $10m instead of $20m per movie.

They wouldn't be "forced" to accept anything, they just wouldn't do the deal in the first place, and would go to someone more willing to pay them the fee they want.

That's why I said "oversimplified". Is there someone willing to pay them more at the moment? Why not? Is $20m per actor a universal constant? That's how inflation works.

It’s probably a simple calculation. Those who watch less than $monthy_price hours of content are likely to cancel.

> with 149 million hours viewed.

When I open Netflix in my browser Tiger King starts playing in the background. If you click on a film to read the synopsis often the film starts playing.

I wonder how many of those hours are people repeatedly watching for 20 seconds in the background before moving on?

Considering that this kind of overcounting data was communicated to paying customers of Facebook (advertisers), I imagine Netflix will go for the biggest numbers possible. It's not like that data is audited like a financial statement would be.

I wonder how Netflix could structure the deal with the main cast. Rock, Gadot, and Reynolds all take a share of box-office earnings, right? Netflix does not have a box office, so it's really curious how they structured such deal.

Is the 149 million actual hours streamed, or is it "viewer" hours (hours streamed x average number of people sitting on sofa in front of tv)?

It’s actual hours streamed. Given the intense scrutiny this metric will get, including audits by EY, there’s no way they’d get away with some handwavy assumption multiplier of butts in seats.

I mean, Wirecard was audited by EY. So I wouldn't necessarily agree with the assumption that they wouldn't get away with something.

A success... but entirely forgettable.

Not sure.. It was such a terrible movie, even half an hour I tried to watch will require therapy to forget.

I thought it was a great movie!

It's only been out 11 days though, so 17 to go?

Viewing is non linear over time. Just guessing but maybe half occurs in the first week nominally.

I'll probably stream it tonight. That's anther 1.5 hours towards their goal.

Reply on news.ycombinator.com