An Anime Insider's Take on the Netflix Engagement Report

In an era where one of Netflix's top new releases for the year is a live-action adaptation of a manga and the streaming giant is starting to lose subscribers in much of the world outside of Asia, how does anime fit in? It's a question that's been difficult to answer for most of the 16 years of the company's experimentations with anime – Netflix has famously been a black box of data, even to its closest production partners.

However, last month, Netflix shared details beyond top 10 lists – finally, the hours viewed for their top 18,000 or so titles are available for review, covering the first six months of 2023.

As someone who enjoyed a huge amount of data to play with in roles at both Crunchyroll and Anime Limited over the last decade (and conducted regular content analysis at both), I couldn't help but dig into the numbers the moment they were available for download. I've done extensive analysis using information similar to what's been provided here in my professional roles. While I'm on a sabbatical from the anime industry, I hoped to share some of my insights and perspectives with all of you.

It's hard to look at the complete list of anime from Netflix and understand how they compare to one another, even if you account for hours viewed by hours available. Dramatic differences in regional availability, a lack of episode-by-episode data, and the other limitations of the information provided by Netflix detailed in the introduction require broader statements than I'd typically be comfortable with. That said, there's just enough here that we can start to understand Netflix's unique relationship with anime.

While the Asia Pacific (APAC) region — for our purposes, a territory including East Asia, South Asia, and Oceania — is of particular interest thanks to its disproportionately high consumption of anime compared to the rest of the world, it only accounts for roughly 17% of Netflix's subscribers. To that end, it's best to start with the 30,000-foot view of what Netflix users are engaging with when it comes to Japanese animation.

Regarding the breadth of its anime library outside of Asia, Netflix is not a contender. The first half of 2023 saw them release just nine anime titles with enough exclusivity to be considered "Netflix Originals" – comparable with Disney+/Hulu's offering in terms of quantity, but without breakout titles in the anime community like Heavenly Delusion. Compared to anime-focused streaming platforms like HIDIVE and Crunchyroll, the offering is scant: HIDIVE released 21 unique titles from January through June. Crunchyroll's line-up was more than an order of magnitude larger than Netflix's.

It's true that some of the most celebrated anime of the year were Netflix releases of one kind or another – the recent Pluto adaptation and Scott Pilgrim Takes Off come to mind, both with particularly strong claims for "anime of the year" – but the first half of the year was not as kind to the category.

Record of Ragnarok: Season 2's performance is fascinating here. Though ranked 150th overall in terms of hours viewed, it was the fourth most-watched new release on Netflix for the first half of 2023 without worldwide availability – though the reasoning for that is interestingly not because of licensing agreements. While over-performing the most-watched anime of the period, Demon Slayer's first season, if one accounts for hours watched compared to the available hours of programming, that achievement is made less impressive by Demon Slayer's cross-platform availability (in the United States, it's streaming on Netflix, Hulu, and Crunchyroll), as well as the fact that the shonen hit has been out for five years.

As the top new anime of the period, there are few other favorable comparisons to the larger Netflix catalog; Ragnarok serves as a representation of how weak Netflix's new anime line-up was at the start of the year, missing the top 50 new releases across all genres in hours viewed. The series made two brief appearances on the Weekly Netflix Top 10s in the non-English TV show category when it was first released. This is not in itself significant – there were few new releases in January of note to compete with in the non-English section – but it does speak to Netflix's ability to mobilize fans of the franchise, which had not had a new release in over a year. Of the 78M hours viewed of the series during the six months of the report, 48M – an impressive 61% – came from the first two weeks of availability. As Netflix highlighted in both its statement and press conference tied to the release of this report, "Success on Netflix comes in all shapes and sizes, and is not determined by hours viewed alone." A title that can encourage less-engaged subscribers to smack the red button on their universal remote for the first time in weeks is worth more than an additional few hours watched for a cohort that already loads up Netflix on their Roku as a daily ritual.

How much of Netflix viewership is anime? With the database provided and a bit of elbow grease from yours truly, we now know: approximately 3 billion hours, or 3.3%, of the 93 billion hours of content consumed on Netflix from January to June 2023 was anime. With 30% of viewership going towards foreign-language shows and movies ("foreign" from a Palo Alto POV, granted), that means more than one hour of every ten watched in a language other than English.

Is that a lot? It's complicated.

It was big news last year when Netflix's creative head for anime, Kohei Obara, shared with Variety that "more than half of [Netflix's] members globally tuned into [anime] last year" — a figure that would put Netflix's penetration into the anime scene in 2021 at more than a third of the global 300M fans, using Crunchyroll's estimate via internal research as the total addressable market. From that perspective, it's a lot!

The average anime-watching account on Netflix will watch at least 50 hours of Japanese animation on the platform this year, if current trends continue.

If we're to assume that the rate of anime consumption on Netflix has stayed steady at a rate of roughly ~50%, that would mean that just under 120M accounts consumed anime or Western titles that Netflix categorizes as anime in 2023. At that rate of viewership, each anime-watching account would have watched an average of 25 hours during the first six months of the year, or 8.4 minutes a day. As a point of comparison, Crunchyroll subscribers reportedly watched an average of 85 minutes of anime a day. The average Netflix household does watch more overall – 2 hours and 8 minutes a day – but even among anime watchers, only an average of ~6.5% of those hours are spent on anime. Other streaming services do not have these kinds of numbers available for public interest, unfortunately.

Particularly when we explore how the bulk of those hours are spent, though, we reasonably would expect a more Pareto-efficient distribution of viewership, with ~20% of anime-watching accounts accounting for a disproportionate ~80% of all hours watched. Only a few "seasons" of Naruto or One Piece have competitive viewership with the Netflix originals released during the winter and spring seasons, but in aggregate, these titles have a dominating presence.

The chart above represents only the 30 most-watched franchises out of a few hundred, but they each represent an order of magnitude more consumption than the median anime found on Netflix – another way the Pareto Principle plays out in the world of streaming. I've found that some of the most invested anime fans have the most skewed perspective of popularity, with a key reason that the anime most commonly discussed online are not representative of the most-watched in aggregate. Kuroko's Basketball had nearly triple the views as fan-favorite Violet Evergarden in the first half of 2023, even when accounting for the difference in episode counts between them. DEVILMAN crybaby may have been a sensation in 2017, but it's not even among the top 100 most-watched anime on a per-episode basis now, with roughly 338,000 "views," as Netflix would define them.

One of the major trends Netflix calls attention to in their summary of the data is the "value of licensed titles," something that came to particular attention in the public discourse over the summer when the 2011 USA drama of questionable quality Suits topped the streaming charts provided by Nielsen. A reasonable 45% of Netflix's viewership comes from licensed titles, with over 500M of that coming from just three shows: The Walking Dead, Breaking Bad, and Suits. Netflix smartly frames this as a value-add that they can bring to give a longer life to popular shows, just as syndication did in the prior era of television.

With anime, though, they're relying on licensed titles to a much higher degree than the rest of their catalog. If one considers the titles that Netflix licensed as exclusive outside of Japan, but that simultaneously aired on linear television in their native country, only 7% of anime by hours watched are "Netflix" anime. Broadening this statistic to all anime considered "Netflix Originals" by the streamer, the number is still a far cry away from the platform average: 67% of viewership is still third-party content.

Particularly in comparison to other mediums, these data represent that Netflix is riding off of the success of anime rather than the other way around. Netflix exclusives and originals in the anime space don't see the same kind of attention as the Stranger Things and FUBARs of the world, and so audiences are drawn to the franchises that are getting promoted elsewhere. The Toei Animation machine has spent the last quarter of a century building up One Piece as an anime franchise. Platform-agnostic publishers like Viz Media have similarly spent more than a decade making sure anime fans continue to religiously rewatch Death Note, InuYasha, and Naruto no matter what streaming service carries them: a slightly more salient fact considering many of their titles left Crunchyroll at the end of last year. The critically-acclaimed Vinland Saga is available on every major anime streaming platform, and benefits from Netflix having the most popular video player of the bunch. The massive grassroots audiences that made Demon Slayer an organic-driven hit in 2019 – Netflix picked up the title only after it had gotten a shoutout from streamer Ninja.

In English-speaking territories, Netflix does not regularly promote anime on its main "Netflix" channels on social media, leaving it to the "anime ghettos" of their much smaller Netflix Geeked accounts, and to an even lesser degree, on Twitter and YouTube exclusive Netflix Anime channels. The first time Netflix's Facebook or Instagram had acknowledged an anime for English-speaking audiences on its main accounts since 2021 was for the Scott Pilgrim series, which has seen more promotions than all of the other Netflix Original anime released this year combined on the Netflix Geeked accounts.

Of the five anime franchises driving the most viewership on the platform in the first half of 2023, none of them have been mentioned on public marketing channels…but they have been spotted as part of acquisition campaigns in programmatic advertisements around the web. For as well-received as many anime tagged as Netflix Originals are, the biggest marketing push the streamer could provide would be making new titles front and center for anime-curious viewers, but based on the low performance of these titles compared to even less popular franchises whose audiences have been built-up elsewhere, it seems that's not happening as much as it could…except for live-action One Piece. Beyond hitting #1 on Netflix in nearly 100 countries, I'm curious to see where it lines up in the 2H 2023 data, especially considering its ubiquity in Netflix's marketing throughout the year.

While this merits a complete examination of its own, the year-over-year Nielsen Bookscan data released to ICv2 subscribers regarding manga sales are also worth paying attention to as an additional piece of evidence to understanding who benefits most from Netflix publishing an anime. Outside of the most popular series tier, manga sales are much more correlated with online community discourse and promotion than availability. It's practical knowledge that a manga that receives an anime will see a significant jump in sales that year – but the degree of that jump is based on many factors. While I have not built a regression model to account for possibly conflated factors in many years for this, the trend I've generally seen is that Netflix Original anime does not get as much of a bump as titles released elsewhere. While I've personally attributed this to the "binge" release model of Netflix anime, which negatively impacts the scale of conversation around a title, the lack of promotion of anything that's not at the level of Scott Pilgrim may influence this as well.

This conversation surrounding the impact of the typical Netflix anime release can be complicated in a few ways, however. A larger trend is not absolute, and there are plenty of counter-examples that are of note. One I would like to call attention to in particular is Hajime no Ippo, which dropped in several dozen countries like India, Indonesia, Japan, the United Kingdom, Brazil, and France (excluding the United States) on the first of January 2023.

And the impact was incredible. On a total views basis, as discussed earlier on Anime News Network, Hajime no Ippo was the #2 anime during the time period using Netflix's method of aggregation by season. Even when accounting on a per-hour watched basis, Ippo just barely misses the top 10 anime list for Netflix, an impressive feat for a series over 20 years old without a manga localized into many languages, including English. But beyond the impressive viewership, Ippo became a topic of massive interest around the world.

Much has been said about the boom of anime-inspired titles coming out of platforms like Netflix in the last half-a-dozen years or so. While this trend is not new – anime has served as a significant source of inspiration for animators the world over for generations, and famous titles in this vein, like Nickelodeon's Avatar: The Last Airbender continue to rank competitively with seasons of Friends and Suits on Netflix – it was the straight-to-streaming revolution of the late '10s that gave the comparatively-cheap adult animation medium an opportunity to explode onto the scene. Netflix, without a deep Rolodex of anime producers in the earliest years of its Originals program, thus started to use the battle-proven "anime" label much more loosely for titles made entirely outside of Japan.

It was somewhat strange to see, then, that one of the most beloved titles in this vein, Castlevania, does not seem to have the rewatch equity of anime classics like Death Note or InuYasha. Even Arcane, which topped 2021's years' best lists among anime fans notable and otherwise, is shown with these Netflix data to get revisited about as much as anime from the same time period like Komi Can't Communicate, JoJo's Bizarre Adventure: Stone Ocean, and Shaman King. This effect is even more dramatic in less popular anime-inspired titles. Dragon Age: Absolution aired around the same time as the anime Romantic Killer and the critical dud The Way of the Househusband: Season 2, but saw a fraction of the viewership of either of the Japan-based titles.

This is not to talk negatively about the merits of these titles – indeed, I've seen all but Blood of Zeus and thoroughly enjoy them – but only to note that, of all of the Netflix Originals considered anime, the ones that are the "sparkling wine" to anime's Champagne tend to get a larger share of the promotional budget and take up supersized roles in online discussions, particularly on news sites and YouTube discourse. Additionally, these titles were touted for their connections to video games or Western mythology, upping the accessibility and, therefore, the total potential audiences – the idea being that anime-influenced titles may be an easier sell to a non-anime watcher than something that comes from what is, to the majority of Netflix viewers, a more foreign culture. I would have thus expected their rewatch equity to be significantly higher than anime of the same general release window of similar popularity, but that was not the case.

Because the three anime cited here are all in Netflix's largest territory of North America – and none of them are in the top 30 IP of anime – I feel as confident as the data allow me to say that anime is not the concern it may have been in the minds of Netflix executives ten years ago when they started to seek out anime-inspired IP. I strongly believe that Netflix should continue to make/support/acquire adult animation like Castlevania and Arcane for a million reasons, but I would consider not calling them "anime." These titles likely benefited from the association, but you can use plenty of other curation tactics to represent the relationship and inspirations without misappropriating the term.

For Netflix, Asia is the future. And that future is anime.

In the first quarter of 2023, more than 80% of all subscriber growth came from APAC countries. In many places, like Netflix's home in the United States, the net subscriber count actually dropped. There have been notable increases in home-grown programming and paid media in Asian territories, and campaigns like cutting the price in Indonesia by almost 50% to make up for the crack-down on account-sharing earlier this year represent Netflix's continued ambitions for APAC.

While the data provided by Netflix are quite the blunt resource, without any cross-tabs or considerations on a by-country basis, by examining Netflix's "Top 10" microsite, we can make an approximate popularity ranking, understanding that it will likely inflate any of the four categories that are relatively less competitive (English TV, English film, non-English TV, non-English film).

Visually, two trends are impossible to ignore in this top 10 of top 10s: Harry Potter and anime. As I tried to hedge above, this methodology is even more imperfect than pulling from the aggregated data provided by Netflix. Still, it's impossible not to be impressed by how vibrantly these two categories stand out: only a pair of inclusions don't fall into either group. Anime simulcasts, in particular, lead the pack, counter to the worldwide list where Demon Slayer's first season (from 2019!) is tied for the lead as "most frequently in the top 10" for the first half of 2023. Unlike with Netflix's typical exclusively-licensed anime release, which sees the streamer hold back a title until the show's completed a run on Japanese television to dub it into nearly a dozen languages, Netflix Japan releases currently-airing anime concurrently with linear channels. This weekly release schedule has clearly made some sort of impact in Japan for Jujutsu Kaisen and Zom 100: Bucket List of the Dead to headline a list of their most popular programming in the territory.

India, another key growth region for Netflix, also featured Jujutsu Kaisen as one of the top 10 most-frequent top 10 titles. In a recent celebration of their success in the region, a Netflix press release cited Japanese-language content as a key element. It's not just Netflix who's noticed India as a growth region for anime, though – anime-focused rival Crunchyroll has spent 2023 with India at the center of expansion, with then-COO Brady McCollum eying what he called the "second largest" anime market with new Hindi dubs and updated pricing, among other efforts.

A second anime can't be found on the list derived from top 10s until you get to Demon Slayer ranked in the 70s (tied with Rick and Morty), but that doesn't mean Indian Netflix fans aren't increasingly watching anime as well: the six million plus hours viewed of new titles like Campfire Cooking in Another World with My Absurd Skill, which wasn't released in India in full until the last week of the period covered by these data, shows that for territories where new anime is published week-by-week in line with Japanese TV, anime is being consumed above its weight-class.

To that point: using the "views" metric as it's calculated by Netflix (total hours watched divided by hours available, or how many times the show theoretically was watched all the way through), we can find some impressive penetration rates for anime in the APAC region, where many simulcasts are exclusive to. Conservatively, 6% of the 40M APAC subscribers on Netflix are watching Jujutsu Kaisen, with actual numbers likely higher due to the typical completion rate for the top tier of streaming shows tracking in the 40-60% range, which would suggest as high as a 10-15% statistic being reasonable. The numbers for Zom 100 and Spy×Family's second season are not too far off, with 4.8% and 3.8% engagement rates, respectively.

One of the biggest headlines concerning Netflix in the world of anime in the last few years was its acquisition of 23 films from Studio Ghibli, the most famous anime production house in its native Japan and internationally. While the tech giant wasn't able to claim the rights to these films in their native United States (Max, then HBO Max, would scoop of those desirable rights), the other estimated 72% of Netflix subscribers still had access and watched over 82 million hours of Ghibli wonder during the first half of 2023.

Compared to Google search interest or Wikipedia page views, these lists align fairly well – the Wikipedia metric has a correlation coefficient of 0.9, for those curious! However, the more adult-oriented films from the catalog tend to have more interest on the web than on the Chromecast, with Princess Mononoke and The Wind Rises performing significantly worse on Netflix than other metrics would suggest, and the more child-friendly Ponyo sees a pop in the ratings for its rewatchability.

Howl's Moving Castle's longevity in interest is a point of constant fascination to me. While I consider it one of the most rewatchable of the Ghibli catalog, critically, it's near the bottom of the list. And yet, the only title with more interest on Google or Wikipedia is Spirited Away, and it's the clear winner in terms of Netflix views. That said, were critical recognition more closely tied to viewership, we would see Whisper of the Heart or Only Yesterday in the top 5, and Miyazaki's best film, Porco Rosso, would not be his worst-performing of the bunch.

I have to wonder how much effect the Japanese audience had with these films, or if they even make up a disproportionate viewer base relative to the rest of the world. If the Ghibli catalog were exclusive to Japan, I would wager Castle in the Sky, a favorite for rewatches on linear TV in its home country, would be much higher than 9th place.

If you'll allow me the indulgence of returning to the Pareto Principle for a moment, it's fascinating – though not at all surprising – to see such a massive distribution in viewership across these films. Studio Ghibli co-founder Isao Takahata is a beloved director in his own right, but his best-performing film on Netflix, The Tale of the Princess Kaguya, only drew 10% of the eyeballs of his contemporary's number one film. Whereas the gap between Toonami's most-watched and least-watched titles, for example, was typically no more than 20-30%, the delta in viewership between Totoro and My Neighbors the Yamadas is more than an order of magnitude.

This example exposes one of the challenges with streaming: the seemingly endless choices one has for consumption means that, even with careful curation, most of the titles brought to the platform will die on the vine. No matter the quality of the content – and it's hard to argue that even the lowest rung of Studio Ghibli fare is anything less than exceptional – there's not enough time in the day or interest to fully exploit any catalog. From your perspective, having Nausicaä of the Valley of the Wind and The Wind Rises available for streaming may be non-negotiables regarding the Ghibli catalog. But for Netflix's sake, outside of the positive press of the initial announcement, they would have benefitted just as much to only get the top dozen films, even at the same price.

From the launch of its streaming service in January 2007, Netflix enjoyed more than a decade as the undisputed category leader for scripted programming in the English-speaking world. With this interesting position, Netflix could do what no major name in entertainment was able to do before in the United States: they could leverage viewer data for themselves. The asymmetry of knowledge advantaged the Palo Alto-based streamer in negotiations: how could producers, actors, or writers know their real worth when the title-by-title performance was obfuscated? Nielsen ratings, the industry standard for television, took a long while to come and then were rough estimations in an era where perfect data did indeed exist.

Finally, addressing over a decade of criticism from creatives, producers, and analysts alike, Netflix began to reveal its top 10 lists in 2020. Then, this December, likely concerning last year's Writers and Actors strikes, Netflix finally gave us something to work with to understand just what is happening on the world's biggest streaming platform.

Strategically dumping the data as a barebones spreadsheet without any delineations for historical comparisons, content type, country of origin, territories for availability, week-by-week viewership, or demographic cross-tabs means that meaningful analysis would be challenging. By only releasing the data from January to June of 2023 on the week of Christmas means that meaningful analysis is less likely. But as someone who was sending regular viewership analysis to the Crunchyroll C-suite a decade ago and has refused to give up the habit even when internal streaming data was no longer available to me, I hope I was able to bring some insight into the mess, and answer out some of the many questions proposed by the data.

---

Miles Thomas Atherton is the CEO and founder of White Box Entertainment, a consulting firm focused on the promotion of anime in the English-speaking world. He was previously the Chief Marketing Officer of Anime Ltd., a boutique anime distributor in Europe. Prior to that, he spent nearly a decade at Crunchyroll across a half dozen roles, ending his tenure as the Director of Social Media, Editorial, and Curation.