Netflix Originals Movies: An Exploratory Data Analysis

Netflix Originals 📽

• What’s a Netflix Original movie?

Movies that had their original production commissioned by Netflix. They include documentaries, comedies, one-off specials, and a whole load of regional programming.

• How many movies are there?

I’ve got 584 in my data set, so let’s say 584. The first went live in December 2014 and the last one I have launched on June 1, 2021.

• What data have you got, then?

Well I scraped the Wikipedia page, added in viewer ratings from IMDb, then cleaned/formatted that into a dataframe in Python.

All in all, we’ve got:

Title, Genre, Language, IMDb Score, Runtime, and Premiere Date for all 584 shows.

Question 1: Is Netflix releasing more Originals every year?

That’s an easy one: Yes, clearly.

Netflix is estimated to have spent over $17 billion in cash on original content in 2020. It will spend at least $26 billion a year on content by 2028, BMO Capital estimates.

2021 has been affected by some major delays but Netflix is still scheduled to release more Originals than it did in 2020.

I’m most excited for Cowboy Bebop, slated for a fall 2021 premiere.

Question 2: For which genres is Netflix creating most original content?

Documentaries, followed by comedy and drama.

But is that split changing each year?

I’ve used a different colour palette here in the hope of making this legible.

Key trends:

• A significant increase in drama content every year.
• Documentaries were the most popular genre for Netflix every year - until 2020, when they created more comedy content.
• It looks like drama is on course to be the most common genre this year.

Question 3: Is the content any good?

We had a look at this question for all Netflix content in the fan-favourite Netflix special of hi, tech. We also analyzed content from their rivals in that one.

This time, I have assessed Netflix Originals in more detail.

I’m using IMDb scores since there are lots of them for almost every show.

Are they a reliable metric? Well they’re the best we have, although The Trial of the Chicago 7 has a score of 7.8/10 and that was a reprehensible use of everyone’s time.

Still, let’s take a look at the distribution of Netflix Originals by IMDb Score:

David Attenborough’s ‘A Life On Our Planet’ earns a well-deserved 9/10, while ‘Enter The Anime’ props up the rest with an execrable 2.5/10.

A recent review of ‘Enter The Anime’ says:

I love anime, but this movie made me wanna shoot my brains out after 15 minutes.

Or how about:

It has absolutely no value to anyone that is familiar with anime and even less for people wishing to dive into the medium.

I’m starting to think 2.5 was generous.

Let’s look at the top 10:

Documentaries and music shows are highly regarded, it would seem.

The mean score across all Originals is 6.3, which generally means they’re of acceptable quality.

Question 4: Does the quality differ by genre?

This chart will show us the range of scores within each genre:

Documentary has the largest range of scores of all genres, including a collection of content that scores 7.5/10 or above. Netflix has evidently figured out how to churn out mediocre drama and comedy content that people will tolerate.

Horror content has a low average rating, unsurprisingly.

Animation content is thin on the ground but when it does appear, people seem to like it. We need to factor in the niche appeal of these shows and the small sample size, however.

Question 5: Is Netflix Originals content getting better?

I’ve gone for a box plot here both to mix it up and to help show the range over time.

The instant reaction to our question is: No, not really.

Netflix is making a lot more content but it is not reliably of a higher “quality”, based on viewer feedback. There are fewer duds in 2021 so far, but the year is young(ish).

We could say that the range of scores is shrinking a little since 2019, albeit not to the extent that we can conclude Netflix has cracked the formula for making dependably excellent TV. If anything, the mean creeps ever closer to 6/10 - the hallmark of mediocrity.

Let’s look at this from another perspective.

Question 6: In which languages does Netflix create original content?

The vast majority of Netflix Originals are in English:

But is that changing over time?

Yes, very much so. From a slow start, Netflix is investing much more in non-English langauge content.

To put this dramatic shift into perspective:

In 2021 to date, Netflix has launched more shows in non-English languages than it has in English.

To demonstrate that in a pie-shaped form:

“Non-English language” is a very broad category, of course. Yet this is evidence of a purposeful transition in Netflix’s strategy.

It wants to increase its reach in international markets; but it also wants to create non-English content that proves popular in English-speaking markets.

As we have discussed before, shows like Lupin and Casa de Papel have been huge hits in the US and UK. In a crowded competitive market, Netflix is using this as a way to fend off the pure, Anglophone threat from Apple and Disney.

Netflix is even developing software to help international collaboration, as detailed in this post.

Question 7: Is this non-English content any good?

Well it has a mean IMDb score of 5.9, so it is “worse” on average than English content (score of 6.4/10).

Only two non-English language Originals have an IMDb score of 8.0 or higher:

Emicida: AmarElo - It's All For Yesterday (Portuguese)

The Three Deaths of Marisela Escobedo (Spanish)

But we should expect to see that cohort grow as Netflix increases its investment in non-English content.

Below, we can see the top 10 languages for Netflix Originals content with the IMDb ratings for all titles in each language:

There are some notable variations in the ranges here. Netflix is yet to make a French movie that can even scrape above 7/10, par exemple.

Additional points of interest

(But not interesting enough to end up in a chart.)

• Over 70% of Netflix Originals are released on a Friday.
• October is by far the most popular month for releasing content. (This also helps to explain the relatively low number of shows released in 2021 so far.)
• Mean runtime: 94 mins.

1. Shows are getting slightly longer each year, on average. That makes sense from an engagement point of view, but doesn’t really tally with your TikTok-addled modern viewer.
2. There is a very weak correlation between runtime and IMDb scores, as you’d expect. Shows can be short and bad or long and bad.

• Short animations seem to score higher than average, yet we should factor in their niche appeal.
• Shortest Original: Sol Levante (4 mins)
• Longest Original: The Irishman (209 mins) (Did anyone watch the whole thing? Be honest.)

🤔 Main points

• Netflix is creating a lot more original content each year. From just 27 in 2016, it increased to 179 in 2020.
• Documentary is the most common genre, but drama and comedy are catching up.
• The quality isn’t getting better - there are just lots more shows. Netflix hasn’t cracked the formula for creating reliably excellent TV. If anything, it has settled for tolerably mediocre - with the occasional, pleasant surprise. More importantly: We, the viewers, have settled for this deal.

1. Attenborough is the best.
2. Enter the Anime is the worst.
3. But this measure of “quality” (IMDb scores) is dependent on the opinions of viewers. Standards, tastes, and preferences all change markedly over time. Still, it’s the best metric I could get my hands on. And I do believe that Enter the Anime is rubbish, based on the reviews I read.

• Netflix has rapidly grown the selection of non-English content

1. It is now creating more non-English Originals than it is English Originals. That seems like a big deal.
2. Hindi, French, Indonesian, Turkish, and Portuguese are the fastest-growing languages for new content.

And I think that’ll do us for this data special. Give it a share, if you think it’s better than Enter the Anime.