Category: Visual of the Week

Visual of the Week – Netflix Top Ten Series by Total Minutes Viewed

One of the big questions every quarter is whether or not Netflix will hit its quarterly subscriber growth estimate. This leaves analysts scrambling to read the tea leaves from app downloads and what not to try to figure out if they are on track or not. Tomorrow (Tuesday Oct 20th) will tell us one way or another.

My contribution to this is to note that often having valuable content drives adoption and usage, and hence subscriber growth. This sounds relatively benign, as a statement, but has profound implications for whether or not Netflix has a “moat”. Or indestructible defensive position. If Netflix is simply another content creator whose success depends on producing good content, well they’re as mortal as the rest.

So my data of the week is one look at content. There are lots of ways to do this (Hedgeye Communications used Google Trends to brilliantly show this in a newsletter last night), and the data set I’ve been playing with recently is Nielsen’s top ten streaming shows each week. Here is the total minutes viewed for Netflix from Nielsen by week for the United States from end of March to present, with a big gap:

Screen Shot 2020-10-19 at 10.45.22 AM

And here’s the table for folks who want the raw numbers. I also included how many titles they had in the top ten.

Screen Shot 2020-10-19 at 10.50.29 AM

Ramifications/Thoughts/Insights

– Hits drive the ratings. Again, this is so obvious and has been true for decades it sounds silly to restate it, but in the “digitally disrupted” world, we have to relearn old lessons.

– Man, look at March! It turns out Tiger King and Ozark drove huge viewing to the platform. Almost 2.5 times more viewing to the top ten.

– Likely this means that content in Q2 was much more popular than Q3. Tentatively, this would portend a drop in US and Canadian subscribers in the next earnings report. (Some application sign-up and download data is presaging this outcome as well.)

– Yes, the last three weeks have seen 1 to 2 non-Netflix shows make the list, making this time series not totally apples-to-apples over time. That said, I ran the list with the Amazon and Disney shows, and it looks mostly the same. Meaning that the top 3-5 shows tend to account for most of the viewership, so having one or two small shows with 500 million minutes viewed doesn’t radically change the numbers.

– I have a ton to unpack for these Nielsen numbers to learn/prove more insights about how content behaves on Netflix and other streamers. (Trust me, I know a ton from my previous role about how the content behaves, but I want to show/prove it in the data. And for the most part, it behaves the way I expect.)

– Long term, I hope to compare Nielsen’s data to Netflix’s Top 10 data (provided by Flix Patrol) to Netflix’s own datecdotes to Google Trends and more, but that takes time. Also, if you have a data set you want to share, my email is on the contact page!

– Specifically, I’m on the look out for the missing weeks of top ten data from this Nielsen data set. Someone sent me the April and May numbers, I’d love to have March, June and July if anyone has them. Your confidentiality assured.

Visual of the Week – Netflix Produces 3.3% of Its Top Streaming Shows

Over the last six weeks, Nielsen has released a top ten list of the most streamed series/films by total minutes viewed. I’ve been taking this data and adding a layer of detail on top, specifically who produces and who distributes what shows on Netflix, Amazon and so on. Now that we have six weeks of data, we can start to parse some insights. 

(Thanks to Kasey Moore of Whats-On-Netflix for saving the Nielsen lists for me.)

The visual of the week for this week is just a look at who owns what in the streaming wars. Of the 52 billion minutes of TV tracked by Nielsen, here’s who produced what and what shows they own (by parent company):

Screen Shot 2020-10-14 at 8.48.12 AM

And here is the table if you want to see how the sausage is made.

Screen Shot 2020-10-14 at 8.48.19 AM

Now some insights/details.

— Some shows were co-productions, in which case I split ownership between the two companies. Meaning, the percentages won’t add up to 100%, since some shows were counted in both owners’ percentages.
— Two films/series were not on Netflix (The Boys and Mulan), but that only boosts Netflix to 3.3% in “Netflix-only” series.
— I focused on major producers only. The traditional conglomerates. Usually, any of these shows has a bunch of smaller producers attached; I counted who likely paid the production budget.
— I use Wikipedia to determine producers with another source who tracks everything on Netflix by copyright ownership. The closest call was Umbrella Academy, which is also co-distributed by Netflix. However, NBC Universal owns the copyright outright so Netflix will not own it in perpetuity. Moreover, they aren’t listed as a producer, so didn’t make this list.
— That’s really what I’m trying to get at by focusing on producers versus distributors. The idea that who “owns” a piece of content so they can eventually maximize the value of it.
— I can hear the criticism, “Well this list is mostly library content.” And that’s true, but not 100% correct. Even the list of first and second run content by Netflix is almost entirely licensed content.
Seriously, don’t use “Netflix Originals” as a descriptor. It really doesn’t capture the key parts of ownership in content.
— I will run this same analysis on the FlixPatrol data for Netflix’s Top Ten list, but I haven’t had time to do that yet.

Bottom Line: A core thesis of Netflix’s content spend has been to build a “moat” of original content they own in perpetuity. Clearly they have a ways to go before they truly own their content.

Visual of the Week – Is Netflix is a (More Watched) Broadcast Channel?

We had a fun bit of data dropped via Nielsen in August which allows me to update my most popular article of the year, “Netflix is a Broadcast Channel”. Nielsen let us know how viewership looks through the Coronavirus lock downs as of August 2020. Here’s the original 2019 data and the update:

Screen Shot 2020-09-02 at 9.19.51 AMSince I promised this is in visual form, here’s the stacked bar charts…

Screen Shot 2020-09-02 at 9.20.15 AMQuick Insights

First, is this statistically significant?

Yes, tentatively. It all depends on what your confidence interval is, but with their panel of about 1,000 folks, Nielsen can have a margin of error either direction of about 3%. This is right on that border line.

That said, why use a 95% confidence interval? If you use a 90% confidence interval, than year we’re reasonably confident Netflix saw a bump. I’d add, everyone else was flat and next grew or declined. (Except for Disney+, which wasn’t on the platform last time.) That’s hard to interpret as anything but good to great news for Netflix. Contrariwise, if you want 99% certainty, then this is firmly within the margin of error.

So we’ll see how this number grows, but I’m inclined to think it measured a real trend.

Second, why not update your Primetime chart from last time?

You mean this one? Image 1 - Estimates

If this were extrapolated to Primetime, then Netflix has exceeded even CBS and taken the top broadcast spot. (They’d be at 8 million primetime viewers if we used the same math from August.)

First, and simply, I don’t have the linear TV viewing numbers to compare. Broadcast ratings could have increased by a similar rate, so it wouldn’t be apples-to-apples. 

Also, while the 3% increase in Netflix viewing is good, and the 7% surge in streaming video is even bigger, I’m skeptical that viewing came during primetime. Sure, folks can’t go out so TV viewing is likely up across the board at Primetime, but the 7% surge in streaming likely came from elsewhere. I see two options.

Option 1: People watching TV during the daytime. The notable thing about coronavirus is that everyone is sitting at home streaming during work. (Are those two things incompatible? I think so, but that doesn’t mean it’s not still happening.)

Option 2: Children. The other group that is probably streaming even more than ever are kids. And children. And teenagers. Again, not during primetime, but throughout the day. And my initial comparison was about primetime viewing. That’s why Disney+ went from not existing last fall to getting 4% share of streaming.

Visual of the Week – The Biggest Broadway Musicals of the 2010s

Well, the race is on to get your Broadway musical. First, Disney set the standard with Hamilton. Now, Netflix is fast on their heels, getting the rights to a Princess Diana musical.

This got me wondering, especially the Hamilton news, about how big was Hamilton in the 2010s? Was it the biggest live musical show in the world in the 2010s?

Fortunately, Wikipedia has us covered with data from The Broadway League, so here’s a chart that didn’t make it into my Decider piece. When you look at “Per show” revenue, Hamilton was a popular monster that has few peers.

IMAGE 1 Revenue Per Performance

Some quick insights:

– First yes, winner take all. It shouldn’t even be a surprise at this point. Which was Wicked, until Hamilton, which will likely earn twice over however long it’s lifespan runs. Moreover, the Disney+ platform will likely only boost long term receipts as more folks want to see it in person.

– Second, here’s the table if you want the data yourself. This is sorted in total Gross Revenue to provide a different look.

Image 2 Table

– Third, if you look at “Revenue per Year”, you can see another look at just how much Hamilton was making. 

IMAGE 3 - Revenue Per Year

HBO U.S. Subscribers Over Time – Visual of the Week

Inspired by AT&T’s release of HBO Max “activations” and total HBO subscribers, here’s a timeline of HBO subscribers and HBO+Cinemax subscribers over time:

IMAGE 1 Chart

If you’d like to see that in table form, along with some financial numbers, here you go:

Screen Shot 2020-08-03 at 11.13.55 AM

What about total subscribers? Again, we only have data from 2011-2017, but here you go:

Screen Shot 2020-08-04 at 9.35.20 AM.pngSome quick points and explanations:

– This data was cobbled together from random leaks, Time-Warner’s annual reports and AT&T’s earnings reports. (Links here, here, or here for leaks and here for Statista.) If you know of any I missed, send them my way.

– There is a chance that the reason AT&T didn’t release 2018 numbers for HBO, in addition to the merger being ongoing is because their numbers during Game of Thrones season 8 last spring were higher than they are right now. We don’t know because of gaps in the data, but looking at 31.4 million HBO subs alone in 2015, then considering they had 5 million digital only subscribers in 2017, that could easily have been higher than the current 36 million.

– With only 3 million subscribers having “activated” HBO Max, that service has a lot of room to grow. I’d compare that to the early days of Amazon Prime Video; it too had a lot of time to convince people to try it out, but also the free cash flow to wait. Math and explanation of activations over at Variety.

– If you want more on the financials of HBO, and discussion of their subscriber counts over time, read my article at Decider and the Director’s Commentary.

– Comparing multiple subscriber counts with different definitions reminded me of this table I built for Netflix last fall. I’ll update it this fall with yet ANOTHER definition for Netflix.

Visual of the Week – The Performance of Netflix Top Films Over Time

(This is a new feature from the Entertainment Strategy Guy. It’s a weekly “visual of the week” that will come out every two weeks. If you like it, consider sharing it on social media, just toss me credit.)

The big Netflix news last week was their earnings report. But the most fascinating story for a data wonk like me was Lucas Shaw’s scoop on the top Netflix films by viewership (2 minutes of a film) of all time. With this scoop, I’m up to 30 different “datecdotes” on Netflix film viewership over time.  

This visual of the week has two different presentations. First, Netflix raw viewership overtime, by quarter:

NFLX visual 3

(Details: This is by my estimates for 70% completion of a film by Netflix subscribers. This is global data. Time period is Q4-2018 to Q2-2020.)

Of course, that doesn’t account for the size of Netflix, so here’s the percentage of viewership:

NFLX visual 2

(Details: This is by my estimates for 70% completion of a film by Netflix subscribers divided by subscribers at the time. This is global data. Time period is Q4-2018 to Q2-2020. Constraint: Only films getting over 20 million subscribers are included.)

If you want more details on Netflix feature film performance, I started a big thread on it on Twitter.