In this unique episode, Michael Beach is joined by the anonymous industry insider known as Entertainment Strategy Guy (ESG) to discuss how algorithms dictate streaming viewership, the evolution of streaming economics, and what content spend may look like going forward. Watch our latest Screen Wars Thought Leader Interview here and read the full transcript below!
MB: Entertainment Strategy Guy, welcome to Screen Wars.
ESG: Hey Michael, I’m glad to be here.
MB: I’ve been a huge fan of your writing for a long time. You’re the person that we have quoted the most in our newsletter over the past seven years. I’ve always found interesting the point of view that you’re able to offer, writing from an anonymous account. What is your background on thinking about that, and how’s that worked out?
ESG: First off, I’m also a huge fan of your newsletter. I’m constantly pulling charts or finding new data sources from it.
I started writing anonymously after I was let go from another streamer. I knew I wanted to write about strategy and entertainment because I thought there weren’t enough people analyzing the entertainment industry and the strategies behind it. The odds were that I wouldn’t be able to make the writing thing work, and I would be back to work in the entertainment industry.
I felt that if I wanted to write about entertainment with as much freedom as possible, it made sense to do it anonymously, so that if I didn’t get any traction after six months, I could go back. Along the way, it ended up working out. The surprising thing is that whenever I throw up a poll on this, asking, “Should I stay anonymous or reveal who I am?” It’s 50/50. So half my customers love it, while the other half don’t. So I’ve kept doing it for that reason.
I also work on a lot of creative projects on the side, including some consulting projects with other people. And now that I write so heavily about streaming ratings as well, I think the anonymity piece protects some of those projects. I can write freely about what I think is working or not working in the industry. People always ask me about the anonymity thing the most, but it has worked so far, and I have been going with it ever since.
MB: It’s been interesting. It seemed like you wrote on your own for a while, and then moved to The Ankler a few years ago. How has it been working with Richard Rushfield and the team?
ESG: Richard Rushfield has been great, actually. The entire team has been fantastic. As soon as I discovered Richard’s newsletter, I definitely felt him and I had the same approach, even if we came from different backgrounds. In some cases, we are skeptical about what the studios are saying, and dig beneath the surface. So I was happy to join him.
Richard told me the direction they were going before they launched the big Ankler redo and expanded their operation, so I was happy to start writing a column for them. And for me, it’s been the best of both worlds since I have my newsletter that continues to go out via Substack, but I also get exposure to the Ankler’s great audience and get to work with Janice Min, who’s just a fantastic editor and great sounding board for ideas.
MB: That’s amazing. Your insight on streaming ratings is well-known. It seems like you realized, before anybody else, that the data was available to make consistent analysis of streaming ratings. And I’ve always been interested to ask, what is the hardest part about putting that together? What pieces are still missing?
ESG: From a personal standpoint, currently the hardest part for me, is dealing with “the known unknowns.” I started out polling Nielsen’s top 10 list. I’d like to call it, “The Streaming Ratings Era,” when it started in 2020. COVID happened, and all of a sudden, a lot of the companies that were working on data hopped into the field at the same time. The incident that generated the most traffic was the release of Mulan on TVOD, and everyone wanted to know how well did it do.
I noticed that, all of a sudden, there were five different companies, including Nielsen. Then, they started publishing their top 10 lists. And my rule of thumb is always the content is about half of the entertainment battle. That’s what you’re trying to win. So I was all of a sudden like, “Now we have the data, we can actually look at streaming.” But that that was still only a top 10 list.
And for someone who doesn’t work for a major company, or even for some people who work inside a major company, the big unknown data point is all those other data points for all those other competitors. And even the people who provide that (like Nielsen or Samba TV) also have other blind spots, which gets to your second point: No one actually knows what happens on the other side of the screen. Is it one person watching? Is it two people? Who in the household’s actually watching? And that’s where they have the test panels.
But the problem with the test panels in the living room is tying that data over to mobile devices, like phones and laptops, and trying to figure who is watching. This is still the biggest piece that companies may not solve for a while. In the future, user-generated spaces, like YouTube, and TikTok, will produce even more pieces of content. Translating all those, and figuring out what works across every single thing, is such an issue that we may never solve it.
MB: I find it interesting because we started our core business on the advertising side, and we’re using similar data sets. How do you feel, on the content side, about how they value everything? With the way you put the ratings together, do you feel that they have a view to properly value the content, or is that still a black hole? Because, in the advertising side, it’s probably not getting properly valued.
ESG: I don’t think we’re there yet to have it be properly valued, but we’re on the way there. I don’t think my streaming ratings reports alone could provide the proper value, because we don’t have enough public data points to provide those data points. I also think that once we have multiple sources at every company, that will provide better double checks. So, instead of having just Nielsen and a single point of failure, you can have multiple sources to see if they’re all roughly telling a similar picture.
That said, I am optimistic we can get to what content is valuable simply because we have so much data that, at some point, it becomes overwhelming. We know what the hits are, we know what the slightly lesser hits are, we know what the pieces below those are.
As much as things have changed, I tend to revert to the basic principles with entertainment. This is a hit-driven business, and if something’s popular in one metric, it tends to be popular across the board. The example I love to use is Stranger Things. Stranger Things is probably Netflix’s most-viewed show in terms of total hours, but probably also with unique customers. Its Wikipedia page views are very high, so it’s a social media conversation.
All those things tend to be directionally, and it has a very high completion rate. You mentioned advertising, and I think that has its own different challenges, especially with attribution and figuring out what’s on what platform. Do you think we’re going to get better at telling what content works, and then valuing how much that actually drives value down the value chain?
MB: Yeah. I got three specific pieces of content that I’ve always wanted to kind of get your background on. The first, you were really early in calling the Netflix subscriber slowdown. Can you explain to me how you projected that?
ESG: I actually took that from a business school professor who, in our entertainment strategy class, a long time ago, looked at Netflix, and he applied the Bass diffusion curve to it. The Bass curve is a tool that says that new technologies have an adoption curve that starts out slow for a few years, picks up speed, and then slows down again. It is also called an S curve due to the way it looks.
When I was starting to look at Netflix, they started to have some quarters that were slightly slowing down, and I wondered, “Could we reapply this curve to Netflix using them as a stand-in for all streaming video?” And sure enough, the curve did fit pretty tightly. It still had some uncertainty, but the forecast that Netflix would get to 90 or 120 million subscribers in the US seemed unlikely.
Since Netflix changes their definition of US subscribers very often, I had to do a lot of data searching, and I had to do some estimates, but it ended up working out. And the actual number was right around 65 to 70 million subscribers in the US. Netflix later converted to US and Canada subscribers, but if you subtract out the 10% of the subscribers that are roughly from Canada, Netflix is right around 66 million subscribers in the US. That’s how I calculated that number. I am not sure if it would work for some of the other streamers because the streaming curve has been adopted now. But I’m monitoring those streamers. If I see subscribers slow-downs in the future, that’s going to be my guess for what happens.
MB: It’s great just because, even if you just get Netflix, they’re on the high end of the TAM (Total Addressable Market) for the US. It was just such a good insight so early, and will I keep going back to that the next time we get into this bull market, and people start making lofty projections. That insight was incredible.
ESG: I appreciate that. Whenever you hear buzzy numbers in any field, look up the Bass diffusion curve and look at it. In some cases, you can just eyeball it, and you’ll get the same thing. I also noticed this with Fortnite. I saw the slow growth, and then it took off. I looked at the same thing and said, “Is it headed to keep skyrocketing to a billion users, or is it going to top out?”
And while Fortnite has remained popular and has its user base, it definitely isn’t growing whatever year like when it was the Fortnite Summer. There were people wondering if it would get to 500 million, 600 million, even a billion active users, but I could tell it would slow down because it was so fast on the uptake. It’s a fun tool to use. Another I would relate to with Netflix, which is something I was on pretty early as well, is the churn game, which is everything with them.
Mapping things to cable, cable was an industry that didn’t have churn because if you wanted TV, you had to have cable. And with the streaming services, you don’t have to keep your subscription. And I was trying to quantify that churn issue, because I thought that was going to lower the overall TAM. And I think people disagreed. Thank you for that insight. That’s the one I try to hang my hat on because it worked out so well, and making predictions is a tough game.
MB: And another one, The Algorithm Is A Lie. I want to get your insight, because I think your intention was assuming that Netflix or a Hollywood Studio use data science to pick what content to green-light. But we also look at the content delivery to the consumer, where you’ve got algorithmic delivered TV. Do you apply it to both, or to one more than the other?
ESG: No, I absolutely don’t apply it to algorithms that surface content, with some caveats to that. I think certain social sites leverage algorithms even more strongly than certain streaming sites, and they definitely overhype them. For any of your readers who aren’t familiar, The Algorithm Is A Lie, it’s a dejure for a lot of critical reviews of streaming shows, especially for Netflix, that if they don’t like a show, they’re like, “I can’t believe the algorithm picked this show.” And the algorithm said to cancel a show within six hours, which has never happened because the algorithm wouldn’t even have enough data to make the predictions in six hours. It would give some trends, but not predictions.
It is true that the streaming services and all the studios have more data than ever at their disposal to help guide the development process, to help see what’s working, to help make some calls, and things like that. But there still isn’t enough data to actually forecast that a script is going to be a hit TV show. From a data perspective, the main reason for this is that there’s just not enough data points to be able to tell, because there are too many variables going into any individual TV project to say, “This thing’s going to be a huge hit,” or “This thing won’t flop.”
If studios had algorithms, they wouldn’t make bad shows, and nothing would flop, but that’s not the situation. They have more data than ever before, and this includes Nielsen ratings, box office, social media, internet traffic, all those things. People create content silos and deserve all the credit and blame for all content decisions. On the other side, is the algorithmic delivery of content, especially for social media platforms. Something like YouTube, especially because they have a great number of data points, since besides having millions of users, they have thousands of videos. Those are absolutely powered by algorithms that surface content to people and deliver those things, and they’re powered by advanced machine learning algorithms that actually work.
The one difference between social media and streaming services is that algorithms become even more important when you have millions of data points like YouTube, Facebook, Twitter, or TikTok, especially. For many streaming companies, the algorithms are important, but they don’t have nearly as much content to surface. So if they say, “Hey, this customer likes X type of content.” there are only 10, 20, maybe 30, examples that surface, and this is why people complain that there’s nothing to watch.
Plus, most of the streaming companies have their fingers on parts of the scale of those algorithms, though they don’t admit it. For example, putting something at the top of a homepage does a ton to drive traffic to it, which will boost its engagement. So automatically, it’s giving it a thumbs up. The marketing team designs all those top spots on Netflix or Disney+’s top row, usually based on what’s recently releasing, not the algorithm. So, the algorithms on streaming are very important, but they’re not quite as powerful as some of the algorithms out there in other websites.
MB: The American Viewer series you put out is my favorite. I reread a lot of this recently. I thought you had a ton of insight. One was about how Hollywood designs or develops creative for the coast, and that’s different from the overall American viewer. What key take away is in that project?
ESG: I love putting that together. I actually have a researcher who works with me, and he had the idea. We actually cribbed it from Matthew Yglesias, who had a series called “Who is the Average Voter.” So I wanted to do the same thing for the average TV watcher, with the one difference being that a lot of people don’t vote, but everyone watches TV. So it ended up being across the board. The main takeaway is that everyone watches a lot of TV, and that TV usage is pretty steady. That one was a long time ago. I’ll actually say, one of the interesting things is the demographics question keeps coming up, especially with ratings.
I get a lot of emails about that with people asking questions. I don’t have the data to dive into it, but it’s definitely one of those areas to dig into, and it’s a complicated one because you need to have a good enough sample size, but demographics is one.
There are a couple of the key takeaways that I thought were the most surprising. First one is social media use. It was amazing to me how big Facebook really is. And then, YouTube would be the other one. The question is, are they a video platform, or are they a social media platform? But Facebook, in particular, is still far and away the king, even though most people on the coast would say it was either TikTok or Twitter. Most of the chattering classes tend to use Twitter. But if I had to pick one social platform to only focus on for that conversation, it would be Facebook.
The other one is the prevalence of living room TVs. Many times we think that people watch a lot of videos on phones, but the living room TV still dwarfs them. You’ve written a lot about CTV. And this is where I would focus a lot of my energy compared to phones, because the viewing experience is just different. With a living room TV, you sit down, and tend to watch it. You might still be on your phone, or doing something else, but it’s like a centerpiece.
Where a lot of mobile viewing is either TikTok scrolling, or it’s watching a video on a webpage, which YouTube actually benefits from. So social media is a big one. The last take relates to the coast. Something I think was really fascinating was that we still underplay how religious a lot of America is, and that type of content hasn’t really been served. Every so often, you see a movie or a TV show like Rise Up that does really well at the box office or the ratings. And it’s because there are many religious people, who are very diverse, with different sets of beliefs. But that’s something that a lot of people in Hollywood tend to forget.
MB: Interesting insight. Another thing you talk about a lot is the business model around streaming compared to the linear TV side. A lot of this came to a head with the HBO Max and Discovery+ merger, and the changes they made. Do you see this evening out over time, or are the economics built into streaming incapable of creating the same economics that you had in linear TV?
ESG: I think they’re incapable of creating the same economics, especially if we look at the economics across the whole lifespan of an individual film or a TV show. Too many of the windows have disappeared on the film side to ever recreate the same upside, especially compared to the DVD boom in the early 2000s. You had something where films, if they could afford the marketing budget, could make their production budget on DVD sales. People were buying DVDs for everything, and then everything else was easy. When it comes to TV, my biggest concern is still the churn piece.
When you’re able to churn out, it just means that people are going to pick and choose what they’re watching at a given time. And right now, it’s much more likely that Netflix is at the best case for that, than Netflix is at the worst case. It’s going to improve over time. I think more people will cut the cord, but as those people cut the cord, they’re going to use a variety of free options or other streaming services. We are going to see some evolution, as Zaslav had mentioned, can we do a big bundle with all the streamers at a lower price point? I see the logic for that. But I also see that it is incredibly difficult to convince all these different studios and to figure out who is worth what, as they try and lower prices and offer a year subscription to people.
The economics are lower, and for individual TV shows, the syndication window just won’t be what it was. The second window, licensing, also won’t be what it was. The TVOD sales won’t be there because people will churn in and out. It’s going to be lower. I want to try to stay positive, so the one bright side is that I think the global opportunities will continue to be there in ways that weren’t before. All the traditional studios had, for example, some Latin American channels, or they partnered with European companies to sell their content.
And Netflix has really shown that with D2C you really can go global with the content, with the caveat that you often need to make a lot of local content for all those new places. If the Hollywood studios can grab more of that global share, that will happen. But on individual market perspectives, like in America, I don’t think we’ll see what we used to see, especially with the revenues. We won’t see the same operating income or profit margins that we saw back in the 2000s or early 2010s.
MB: In the last couple of months, we have seen in the news that are streaming companies are actually removing content from their library, and this caught a lot of people outside the core industry off guard. What does the decision-making for that look like?
ESG: This is one of those situations where I wish I could be inside some of the streamers to hear some of those conversations in real life. When I started in streaming, the idea of making a show and then pull it later was inconceivable. But I think there are a couple of different pieces. First, the big one for a lot of the traditional studios is that they’re still trying to get their profit margins under their control, and the debt financing is not going to be an option going forward.
So if a show underperforms badly, and you can essentially write off that value and take the tax losses, you can protect some of that revenue you would have to send to the government. Getting those write-offs off the book, so you can take the content and monetize it elsewhere, is one of the biggest ones. Related to that, in the initial exuberance of the streaming wars, especially from the new entrants to catch up to Netflix, a lot of the deals probably weren’t good deals in the first place.
If I’m right that the future TV ecosystem is smaller than the current TV ecosystem, then that means, by definition, that the budgets have to come down, or the amount of shows has to come down. In some cases, deals probably didn’t make a lot of sense or who wrote them didn’t structure them well enough to give streamers flexibility for how to sell them in secondary markets. And then, the last piece, even though it is very small, is that the streamers do pay residuals on a lot of content. There has been new stories that say that residuals are dead. This is not true for streaming. But the way it works for streaming is that residuals are one flat rate.
And when you think about this, you can see why it would motivate some of these decisions. So Wednesday, on Netflix, paid the same amount to its writers and actors in residuals, that the Rise of the Pink Ladies, that Paramount+ just removed, did. One of those shows was the biggest season one on television in 2022, and the other show never made any of the charts I tracked. When you consider that, there’s this benefit that, if you have to pay these recurring costs to keep content on the air, from a cashflow perspective, it might be cheaper to pull them off, ignoring the accounting pieces.
MB: Two follow-ups to the point about the overall market being smaller. I’ve been working on a story about content spend and churn. You’re still looking at a lot of Morgan Stanley estimates and similar things, that have scaled back from where they were, but they’re still projecting growth into the future. Do you see content spend being the same or smaller? And if yes, is it a dramatic reduction in the number of shows, or the spend per show?
ESG: I do not have a strong take. I don’t think it’s going to keep going up. Whether or not content spend as a total goes down, I don’t have a strong feeling on that. We have seen across the board that it really does seem like a lot of people are trying to correct the size of their content spend. So I am more inclined towards believing the spend will stay flat, which, counting inflation, is the same thing as making it smaller over time. I think both will be an option, though. To answer the second part of the question, what happens to the number of content?
We’ve already seen budgets come down in certain cases. For example, on the broadcast season, they’re having fewer actors per episode and smaller guarantees doing smaller writer rooms, things like that. They will try to figure out ways to keep some of the content costs down. And they will try to squeeze in some reality TV shows for their smaller budgets. I think we’ll see a combination of both those things.
The other piece, not really related to removals, is that I think we might see more examples where streamers are selling the same shows to other people because that’s a win-win. Another streamer gets a new show, and then the other streamer gets cash flow. So it’s more circular like that. Providing different streamers with new content, but not having to go out and pay full freight of a brand-new production.
MB: With that last business model question, are you a bull or a bear on ad-supported streaming?
ESG: I am a bull on it, especially with the free ad-supported streaming TV. I call myself the Entertainment Strategy Guy, so whenever is possible, I come back to the core strategy principles. For me, my favorite chart is just a value creation model, asking how does this create value for the customers? And with the free ad-supported streaming in particular, I think it solves a customer need that has not been figured out by the on-demand part of streaming, which is the idea of just wanting something on in the background. It can be reruns of Friends, or it can be or a movie midway through. There are some residual implications for that, but I’ve always thought that, for example, Disney+ misses a huge use case for why TNT, TBS, FX, and USA Network, kept movies on all the time.
It’s not because you plan to start watching Avengers from the beginning, but you see, “Hey, what scene is Avengers on? Oh, I like this scene. I’m going to turn it on in the background, and I can go finish dishes or something like that.” I’m a big believer in that. Ad-supported streaming, in some cases, it’s also just going to lower costs. So it allows more customers to have different options for what to watch at the same price. The one caveat I have for the streamers doing it is that the data I’ve seen is that, the moment you put advertising on something, the customer value for that item drops dramatically. When ads support something, their willingness to pay goes way down compared to when it doesn’t have ads at all. I recently heard a family member complaining to me about Netflix, which doesn’t even have ad-support.
He said to me, “They’re running trailers on things. I can’t believe it.” And that’s the smallest thing you can do. I’ll admit I felt that way with HBO Max, where they will run a trailer, and I’m like, “What are you doing? Just go to the show.” Old HBO used to do it all the time. But I would still say I’m bullish with the idea. I would be the most skeptical about the monetization forecast. I’m sure there are people who would like to take the line and draw it up to the moon, and I always tend to be skeptical of that. I’m skeptical of every business model by nature.
MB: Our thesis is that we look at the revenue per hour, with that being most, or all, of their revenue from advertising. We’ve got to create more value in the ad unit, unless they’re just going to recreate the linear ad load. In order to do that, they got to add your measurement and make it more impactful for the buyer, but also bring in more buyers.
There’s only 250 brands in the majority of national TV advertising, there are around a million video advertisers in the US, and the demand side of the equation is really growing. That’s a white space that we see in the market today. But, on a side note there, do you think Fox did a smart move by only having a fast, and not having the other pieces, or are they going to fall behind because of that?
ESG: I think it’s a smart move for now because they’re really focused on the cashflow question. One of my rules is not to bet against Rupert Murdoch because he actually has done the thing that everyone says, “Buy low and sell high.” Someone told me once that Rupert has a price for everything, and he knows what he would sell everything at, and what he would buy everything at. I think he judged that streaming was going up, sold at the right time, then he could come in and buy low while he’s building up this cash pile.
So I think it’s smart. That said, I think you identified the big key piece. Going back to the traditional TV, while those 250 advertisers did buy the bulk of the advertising, almost all local channels always sold local ads to car dealerships, appliance shops, things like that. So there are definitely a lot of opportunities for that. I’m actually working on an article right now on AI and how Generative AI could impact everything, because that is a huge subject.
One of the pieces that I think would be interesting for the local people is about the question, can you make an ad that looks almost as good as nationally produced ads but for local restaurants? I think that could help solve one of those demand issues. That could also help the supply, and that if the ad takes a company one day to make, it’s way easier than having to pay someone to make a whole film. I think they have to solve the demand part, but I am still optimistic for it the way you are.
MB: They got to look at the ad unit now, not as a spot running once for everybody, but it’s 35,000 ad impressions. And again, Facebook has just crushed this problem of figuring out how to get that ad impression in front of the person that finds it the most valuable. That could be location based, or it could be a direct-to-consumer brand that doesn’t care where their customers come from. It could still be the national groups.
When I talk to people that come from the national ad side, they don’t think that way at all. Because they’ve been so successful with the scarcity of selling that one spot. But now having impressions, you could sell this to anybody. It doesn’t matter. You could take 10% of your inventory, focus on that, and let 90% go to the 250 groups. And I’m bullish on people, like Roku, or a fast service that aren’t at the upfronts dominating that market today.
ESG: Right. I think that makes sense. I think fake views are the one other obstacle that the connected TV market, and anything digital, has to handle. That’s why when it comes to streaming ratings, I tend to prefer any platform that has panels. Panels are a little more reliable than just tracking straight usage, because you can already find stories about server farms out there just trying to generate click-through on ads.
That is a big hurdle, but I think you have absolutely identified one of the opportunities. And again, that’s what makes me bullish on ad-supported TV, because a lot of traditional is going to transition over. And to your point, you can do better targeting, and better optimization for some of those ads for smaller people who would never have considered it before.
MB: All right, last question. What is one insight that you have that we’re not talking about that we should?
ESG: I write a lot about strategy, and one of the biggest pieces that continues to amaze me is that, when we tend to think of strategy, most people tend to look externally on it. Where is our product positioned? What do the customers want? What are our competitors doing? And much fewer people look internally at things. And I think one of the big revolutions is going to be in how we work.
I’m a pretty big believer in the books by Cal Newport, including “Deep Work,” and “Digital Minimalism,” among others like “A world without email”, which would be the most applicable to big organizations. I think we’re going to see a big trend in the next 10 to 15 years, with companies moving away from managing most things in their inboxes to project-based workflows. They will be doing things slower, but at higher quality. They will manage fewer projects, and things like that. I always recommend Deep Work whenever I’m asked about a recommendation.
I think the way we’re working is very much like when factories were first electrified. This is an analogy Cal Newport makes. That change required rethinking the conception of an entire factory floor, but that process took decades.
That will be one of the big changes in the next 10 years, and I don’t think enough people are talking about it.
MB: All right, ESG, I loved this conversation. Where can our audience find you?
ESG: I’m at entertainment.substack.com. That’s my newsletter, where you can get a weekly streaming ratings report, which uses multiple data sources to provide a comprehensive look of what’s working and what’s not working in streaming wars. I also publish regular strategy columns over there as well. I publish every other week for theankler.com, a strategy column as well. And I’m on Twitter and LinkedIn as well. So if you want to connect on LinkedIn, go ahead.
MB: I’m grateful for your time, this has been a blast.
ESG: I enjoyed talking to you and appreciate the invite on the podcast.
See the rest of the Screen Wars Thought Leader Interviews here!
A former strategy, research and business development executive at a major streaming company, The Entertainment Strategy Guy writes about business strategy, streaming ratings and entertainment for The Ankler and his own website. A veteran with his MBA, he prefers writing more than sending email, so in 2018 he started his own website and has since expanded into newsletters. The EntStrategyGuy has worked at entertainment companies large and small, and leverages this experience along with his love of numbers to explain the business of entertainment.
Cross Screen Media is a leading CTV activation managed service for marketers and agencies, built on a proprietary technology platform that enables advertisers to plan and measure advertising across Connected TV and audience-driven Linear TV at the local level. We seamlessly fit into existing workflows to help agencies scale, differentiate and deliver high-impact campaigns for their clients. For more information, visit CrossScreenMedia.com.