All posts by EntertainmentStrategyGuy

Former strategy and business development guy at a major streaming company. But I like writing more than sending email, so I launched this website to share what I know.

No, Seriously, Why Don’t You Use Data to Make Movies?

If you want to know the “holy grail” for data scientists, I’ll tell you:

Predicting box office performance of movie scripts.

Here’s how it goes. An aspiring data scientist—ranging from bright undergraduate in computer science to a Ph.D. candidate in statistics to even tenured professors—looks for a new topic. They’re bored by analyzing mortgage applications and discover that no one is very good at predicting box office for movies. So they say to themselves, “I can do that.”

Sometimes they even create a model and/or publish papers. Then they go to the Hollywood studios and claim they can use an analysis of a script to predict box office success. Often this is touted alongside advanced analytics, machine learning and neural networks, or other similarly jargon.

We shouldn’t shame these data scientists for trying, though. I mean, the executives at streaming services like Netflix and Amazon Studios/Prime/Video claim they can also use complicated algorithms to predict how well they pick TV shows or movies. Both those streaming video platforms are constantly asked about—and they in turn release vague hints about—the data and algorithms they use to pick TV series.

I have also fielded those types of questions since I helped work on strategy at a streaming platform with tons of data, as I mentioned in my second post “Theme 1: It’s about decision-making, not data”. It typically went, “With all the customer viewing data, how did you use that to pick TV shows?” In my initial post, I specifically didn’t answer the question, but went off on a tangent.

But it is worth answering, because it will illuminate a common Entertainment Strategy Guy theme, “Be skeptical”. In this case, “Be skeptical” of the streaming services claiming they have esoteric data knowledge and the entertainment journalists who let them repeat this.

Of course, I don’t blame the executives per se for claiming they have complicated algorithms. I blame the journalists who repeat it without questioning it. These media members don’t probe that audacious statement. A quick push will reveal those statements to be a house of cards, if you will. (Wow, brutal pun.) In reality, Netflix/Amazon/Hulu/other streaming services and traditional studios don’t have enough data to actually use data to help them make decisions.

So let’s push back, just a bit.

Read More

Most Important Story of the Week and Other Reads – 11 May 2018

Welcome to my first “weekly news round up”!

This isn’t a comprehensive list of every story that premiered this week or last week or whenever it came out. (I may have just found it and wanted to share.) You have Twitter and access to Variety/Hollywood Reporter/Deadline so you don’t need me for that. (Or try out MediaREDEF’s feed.) They do it better than I can, so I won’t compete.

My goal, instead, is to provide a bit of commentary on the news. Like calling out the news that might be slightly misleading or over-exaggerated in its importance. Just because something gets the most clicks doesn’t mean it will have the biggest impact in the future.

The Most Important Story of the Week – AT&T and Time Warner deal

I’ve been writing “test updates” for the last few weeks to see how long these would take to put together. I wanted to make sure I wasn’t over-extending myself by committing to one per week. I bring this up because I was tempted to refer to a now unwritten update, and I realized I needed to explain myself.

My goal each week is to call out the most important news story. Not the biggest. Not the flashiest. The one with the longest term impact on the business we call entertainment. The other sections of this update will rotate, but I’m going to try to call this out each week. The story I think will have the biggest impact on the future of entertainment.

The success or failure of the AT&T and Time-Warner merger is clearly the biggest news story of the week, and only dwarfed by Disney and Fox acquisition as the most important story of the last year or so. The ramifications of further media consolidation will force every other player to consider how they react, and will likely spur further mergers.

Also, regulatory context is always important, especially in media and entertainment. Trump’s regulatory environment is, frankly, unprecedented in American history. I don’t like consolidation in most industries because I believe it hurts consumers. This is ostensibly why the DoJ is pursuing the anti-trust lawsuit in the first place as they wrote in their final post-trial brief. If that were the whole reason, I’d be really excited because it would mean the next step is breaking up Comcast-NBCUniversal and forcing other content makers to divest their distribution arms. (I’ll have a post on that in the future.)

But Trump’s Department of Justice isn’t pursuing action against AT&T for that reason; it’s likely pursuing it because Trump hates CNN, which is owned by Time-Warner. They said mean things about him in the election and he calls them fake news. That’s why Trump is okay with the Disney-Fox acquisition, because he likes Ruper Murdoch and Bob Iger hasn’t gotten on his bad side. Again, unprecedented.

(Notice, the Michael Cohen and AT&T payment/consulting story isn’t important in an entertainment and media sense, but it is salacious.)

Fun idea

I’ve said before I’m a sucker for frameworks. Well I looked, and I haven’t said that before. Well I’m a sucker for frameworks. (A professor once said this about me; he was so right.)

So when I see an HBR Ideacast with a framework for launching a start-up, man, I’m probably going to recommend that to my audience. And here I am doing just that. Take a listen to “Choosing a Strategy for Your Startup” by Sarah Green Carmichael interviewing Joshua Gans.

I’ve always thought that entrepreneurship is overrated from the standpoint that most companies should be constantly thinking about entering new lines of business. All companies should launch news businesses or business units. Thus, the lessons from this podcast apply just as much to people in big companies as small ones.

Long Read of the Week

One final caveat, I recommend long reads, even if I disagree with the ultimate point. My favorite type of articles are ones that inspire ideas. Sometimes that’s really good articles that I can apply elsewhere (see the podcast above); sometimes that’s articles that make me so upset, I have to write a rebuttal.

The article I’m recommending this week falls somewhere in the middle, “The Revenue Stream Revolution in Entertainment and Media” by PwC’s strategy+business. Let’s start with the negative. In the title. “Revolution”. France in 1789 was a revolution. Russia in 1917 was a revolution. (America is 1777? Depends who you ask.) New revenue streams are a change, they aren’t a revolution.

There are three ways to make money in entertainment: sell ads, sell subscriptions, or sell product. I call these three–and I didn’t come up with this, I learned this in class–advertising, subscriptions or transactions.

And a quick look at the “revolution” is really just different forms of those three things. “Platform” is basically owning the distribution channel…much like every studio already owns TV networks. (Subscriptions.) So yes, the platform is changing, but it’s not a new business model. Or take “the Omni-brand” which recommends moving into different revenue streams for successful IP. I’d say, “What are you, Walt Disney in 1930?” (Transactions, advertising and subscriptions.)

My other criticism is the use of qualitative descriptions that could be quantified but aren’t. Take these sentences from the opening:

“The competition for user engagement and spending has never been more brutal. All these developments have significantly disrupted the flow of E&M revenues. Gone are the days when TV networks, film studios, or companies of any kind could thrive on one, two, or even three reliable revenue sources. Today, profitable growth increasingly depends on having five, six, or even more revenue streams.”

I think I could take each one of those sentences and interrogate it with data. Now I don’t have it all, but I’m sure PwC does. So are entertainment revenues being disrupted? Does that mean they are down year over year? If so, by how much? Competition for spending has never been more brutal? How do I quantify that? Compared to what time frame?

It’s not that the sentences are wrong, just that I don’t see any proof. Otherwise, it could be conventional wisdom that may not be true. If it isn’t true, and you’re making decisions off of it, you might be misled.

Final point, this is a good article and I like trying to define new business models. The idea I had is to take their framework and try to pin additional companies into their buckets. Not just using successful businesses, but all businesses.

How much money will Disney make (or lose) on the Lucasfilm deal? Part I

(This is Part I of a multi-part series answering the question: “How Much Money Did Disney Make on the Lucasfilm deal?” Previous sections are here:

Part I: Introduction & “The Time Value of Money Explained”
Appendix: Feature Film Finances Explained!
Part II: Star Wars Movie Revenue So Far
Part III: The Economics of Blockbusters
Part IV: Movie Revenue – Modeling the Scenarios
Part V: The Analysis! Implications, Takeaways and Cautions about Projected Revenue
Part VI: The Television!
Part VII: Licensing (Merchandise, Like Toys, Books, Comics, Video Games and Stuff)
Part VIII: The Theme Parks Make The Rest of the Money)

Let me take you into the mind of a business school student. While in school, you’re taught to be super critical of any business presented to you in the form of a Harvard case study. Even if things look super rosy, there are some bad numbers hidden in the appendix you need to find.

At the same time, you’re taught to be super positive for any business that is doing well on the stock market at the moment. You don’t have appendices to go searching through to find flaws. It’s a weird dichotomy of simultaneous criticism and optimism.

The company that was flying high when I was at business school—and has been a darling of Harvard case studies since the 1990s—was The Walt Disney Company. During my first year as an MBA student, Disney acquired Lucasfilm, the maker of Star Wars and Indiana Jones (if somehow you didn’t know that). I was so enthused by the deal that I used it as my topic for our speech class. To call me “enthusiastic” would undersell my opinion: I thought it was a guaranteed home run.

So when I came up with my first “analysis” article for this website: “How Much Money Has Disney Made on the Lucasfilm deal?” I remembered I’d already tried to answer that question. In that presentation above.

So I pulled out the presentation. I searched for my numbers to see what I said. I only found one slide with any numbers on it. I laid out the challenge for The Walt Disney Company: to make a good return on its investment, Disney would need to earn nearly $550 million per year to make up its money.

Slide1

That’s a huge number. So did my speech predict how much money Disney would eventually make on the deal?

Nope.

If I could make one change to the entertainment business press—and I’d make a few—it would probably be to enforce this rule: You don’t have a strategy if you don’t have any numbers. Looking back on that presentation in speech class, I didn’t obey my own rule! My presentation didn’t have any numbers. Well that’s not quite true, I had some tables showing that Disney does well at the box office and international growth is important, but I didn’t project how much money I thought Disney would make or lose by that deal. I just said I loved it and listed some general strategic points.

That’s what most of us do day in and day out in business, and I want to change that.

Strategy is numbers, and today I want to look back at that deal. It feels like a good time to update our thoughts on the Lucasfilm acquisition. While the last film was a box office smash (the number one movie in 2017), it had the worst customer feedback since The Phantom Menace. Worse, Solo had a troubled production, worrying fans on the Twitter/Reddit. And when Disney announced a new trilogy with Benioff and Weiss, it was met with a giant “Eh” and articles worrying about “saturation”.

Today, as a business analyst who loves Disney’s model and a Star Wars fan who loves the franchise, I want to combine these two things and answer a question I haven’t seen anywhere else: How much money will Disney make on the Lucasfilm deal?

Blink and Gut Analysis
When I write an “analysis” article, I’m going to try out an approach different from most other analysts. I laid out my rationale here, but to summarize, too often when we think about complex questions (like the one I laid out above) we don’t clearly own up to our initial reactions and gut thinking, even though that inevitably informs our final analysis. To combat that, I’m putting my “blink” and “gut” reactions right up front, then seeing how they change as I run the numbers. Read More

Introducing: “Analysis Articles”

Tomorrow, I’m going roll out part one of a potentially four (or more) part series analyzing The Walt Disney Company’s acquisition of Lucasfilm. I’m going to ask a simple, yet tough to answer question, how much money did Disney make off that deal?

When I launched this website, I decided specifically not to call it a “blog”. And I don’t want to call it that because of that article coming out tomorrow. And I want to explain why.

Since I left my last job, I’ve been thinking about writing about the business of entertainment. Specifically, the strategy of entertainment companies, defined as the decisions they make and the quality of those decisions. I selected this vague area because in my analysis of the entertainment journalism landscape, I didn’t really see anyone covering it.

Trade goings-on? Covered by Variety and Hollywood Reporter and others.

Industry analysis for financial firms? Covered by stock brokerage companies, including Rich Greenfield of BTIG.

Statistics of entertainment? A lot of statisticians dive into that field with websites.

Strategy of technology? Obviously StraTECHery has us covered.

But analysis and explanation of the strategy of the entertainment business? I just don’t see enough of that on a regular basis.

I had a ton of ideas for smaller articles. I’m calling those my idea posts and they’re most similar to what an old-fashioned blog was: a bunch of posts going up on a variety of topics with my opinions. Some of these will be on the business of entertainment, or industry trends, or recommendations or explaining updates to the site or even vaguely political posts (if the politics impacts the entertainment industry, and it does). But those ideas aren’t enough to distinguish me in an already crowded entertainment journalism landscape.

I also know people love updates on the news. I could try to distinguish myself there. But honestly, I’m not good at devouring news constantly and spitting it out. That said, I do read a ton, and I get thoughts. So on a weekly basis I’ll put out a review of the news of the week, with my own spin on the most important stories.

Still, that’s not enough. Which is where “analysis” articles come in. These are the magazine pieces of this website and hopefully the centerpieces for why entertainment execs—senior, novice or aspiring, creative and business types, all are welcome!—will come here. I’ll take big questions and try to answer them analytically, in a readable style. My first one goes up tomorrow, and it’s a fun question with a lot of serious learning points and deep analysis, that ends up touching on the future of the film industry.

Types of Analysis Articles

To give you an idea of how deep, I’ll give you an idea of some of the topics I want to pick.

Big Questions: I’m starting with this tomorrow, which is a good example of a big one, “How much money did Disney make on the Lucasfilm acquisition?” It requires running the potential financials on four different lines of business, and will allow me to explain the economics of feature films. Other ideas in this category are “How does Netflix value original content?” and “How much should Disney charge for its SVOD platform?” In each case, I hope to explain how I got my numbers so other people can learn from it.

Versus: As I was brainstorming potential topics, I started thinking of companies to compare to each other. Sort of like a preview of a football or basketball game, but for entertainment companies. And instead of a one off match, it would predict out the next two to three years. Once I started, I came up with a ton of fun ideas like, “Who won the deal, Disney or 21st Century Fox?”, “Who has a brighter future, Atom Tix, Fandango or Movie Pass?”, “Who should Vox sell a TV show to, MSNBC or Netflix?” or “Who did better Grantland or The Ringer?” Really, the sky is the limit and I want to compare everyone from Snapchat to Twitter to NBC-Universal to Warner Brothers.

Power Rankings: So since I”m already declaring winners above, my next idea is a natural extension of that. I’m going to rank companies in different areas as I come up with topics. The key will be grounding them in specific numbers, and that’s the hardest part so far, but could be the most insightful. I’ve come up with a few fun ideas including “Streaming Services”, “Franchises” and “Kids Platforms” and we’ll see how many more I can do and how long they take.

What “Analysis” articles will include

What can you expect from analysis articles (assuming they don’t kill me in the meantime)?

First, numbers. One of my future themes will be “strategy is numbers” and one of the most frustrating things for me when reading coverage of entertainment is the willingness to ignore numbers when proclaiming a winner in strategy (especially with Netflix). That’s why I’m going to try to push the questions to always having a financial winner or loser, if I can. Hopefully, this is grounded in really good data, though that depends on what I can find.

Second, learning points. As I was writing my first few analysis pieces, I felt the need to explain parts of entertainment business models. These are things I just don’t think a lot of people know and even I learn more about as I do them. So I’m going to explain what I know often to explain how I’m coming to my conclusions. I think it will make us all smarter.

Third, they’ll be long. In a previous writing enterprise, I tried to keep blog posts to under a 1,000 words which meant I often did series of articles on topics. I’ll still break up analysis articles, but often times the whole article will be several pages long. Again, much more like a magazine article than a blog post. In the end, I may collate the entire thing into a pdf for readability.

The Future

So how often will these articles come out? Probably weekly, with occasional breaks. I have the first two mostly written, but they are tremendously long so will come out in parts. Hopefully over time I can shrink that length.

To be fully transparent, in the future, I may hide these analysis articles behind a pay wall. Or only allow access to people who give me their email address (and some data). Why? Well I need to make money on this writing and if my gut is correct—I think this articles could be popular and/or the most valuable piece of writing I have—then it would be a good way to monetize the site.

Overall, I’m really excited to get started with these analysis articles. They’re the most exciting things I’ve worked on in a few years related to the business of entertainment. I get to improve my skills at analyzing other entertainment companies, and I get to do it outside of a corporate context. (And a consultant context.) Which means I don’t have to deal with the opinions of a lot of senior people biasing my outcomes.

Also, I get to learn a ton more about how different businesses operate. Stuck inside a company, you get very myopic. Plus, there is something about reading news reports on a company, and something else trying to dig into a company in detail. The “versus” concept in particular has allowed me to compare two companies with something interesting linking them.

Finally, I get to hone my process. I think we all need to make better decisions, and analysis articles are my way to do that.

I hope you enjoy the first of many to come.

My Process

Most organizations–from corporations to sports teams to armies to universities to government agencies to non-profits–think they use a sound process to make important decisions.

But they aren’t. They are arriving at the easiest decisions.

I’ll use a media example to show the worst process. Basically, political pundits opine and opine and opine for months about an upcoming election. Then when it happens, they go on air and, if they were wrong, ignore what they said before, or, if they were right, embrace it wholeheartedly. Either way, they develop a narrative to suit whatever the data says. That’s an easy process.

The opposite process is much harder. In my opinion, the gold standard of openness is the journalism of FiveThirtyEight. Nate Silver and his team go out of their way to be open with their process, even explaining how their election model works (or doesn’t). They also write articles owning up to mistakes when they make them. The highlight was when Silver admitted that he didn’t acknowledge the data that showed Trump’s underlying strength during the primaries. He corrected and didn’t make that mistake in the 2016 general election.

I wish corporate America had such an open process.

I want to emulate Nate Silver’s approach to tough problems. He has politics and sports covered (along with some other great journalists), so I’m going to focus on entertainment. I want to ask big, interesting questions, and use a sound decision-making approach to find unique insights and answers.

Honestly, I want to get better at analyzing problems and hopefully making better decisions overall. I want to be more accurate. To do so, I am going to use the decision-making process I wish I could have used in my last job. Yeah, I stole Nate Silver’s philosophy, but I’m tailoring it to my style.

That’s the process I want to explain here, once, so I can reference it in all future “analysis” articles. (I’ll explain what those are in a separate post.) I want to be very open with my process. This way, when people criticize my conclusions, I can point to my process and ask for recommendations for how to improve that. Hopefully, it will also make other people better at understanding their processes.

Part 1: Blink Analysis

I start the process with my “blink” analysis. Yes, I stole the name from Malcolm Gladwell’s book. Like any long Gladwell topic, you never quite know how well the initial ideas have stood up to longer-term academic scrutiny.

But the core idea is fantastic. As soon as you hear something, you get an instant reaction. And I try to capture that as soon as I get an idea for a question. I try not to think about it, but just put my thoughts on paper. Consider this my “looking at a statue and calling it fake” analysis for those who have read Gladwell’s book.

My working theory is that most of us do this ten times a day in conversation. You hear some news, “Did you hear that so-and-so was hired at such-and-such company?” and draw a conclusion, instantly. We just don’t acknowledge it. I’m trying to acknowledge it up front.

Part 2: The Gut

My next step is the “gut”. Here, I try to write down what I think about a question, but my goal is to reflect a bit deeper on it. Whereas the blink is a paragraph, this could run to a page or more.

The best analogy is the “case question” in an interview. In my analysis articles, I’ve asked a challenging business question (to myself) and now I need to see if I can answer it. And like a case study question, I won’t have any data to go from. I can divine some data I think I know or know for sure, but I can’t use the internet to confirm it. So my “gut” analysis is deeper than blink, but without the numbers that come from part 3 of the process, my analysis.

Honestly, when I set up this process, I realized how often this is essentially what we do in a business meeting. Sure we get initial blink answers, but as we have a conversation about an important issue or the future of entertainment (say at a conference talk or in a group or around a table in a meeting), we’re having a “gut” conversation. We can’t pull up numbers, so we do our best guesses.

The reference book here is Thinking Fast and Slow, or The Undoing Project, which are stand-ins for the work of professors Kahneman and Tversky. Two of the founders of behavioral economics, they basically show that “blink” goes wrong when it relies on misleading heuristics. In other words, your blink can often mislead you. The gut stage is supposed to short circuit that quick analysis to allow a moment of reflection.

That said, the gut may be the most dangerous or seductively appealing of the three stages. My “gut analysis” is NOT my final answer. The gut can be tremendously misled by assumptions we haven’t fully explored because we don’t have enough time to do it. Or the data doesn’t support it. It gives us unearned confidence.

This is why I think Hollywood overall makes most of its bad decisions, it stops after the gut decision-making. So how do we make our gut a little more accurate? We do the analysis.

Part 3: The Analysis

This is really the meat of the process. This is where we make better decisions. And it’s the toughest part.

So what is it? It is about deciding on a framework, and putting numbers to that framework to derive an answer. It won’t be the same process each time—that’s asking too much from any framework—but every process should have some numbers. Those numbers are then analyzed. Throw in more quantitative and qualitative data along with history and experience. But it is ultimately boiling down to some numbers, whose judgement I have to accept.

A key to making this work is to do the analysis myself. In most cases that will mean asking the right questions, finding the data, building the models and thinking about all that I’ve analyzed. The key, though, is doing the work myself. I specifically want to avoid the thoughts of others as long as possible in this part.

Why? Because I want to be part of the experts in the crowd, not the crowd-source.

Let’s explain that. Take any one expert. Usually, they are pretty inaccurate. Sometimes they have hot streaks, but overall they can be wrong often. A real world example is NFL experts picking games. Any one expert gets a lot wrong. But if you poll a dozen experts, they’re more accurate than the majority of experts. In entertainment, the gold standard is Gold Derby, whose poll of Oscar predictors is more accurate than most other predictors.

Another example is Nate Silver taking a poll of polls. He doesn’t run his own poll—a new set of data that can be inaccurate—but he polls the polls to make them more accurate. In this analogy, I actually want my analysis to be like the NFL expert or individual poll or the individual Oscar predictor. I’ll be inaccurate often, but if you combine my analysis with others, you’ll have a more accurate picture of the world.

By doing the analysis myself, I’ll come up with unique insights. This means I’m adding analysis into the world, not simply parroting opinions I’ve read elsewhere. I don’t want to be the person collecting the crowd-source, but adding data and analysis into the world you can’t find anywhere else.

My caution is I’m doing this because I can. I don’t do this exhaustive analytical process for fields I’m weak in. Take fantasy football. I can’t watch every football game in the NFL, so I don’t try. Instead, I rely on experts who do and statistics that have predictive accuracy. I think I can add my own individual expertise into the analysis of the entertainment industry because I excelled at this analysis in business school and in my career, and I consider myself an expert. I wouldn’t be writing all these words otherwise.

Part 4: Research and Other Opinions

Doing analysis myself doesn’t mean I ignore other opinions and analysis. After I’ve done my opinion, I head to the crowd. See what they think. Then, I weigh whether any of it has a strong enough weight to change my mind.

In some cases, like comparing “Atom Tickets to Fandango to Movie Pass”—a future article I’m working on—there won’t be a ton of people weighing in on the issue. Or there won’t be a substantive analysis to compare my own to. In other cases, like evaluating The Walt Disney Company-21st Century Fox acquisition, lots of people analyzed that deal at the time.

There is another analogy to polling that provides the one caution at this stage. One of the things Nate Silver worries about is when he sees polls cluster around the same result. This worries him because he thinks pollsters are subtly manipulating their polls to arrive at a conclusion. The polls do this because they don’t want to seem like outliers with the negative connotation that comes from that.

I think this phenomena plays out all the time in the world of business. Doing your own analysis takes a lot of time, so it is easier to let someone else do it. So a lot of the business world (and stock market) cluster around similar opinions even if their own analysis pointed somewhere else. (Or would point to it if they took the time to do a clear process.)

So after I do my own analysis, I’ll read other people’s thoughts and update my thinking. See where I go against the grain and see where I don’t. This may be in my analysis or in later updates, I haven’t decided yet.

So that’s my ideal process. Now it’s time to see how it works.

Theme 1: It’s Not Data, It’s Decision-making

The big buzzword in business is still “data”. Or better yet, “big data”. Or more complicated sounding, “analytics”. Better than just analytics is “advanced analytics” which is like analytics but more advanced. With all that data, a bunch of “algorithms” are figuring everything out.

Take your pick. Sure other trends have picked up in the last few years—how about disruption anyone?—but since I started business school, data (or one of those words related to it) is everyone in business’s obsession. Companies launch whole businesses now who run off business models entirely related to collecting customer data. Facebook’s Cambridge Analytica revelations just brought this to the fore.

And this “data revolution”—another term—has come to entertainment.

Especially the streaming video services. They get so much data that was never there before, including the entire viewing history of a customer (and their shopping history on Amazon). So when I talk to MBA students—either at alumni events or recruiting events or in a classroom—I invariably get asked this question: “How did [your streaming platform] use data to pick which TV series and movies to make?”

I can’t help myself from sighing. And being slightly sarcastic. So instead of answering, I usually flip the tables.

I ask a simple question (if say I am in a classroom). Something like, “Let me ask you, when you decided to take this course, what data did you rely on to make the decision?” Usually, the answer is some stumbling around, ultimately ending on someone had recommended the course to them. I point out that this is data, a qualitative piece, but still a personal recommendation.

Notice that word I paired the word “data” with “decision”. So let’s use a real-world example. From my life, but anonymized to protect the innocent.

A group of Hollywood executives are sitting around a table. They only have the budget to renew one more TV show to release in the upcoming year. The decision is around two shows. The first I’ll call, “Cop Show”. It’s a show about police officers. Yeah, creative title. The other show is called, “Awards Show”. If you went into a laboratory to make a TV show to win awards, this was it.

Now, we have to be clear, there is no lack of data. The studio execs have reams of it. They know have how many customers watched each show. They know what shows customers rated highly. They know what critics thought. They have surveys of customers and have conducted focus groups. They think one show will probably stay more popular, but will never win awards. The other show will likely win awards, but customers didn’t watch the first season. But they have a ton of data to draw on at that table to make these conclusions.

Of course, in addition to all the data on the table, there is a lot of data off the table. Pieces of information influencing the executives, but that doesn’t make the strategy powerpoint that justifies the decision. For instance, some of the development execs have friends helping make one of the TV shows. Some development execs are considering whether a popular but uncritically acclaimed show will help their chances at getting their next jobs. And the marketing execs want the easiest show to market.

To top it off, some people just like one of the shows a lot. It’s their favorite show on television. That happens.

This is why I asked the students about the “decisions” they had to make. The execs have a ton of data…how do they use that data? Data doesn’t make decisions, the people do.

Big decisions, like what MBA school to attend, have a lot similarities with picking a TV show at a network, which is why I used it as an example. A prospective MBA student has tons of data to pour through. Though a lot of that data is qualitative. And some of it is on the table, and some is subconscious. And a lot of it has vague predictions about the future: which school will I enjoy the most? Which school will help me get the best job? And in what field? That’s making assumption about the future. Every MBA student had to decide where to go to business school; likely, they didn’t have a data-driven process. They picked the highest ranked school from a list of school rankings whose methodology they probably barely understood.

Because, honestly, people don’t really like using data to make decisions.

Let me say that again, because it is the biggest myth in our business world right now: people don’t like using data to make decisions.

Most huge corporations already have a well-worn, time-tested, established system to make decisions called: HIPPO. Highest Paid Person’s Opinion. (I’m not the first to write this.) Basically, you have a CEO making anywhere from $1 million to $50 million dollars. What if you gather a bunch of data that says they’re about to make a huge strategic mistake? Will they change their mind?

Of course not! They’re the boss.

To reverse course would admit that they were wrong. And this goes all the way down the chain. Say you are a middle manager who is convinced to make some key decision. You decide to change how your team processes something or other related to payments. But you also have a Business Intelligence team. A BI team that developed a sound forecasting methodology, with clean, reliable data and that team’s new methodology argues against your strategy. Whose opinion do you take, the data or you?

You pick you! And that’s what development executives not just at the streaming service I worked for, but for all of Hollywood, do on a daily basis. It’s what financial executives do when making investment decisions. It’s what corporate strategists do when determining their strategy. You can develop a data-driven approach to making decisions, and really think about what decisions you make, or you can go with your gut.

This isn’t to say business folks don’t use data. They have tons of data to read. Like those studio executives around the table. They make reports justifying strategic decisions that are loaded with data. And they may even listen to it a bit. Let it inform what they think, at some level. But the key is executives are the ones making the decisions, not methodologies. Not the data. And with tons of data, more data than ever, it is even easier than ever to construct a narrative that you’re making sound decisions.

This is why I return back to the core theme of this blog: it isn’t about data, it is about decision-making. The goal isn’t to just have data; it’s to use that data to make better decisions. Making decisions is about much more than regression analysis or random forests or neural networks or some other complicated algorithm. It is about asking the right questions, to even know what decisions you are making. Then you figure out your process, your methodology, your data and how you will measure success. Then you let the process make the decision, not your gut.

You cannot make wise decisions without data, but instead of focusing on the “data’ we need to focus on the “decisions”. This idea will come up so often, when I write about entertainment strategy, that I felt the need to make it my first “theme” of my writing, a point I will return to again and again.

As an undergraduate, I heard a description of hierarchy that has stayed with me Basically, the professor said, “salary in an organization is determined by the value of the decisions you make. If you’re decision is, ‘Do I empty this trash can?’, you get paid the salary of the janitorial staff. If you’re decision is, ‘Should we enter China or India as our next market?’ you get paid like a CEO. One of those decisions is valued at cents (the trashcan) the other in millions or billions (entering new markets).”

This site is concerned with those decisions as they relate to entertainment. And making them better.

Data isn’t the solution, better processes and decision-making are. And a lot of that better decision-making will say, ironically to this entire post, to use more data in places where it isn’t being used at all (lots of parts of entertainment) to help make better decisions. In the long run, we can even identify who is making the best decisions and reward them, instead of the executives at the top who likely aren’t making the best decisions, or ensuring they even have a good decision-making process.

Hopefully this website, over time, will help you and your company do that better.

Why Does Hollywood Make Bad Movies?

Before I start writing on a new project, I often feel compelled to explain myself. Why am I putting my words to paper (or coded into bits, since you’re reading this on the web)? Why are my words worthy of your time to read? Why do I want to get these thoughts out there?

On a really simple level, I think I can provide insightful analysis on a topic (entertainment strategy) that I haven’t really seen. I’ve seen good journalism on the goings-on of entertainment; I’ve read good writing on the statistics of entertainment; I’ve seen good articles on the future of entertainment in general; but I haven’t seen the one place that ties it all together.

That’s a gap in the market I think I can help fill. Still, what’s driving me to write to fill that gap?

Basically I want to answer the question in the headline, “Why does Hollywood make bad movies?”

This seems to be the du jour criticism of Hollywood. And it has been since I started reading the Los Angeles Times annual preview of upcoming movies as a child. Critics have consistently bemoaned the state of the US film industry and the quality of its movies. They’ve criticized the number of sequels, which in recent years has turned to criticizing blockbuster franchises and/or superhero movies. But it’s not limited to film. Those same critics have bemoaned the state of TV or video games, even as we entered a golden age of television and independent games studios thrive.

Those same critics who bemoan the state of Hollywood have also tried to offer explanations for why Hollywood makes bad movies: It’s studio executives trying to make more money off sequels. It’s studio executives trying to make movies to sell toys. Or happy meals in the 1990s. Or its studio executives who won’t give creative freedom to creative types. Or it’s studio executives who care about international box office.

So it’s the studio executives. Hmm. It’s like we could modify the NRA slogan: “Hollywood doesn’t make bad movies, studio executives make bad movies.”

The critics are onto something, but they just can’t explain why. Or I’ve never heard an explanation I love. After blaming bad movies on studio executives, they can’t answer the key logical fallacy they just exposed: why would these studio executives willingly make bad movies? They don’t want to make bad movies.

Instead, I’ve always felt the explanation for why Hollywood makes bad movies is fairly simple: it’s the decision-making.

But I feel a bit like Han Solo after Obi Wan Kenobi drops the “without Imperial involvement” line in the Mos Eisley Cantina. If you asked, “The decision-making. Hmm, can you explain that to me?” Well, that’s the rub, isn’t it? Which is how Han Solo responds.

Well, it isn’t just decision-making. Studio execs make these decisions because of how they perceive or understand the business of entertainment. So it’s decision-making in the pursuit of the business of entertainment.

Studio executives make decisions to further their interests and, ostensibly, the interests of their company. And these decisions are are influenced by winner-take-all economics with logarithmic returns. The winners tend to be blockbusters, with additional revenue streams, so from a portfolio perspective, executives decide to make more blockbusters. Yet, blockbusters are poorly correlated with awards success, which is highly correlated with critical acclaim. So that’s what many studio execs try to do, and why many critics criticize their movies.

But that paragraph feels woefully incomplete. Each of those sentences could be its own post explaining the decision-making and economics and business and organizational behavior of studio executives. All the things that influence the decisions they make on a daily basis. Basically, each part of the above paragraph could be one paragraph explaining how Hollywood does and doesn’t make good decisions.

That’ll bring us back to the question at the start of this post. All those ideas–and more–align on the same theme, “Bad movies get made because people make the decision to make bad movies.” That’s right, instead of wondering why this happens in confusion/befuddlement—a sort of throwing up our hands and saying, “Man where do these bad movies come from?”—I want to say, “Bad movies are made because studio heads, development execs, marketing execs, creative talent and others make bad decisions.” And they make these decisions because of how they understand the business of entertainment.

Honestly, I don’t see another website out there that is trying to explain how and why Hollywood works. I want to try to do that.

But it won’t just be movies. Or TV shows. I want to dig into the business of entertainment and media (and tech when they intersect) across all forms of entertainment. And that means studying and explaining and analyzing and critiquing the strategy of various entertainment, tech and media companies. By studying their strategy, we can learn more about the business decisions they make around everyone’s favorite product, content.

That’s why this website exists. I want to explain Hollywood and how it works from a business perspective, with side trips into economics and data. I want to explain Hollywood in the lens of the decisions people make, and how that affects what makes it into the marketplace. I think these explanations could help business people and creative types make better decisions. And hopefully help us all understand why Hollywood does what it does. (But really the people in Hollywood.)

If we know why we make bad decisions, well, maybe we can make fewer bad decisions and more good and great decisions.

Welcome and happy to have you.