T o r t u g a

What if I am an algorithm? Understanding the language of the algorithms and integrating it in a movie strategy.

The keys of the online reality idea is the result of my need for a language or for a map that non-technical audiences like myself can understand when it comes to navigating the overcrowded online ecosystem for the purpose of visibility, promoting ourselves and our work, in ordinary words. So, I broke it all down to keys or anchors.

I don’t know how many of us are aware, but learning more than we were willing to, outside our field, is no longer optional in the algorithmic culture. Not if we discuss film production as a sustainable career.

While diving at the intersection of marketing and IT realms, I extracted 8 golden keys from the blueprint of the online world:

1. Algorithm
2. Marketing
3. Key words
4. Titles/Headlines
5. Content
6. Audience
7. Advertising
8. Monetization

In this post, I map the essentials for The Algorithm, the number one key of the online world.

The Algorithm: the no.1 key of the online reality. What language does it speak?

From the perspective of algorithms, content is not only what we create, but content is also us, namely our data, that is, our target audience. All our movements in online reality define our behavior and help algorithms to better segment us into audience groups and subgroups and provide us with the most sophisticated and satisfying personalized experiences (the object of UX – User Experience departments, among other things this department does).

“It’s hard to remember where something is — was that movie on Netflix or Hulu, or was it Disney+?”

Bryan Striegler, a photographer at Striegler Photography, tells Lifewire – Tech for Humans.

According to a recent survey titled “2021 CTV Growth Opportunity Report” conducted on a sample of 3,000 users of streaming and digital TV platforms in the US by Verizon Media and Publicis Media, 56% of people say they are overwhelmed by the number of services streaming to choose from. The survey also shows that 67% of users say it is difficult to decide what to watch because there is too much content.

“With cable, you have one place to look, you have your favorite channels, and there might be 100 things to watch versus 50,000. Streaming services were created as an alternative to cable to offer unique options at a lower price. But unlike cable, they’re based on algorithms, which means we don’t get the experience of stumbling upon a new show or channel, whereas with cable we scroll through and find something new we wouldn’t have watched otherwise. The problem with algorithms is that they don’t challenge viewers to try something new. They’re only targeting what we’re watching now, giving us similar content, and that can be a big problem.”

filmmaker Daniel Hess of Tony Productions in a statement to Lifewire – Tech For Humans.

This reality was valid until yesterday, in 2021. Since then, the TikTok algorithm appeared and capitalized on this problem, becoming hyper-addictive for all generations that populate it precisely because it offers the public new entertainment content in a continuous stream. According to Statista, TikTok users watch 167 million hours of video in one internet minute. At the same time, TikTok’s direct competitors, Facebook, YouTube and Instagram, launched or relaunched Youtube Shorts, Instagram Reels and Stories, features that are considered specialties that gives creators not only funds to create what they like on their platforms, but also a generous opportunity to bypass the algorithms.

The dialect of algorithms. Or the language of the online reality.

In the language of algorithms, the logic of the world is simple: good values = good content, so it must be pushed as high as possible. Bad values = bad content, so it needs to be buried.

From the book “Hook Point: how to stand out in a 3 second world?” by Brandon Kane, an online marketing voice in the growing crowd of online marketing names, we take away the idea that no matter where we publish our content, on the website, on social networks, on streaming platforms or on any other apps and content channels, the biggest determinant of success is how the algorithms treat us.

When it comes to digital content platforms, all algorithms, no matter how different their subtleties, have a common goal: to give audiences the best possible digital experience while on a website or network. Because if we enjoy the time we spend on a platform as much as possible, it means that we use the algorithms more. The more we use the algorithms, the more they learn about our behavior and preferences, the more the platforms can offer more ads and attract more followers, implicitly more sales and opportunities, thus securing not only their success but also their business continuity in the market.

It all comes down to data, because data means better segmentation and targeting of audiences, i.e. better personalization of the digital experience, essentially better personalization of life.

Paradoxically, the online universe is not so big for video content creators and filmmakers as it is an untapped or underutilized terittory. After all, a handful of creators are constantly churning out content for billions of internet users to see something new every hour of the day. However, from this very small community of creators, an even smaller fraction manage to cross the online universe and reach the public.

Streaming platforms currently have catalogs of thousands of movies, they function as stand-alone search engines. They do not market the films in the catalog, but their own services and products, the biggest marketing efforts remaining with us, the video content creators.

In cinema, it is said that if you understand the “Tarkovski algorithm”, you know how to make a film. In the online world, if you understand the Google algorithm, you know how to make yourself visible anywhere.

Just like in cinema where it is said that if you understand the “Tarkovski algorithm”, you know how to make a film, and in the online world if you understand the Google algorithm, you know how to make yourself visible anywhere.

The problem is that algorithms are just pieces of code. They can’t watch a video and know how entertaining it is, they don’t know how engaging the film we put on a platform is, they don’t know how valuable our work is. Algorithms figure out what is “good” content (versus “bad” content) based on various values/indicators, the most popular being: the number of clicks, user retention on a website or platform, audience engagement (likes, shares, comments, reviews) and the viewing time of a video or movie.

For an overview and to better integrate the “mindset” and language of algorithms into our own mindset and language, here are some essential algorithms that make the internet work and that are behind our online actions or rather our online behavior:

• Google search: it is a giant algorithm that is constantly changing to give us the best possible results when we search the web;

• Routing: it is the operation of transporting information in a network, one of the basic technologies of the Internet, without routing algorithms, we do not have the Internet;

• Encryption: it is the operation of transforming information into characters that are impossible to read and decode, being a security operation on the Internet. Without encryption algorithms we would not be able to securely make online payments or communicate;

• Facebook News Feed (Feed): another giant algorithm without which our news feed would explode with useless information;

• Sendmail: it is an algorithm that ensures that our e-mail reaches where we want;

• Hashing/Hash: algorithms that ensure that the transmitted messages have not been modified by a third party;

• Online advertising algorithms: determine which ads are displayed on a page and where, to ensure they are relevant and generate revenue. One giant advertising algorithm is Google Ads, it relies on our search terms to show us ads related to our search;

• Google News (Google News): Google’s news algorithm analyzes the volume of material that appears on a certain topic and considers the number of other pages that link to it to determine whether it should be displayed;

 • Amazon recommendations: being a platform with millions of products from all industries, the algorithm shows us similar products that we might be interested in buying. Users often use it to discover new authors, bands, movies, and artists;

• Netflix Recommendations (NRE) Netflix believes that it could lose at least $1 billion each year due to subscribers abandoning the service if it did not use the Netflix recommendation algorithm, one of the most intelligent algorithms, the system filters more than 3,000 titles, using 1,300 recommendation groups, all based on an individual user’s preferences. Netflix estimates that only 20% of subscriber choices come from internet search, 80% come from algorithm recommendations.

Netflix’s algorithm helps us find a movie to enjoy with minimal effort by estimating the likelihood that we will watch a certain title based on several factors, including: viewing history and how we have rated other titles, other users with tastes and preferences similar, genres, actors, year of release, time of day we watch, devices we watch on, weather outside, season of the year, how long we watch, titles we choose when checking out, when we watch a scene multiple times, the number of searches and what we were looking for, when we stopped watching if we paused, if we went back or if we watched a movie again.

The last data I could find shows that Netflix estimates it has about 90 seconds to grab a consumer’s attention. This algorithm also helps Netflix produce new, original content, content that has a 93% success rate, 58% more effective than a typical TV show, which has an average success rate of 35%.

It’s not unusual for Netflix to make over 10 different versions of trailers for original content that the algorithm predicts will be popular, or dozens of versions of posters, title cards, or artwork, thus targeting a wider variety of audience subgroups. This tells us that the films and promotional materials are thought from the start considering not only their own algorithm, but also the algorithms of search engines and social networks, where this content needs to be promoted, respectively the alignment between the Netflix universe and the world.

Netflix also uses its own recommendation algorithm to promote one movie or show over another. For example, “The Queen’s Gambit” (2020) blew up on Netflix with over 62 million views, placing in the Top 10 in 92 countries. If we are among the millions of viewers who loved this series, it would not be by chance that “Peaky Blinders” (2013-2022) or “The Crown” (2016) were also suggested to us. It’s about what they have in common. It’s all about data-driven association. Netflix uses the success of one movie or series to support others. Each associated choice has a 90% guarantee that a user will enjoy the movie or at least click and watch the first episode.

The algorithms for writing film scripts and producing successful films. Different beasts for producers to play with.

Although not a topic directly related to the performance of the online content, the attention that the film industry has begun to give to algorithms is worth reviewing, to understand how important their role has become in the decision-making process, but especially how competitive become the marathon for a click, but also to understand the connections between these algorithms and online algorithms.

The battle for a second of the public’s attention is being waged on all fronts with all weapons. If casting one actor over another makes the difference between a box office flop and a hit, producers want to know. If a film that flops in the US, would set box office records in Europe or Asia, producers want to know. Now, artificial intelligence can tell us.

Tech startups like Los Angeles-based Cinelytic are one of many working to amplify the decision-making intelligence of producers. Their software is so advanced that it is a movie in itself, a sci-fi or fantasy even, because it allows producers to play with a movie in every way like a video game, taking the field of financial projections to a whole new level level. For example, producers can introduce a cast, then swap out one actor for another to see how that affects a film’s projected box office. Moreover, it allows the comparison and economic modeling of several scenarios to estimate which one will have better implications for various territories. I know. As independent filmmakers and producers, we are very far in the past and the realization of this reality is crushing.

The Belgian software “ScriptBook” (2015) says that its algorithms can predict the success of a film just by analyzing its script.

Israeli software “Volt” (2015) promises that it can predict which demographics will watch a movie by tracking (among other things) how the trailer is received online.

The “Pilot” software promises to be able to forecast box-office revenues up to 18 months before a movie’s release with unparalleled accuracy.

In 2018, Century Fox explained how it uses AI to detect objects and scenes in a trailer to predict which micro-segment of an audience would find the film most appealing.

“Today, there are robots, drones, super technologies on film sets, but the business side has not evolved in 20 years. People use Excel and Word, fairly simplistic business tools. The data is very siled and there is almost no analysis.”

Tobias Queisser, Cinelytic co-founder and CEO.

The reason why this topic is not talked about openly is due to confidentiality agreements between major producers and AI companies, the mistaken association that many creators and producers make between the use of these tools with the fact that it would alter the artistic value of a film, and the fact that the perception the general consensus is still that AI is a bad thing.

The fact that Netflix has revolutionized the mechanics of the film industry through the use of data and algorithms is perhaps the only argument why major producers have been open to the idea of using artificial intelligence applications and started to consider thinking about film production also from the perspective of data and algorithms.

There are also good news for the smaller film producers. As the technology becomes more accessible, independent filmmakers are also playing with these experiments, such as the BAFTA-nominated British director Oscar Sharp, who made the sci-fi short Sunspring (2016) written entirely by an AI robot, the result of collaboration with his old friend, renowned AI researcher Ross Goodwin from New York University. Sharp described his script as an “average” of sci-fi scripts.

On the other hand, the intersection of art created by humans and art created by algorithms and how the two can intertwine has become an increasingly fascinating territory that can no longer be so easily ignored. At their intersection, extraordinary universes of images, sounds, and stories are born, worlds that we didn’t even know how to imagine and live.

“Instead of asking whether machines can be creative and produce art, the question should be: Can we appreciate art that we know was made by a machine?”

Gerfried Stocker, artistic director of Ars Electronica in Linz, Austria.

Sources:

1. “Why Too Many Streaming Services Will Make Us Go Back to Cable,” Life Wire, July 2021.
2. “Are there too many streaming services? 56% of connected device users are overwhelmed,” Fast Company, May 2021.
3. “TikTok Statistics – 63 TikTok Stats You Need to Know [2022 Update],” Influencer Marketing Hub, August 2022. .
4. Brandon Kane, “Hook Point: How to stand out in a 3 second world?”, Waterside Press, 2020.
5. “11 Essential Algorithms That Make The Internet Work,” Business Insider, August 2011.
6. “Why Netflix thinks its personalized recommendation engine is worth $1 billion per year,” Business Insider, June 2016.
7. “How Netflix’s Recommendations System Works,” Netflix Help Center, 2022.
8. “The Full Guide on Netflix Recommendation Algorithm: How does it work?,” Invisibly, November 2021.
9. “‘The Queen’s Gambit’ Breaks Netflix Record as Chess Interest Explodes on Google and eBay,” Yahoo-Entertainment, November 2020.
10. “Hollywood is quietly using AI to help decide which movies to make,” The Verge, May 2019.
11. “An Algorithm Wrote This Movie, and It’s Somehow Amazing,” No Film School, June 2016.
12. “Best Screenplay Goes to the Algorithms,” Nautil, December 2019.