From DVDs to Streaming: Netflix’s Less Enduring Moat

Netflix's announcement that they are shutting down their DVD-by-mail service marks the end of an era.

The decision, while quite natural at this point, has quite a bit of symbolism in the interplay between technology and scale-based moats. 

Netflix was an innovator in more than one way. 

During its early days, in the DVD-by-Mail era, Netflix’s main innovation was the online subscription-based rental service that allowed customers to rent DVDs without due dates, late fees, or per-rental charges —a significant deviation from the traditional brick-and-mortar video stores. Netflix’s business model included a monthly flat fee that allowed subscribers to rent a certain number of DVDs at a time, but with no time limits. The DVDs were then returned via postage-paid envelopes and the next available DVD from the rental queue was delivered. This model proved to be highly popular but also very disruptive, and eventually led to the decline of traditional video stores and set the stage for Netflix's gradual transition to a streaming service. 

Reed Hastings, the founder of Netflix, foresaw the eventual decline of DVDs and understood that streaming technology was not yet mature enough for widespread adoption. As a result, he launched the DVD-by-mail service with the primary goal of building a strong subscriber base. This approach allowed Netflix to carve out a unique market position, which proved to be difficult for competitors to erode at the time.

However, I will argue that in reality, the advantages the mail service offered during its early days were considerably more significant and enduring than what we now observe with streaming. The DVD-by-mail service provided a solid foundation for Netflix to establish itself as a prominent player in the entertainment industry, creating a strong competitive edge that lasted for quite some time.

I’m not claiming that Netflix shouldn’t have moved to streaming. It had to. Neither am I claiming that it didn’t transition well.

In fact, the company's ability to pivot and embrace the emerging streaming market was crucial for its continued success. 

I’m merely claiming that the current competitive advantage it has is limited and less enduring. 

Let’s delve deeper into this claim. 

Netflix’s DVD Era

From an operations standpoint, several key aspects made Netflix’s DVD model truly innovative, creating complete alignment between the consumer value proposition and the operational model. 

But why did customers choose Netflix over the extensive market of DVD rentals that existed at the time? 

A few reasons: Unlimited variety —Netflix had access to complete movie catalogs from all studios. No late fees —customers hated Blockbuster for their late fees. Convenience —you didn’t have to physically go and pick a movie. They simply arrived in your mailbox. However, there were also some clear disadvantages. The movie you wanted was not always available right away so you had to add it to the queue and hope they would send it in the next round. If you wanted to watch one more movie on movie night… tough luck. And finally, they never had the latest movies. 

But the firm built an operational model to fit that. Let’s see how.

In his paper What is the right supply chain for your product, Marshall Fisher presents the notion that each product’s demand is uncertain. The more uncertain the demand, the more responsive (and thus expensive) the supply chain needs to be to match that uncertainty. 

This is summarized nicely in the following graph: 

​​However, the main idea of the paper is that firms can take an uncertain demand and, through their decisions, they can amplify the uncertainty experienced by their operations even more. Examples of how this can happen are offering customers a very fast response time, a high service level, or a significant variety of products. 

So where would Netflix be positioned on the graph with its significant variety? This is where the model is genuinely innovative. By not allowing customers to choose their next movie (and letting Netflix choose from the queue instead) these fluctuations in demand were absorbed and load balanced. Not allowing people to pick up a DVD from a physical store, but rather shipping them from a central location, reduced the demand that was actually experienced in operations. 

And indeed, Netflix built an impressive supply chain that focused on efficiency (rather than speed) to allow it to handle the massive scale and scope of its movie rental business. Netflix used the notion of virtual pooling: it used the entire network of locations (around 52 at its peak) to decide which movie will be sent to which customer from which location (prioritizing proximity and then availability) without the need to carry excess inventory. 

Netflix’s “no late fees” also had operational consequences. When you mainly incur shipping and handling costs and your customers pay a flat fee, you want to reduce shipping as much as possible. So not only did Netflix want to avoid charging late fees, but they preferred customers to hold on to the DVDs longer than usual. Similar to your local gym: They want your monthly subscription but they prefer you rarely show up!

In fact, Netflix penalized customers that were quick to send the DVDs back —a practice known as throttling:

“In late 2003, a Netflix customer named Manuel Villanueva started a website where he documented problems he had experienced with Netflix, a company that provides DVD rentals by mail. He noted that Netflix had violated its agreement to provide him with ‘unlimited rentals,’ by engaging in a practice known as ‘throttling.’”

Initially, it was a rumor, but… 

“In January 2005 -- four months after Mr. Chavez filed his class action - Netflix conceded, in effect, that they were right. It revised its Terms of Use to state clearly: ‘In determining priority for shipping and inventory allocation, we give priority to those members who receive the fewest DVDs through our service.’ And it specifically warned that heavy renters are more likely to encounter shipping delays and less likely to immediately be sent their top choices.”

This may not portray great customer service, but heavy renters were not the customers Netflix wanted. 

While building the library of DVDs was not that innovative, or difficult to replicate, building the physical infrastructure to sort and ship DVDs at the volume they did was not trivial. 

And then streaming came. 

Netflix’s Streaming Era

Netflix’s journey from a DVD-rental service to a streaming giant was marked by strategic decisions and its adaptability to the changing market conditions. As the popularity of DVDs began to wane and advances in internet technology enabled the possibility of streaming, Netflix recognized the potential to pivot its business model. 

In 2007, the company introduced streaming services alongside its DVD-by-mail service, allowing subscribers to access a growing library of TV shows and movies online. 

Netflix continued to invest in and develop its streaming platform by acquiring and producing exclusive content to attract and retain subscribers. And as streaming became the dominant form of media consumption, Netflix gradually phased out its DVD service (until the recent announcement to completely “kill” the service). 

Although Netflix wasn’t the first ever streaming service, other platforms that offered streaming for movies and video content before them were limited by technology constraints, such as slow internet speed and poor video quality.

Netflix popularized the concept and helped pave the way for others. At the time, their main advantage was the massive consumer base (acquired during their DVD era) and an extensive library (albeit not as large as their DVD library). 

But Netflix quickly realized that simply relying on existing content produced by others was not enough and that they needed to start offering their own content. And in my opinion, the main innovation of their streaming era was launching entire seasons (sometimes even entire shows) at one go, thus introducing binge-watching. 

In a recent statement, Netflix said:

“‘We think our binge-able release model helps drive substantial engagement, especially for newer titles… This enables viewers to lose themselves in stories they love.’”

As the Google Trends chart shows, the ability to watch all of Monster: The Jeffrey Dahmer Story helped drive significant interest to the show.

“‘It’s hard to imagine, for example, how a Korean title like Squid Game would have become a mega hit globally without the momentum that came from people being able to binge it.’... ‘We believe the ability for our members to immerse themselves in a story from start to finish increases their enjoyment but also their likelihood to tell their friends, which then means more people watch, join and stay with Netflix.’”

And the statistics are staggering:

“70 percent of American consumers binge-watch television. For millennials, that percentage is even higher: 90 percent. There are articles that list streaming shows that you can successfully binge-watch over the course of a weekend or even in just one sitting.”

Binge-watching also has interesting operational advantages: it “pools” viewers across time zones and times (different generations). 

Traditional TV shows release episodes once a week and evolve week by week. Succession, for example, requires the buildup from episode to episode to create suspense and ensure that people watch each episode as it is released so they can talk about it the next day at the water cooler. 

Binge-able shows, on the other hand, can be watched at any point in time since the cliff-hangers are only there to entice you to start the next episode right away. Since these shows are designed to be viewed all at once, it’s fair to say that the value derived from watching FleaBag, for example, in 2023 will be identical to watching it in 2033 —very different from most shows that were designed to be released on a weekly basis. 

When a regular channel (like HBO) chooses to invest in a show, they need to ensure there’s sufficient demand during a specific time slot. With binge-able shows the investment can be amortized over an extended period of time (this can also be true when traditional channels show reruns of classics such as Seinfeld or The Office, but it’s rare). 

Netflix, which had a first-mover advantage, got a head start in the market and established itself as the go-to platform for online streaming content, attracting millions of subscribers worldwide. The company invests billions of dollars annually in content acquisition and production, ensuring a diverse and ever-growing catalog that appeals to various preferences.

It’s also better than others in data-driven personalization and makes it easier for subscribers to discover new content tailored to their interests. Furthermore, Netflix has formed strategic partnerships with various content producers, distributors, and technology companies —collaborations that enable the company to secure exclusive content, enhance its streaming technology, and expand its reach across multiple platforms and devices.

But …

Moats in a Bingeable World

Two different things happened at the same time: other streaming services emerged, like Amazon Prime, and producers, like Disney, understood that if they continued to offer their content through Netflix, they would be “modularized” and unable to differentiate, prompting them to launch their own services (e.g., Disney+). 

And indeed Netflix has been losing ground: 

Netflix's market share in the US declined from 49.72% in Q1 2022 to 44.21% in Q1 2023. Hulu increased its share by 3.37% to 21.18% and HBO Max remained stable at 10.61%. PeacockTV grew by a whopping 43.85% to 8.08% from 5.62%, while Disney+ declined slightly to 6.40% from 6.73%. YouTube TV and Paramount+ both experienced significant growth, reaching 5.18% and 4.34%, respectively. The streaming industry showed an overall decline from June 2022 to March 2023. Netflix's lost market share has been captured by competitors like HBO Max, Peacock TV, and Paramount+, which have been growing in recent quarters. 

But why is the competition tightening? Streaming is not very difficult, and the differences in speed are negligible. Recommendations and customer experience? Over time, there is a limit to how good and how differentiated streaming services can be. So in reality, these firms are competing on the quality of content. 

The problem, however, is that once you have enough content, the quality is no longer a scale-based moat. In other words, once a certain amount of content is achieved, bigger doesn’t mean better. Firms still compete every season and every year to ensure good content. 

The trouble is that it’s difficult. 

What makes it even more difficult?

Binge culture!

Why?

According to a recent research paper:

“We analyze the outcomes of two randomized field experiments to study the effect of binge-watching on the subscription of Video-on-Demand (SVoD). In both cases, we offered access to SVoD to a random set of households for several weeks and used another random set of households as a control group. In both cases, we find that the households that binge watch TV shows are less likely to pay for SVoD after these free trials. Our results suggest that binge-watchers deplete the content of interest to them very quickly, which reduces their short-term willingness to pay for SVoD.”

Viewers cancel their subscriptions between seasons. Or worse, hop from one subscription service’s free 30-day trial to another. In fact, the research above shows that after their free trial expires, binge-watchers are significantly less likely to pay for a streaming service than non-binge-watchers are.

“‘From the platforms’ standpoint, creating the opportunity to binge-watch seems good at first because it’s creating engagement,’ explains Ferreira. ‘But, if consumers quickly deplete the content, they’ve eliminated the platform’s major asset. What’s going to keep them around?’”

Although digitizing content cuts down on the need for supply chain efficiency (and the associated costs), a significant investment is still necessary to create content. For example, one episode of Netflix’s The Crown costs $13M to produce. Amazon reportedly spent close to $465M on the first season of The Lord of the Rings show, making it one of the most expensive TV shows in history.

There is, of course, a remedy:

“We also show that recommendation reminders aimed at widening the content preferences of households offset the negative effect of binge-watching, and lessen the concerns of binge-watchers with lack of content refresh.”

But it’s still a competition on quality, and as firms continue to invest in quality, prices will rise and users will have to choose. And indeed, users are canceling their subscriptions. A recent study reveals that over a third of Americans (39%) have canceled a streaming subscription in the past six months.

So what’s Netflix's advantage in making better bets on quality shows?

Looking at the list of nominations in 2022 compared to 2021, the difficulty in being consistent is clear:

In other words: 

During its DVD era, Netflix’s success heavily depended on efficient supply chain management and streamlined operations. But these advantages became obsolete as the market evolved. Today, the primary focus of streaming platforms, including Netflix, is investing in high-quality content to attract and retain subscribers. Achieving excellence in content creation and curation is a considerably more complex and challenging operational skill.

Although Netflix has produced some notable original content, it has not consistently demonstrated a clear advantage over competitors in this area. The shift from its previous operational advantages to the critical importance of content quality has proven difficult for the company. Maintaining a competitive edge based on content quality alone is challenging and potentially more ephemeral, as viewer preferences can change rapidly, and competitors are continuously vying for the same audience.

In the current streaming landscape, Netflix must continuously innovate and improve its content to stay ahead of the competition. This will require not only strategic investments in content creation but also a keen understanding of the ever-evolving consumer preferences. As the streaming market continues to expand and competition intensifies, it remains to be seen whether Netflix can maintain its position as a market leader by consistently delivering top-tier content.

Real Operations Make for Good Moats

But the implications reach well beyond Netflix and the media. 

Perhaps the title of this article should have been “Netflix, DVD by Mail: The Easier Path to Building Operational Scale-Based Models.”

One of the most well-quoted papers on strategy and efficiency comes from Michael Porter, who argues that operational efficiency can never be a competitive advantage on its own. Operational efficiency, in his view, can be copied by others and must be part of a tightly-knit set of activities to generate a competitive edge. 

Despite this, I believe operational moats can potentially be more enduring. In fact, the lack of an operational component in a business model might make it nearly impossible to build a lasting moat. 

Why?

Because for most firms, technology is not the primary source of their competitive moat. Instead, the true barrier to entry lies in friction and long-term dependencies which are amplified by human behavior and interactions.

When I say human behavior and friction, I mean inertia and high switching costs, as well as the willingness (or unwillingness) to create content or take specific actions, such as serving, driving, or creating. Interactions involve creating value by mutually connecting individuals.

Long-term dependencies arise when a company that builds a warehouse, for instance, can lower its prices for the next set of customers due to a high volume of sales which have covered a significant portion of the initial cost. The cost per unit only decreases. 

In essence, it is the physical aspect of the service that creates competitive moats and deepens them over time. These elements, which are rooted in human behavior and historical dependencies, contribute to the establishment and fortification of lasting moats.

In the long run, companies that merely rely on betting on the next best product or technology may struggle to maintain their competitive edge.

The advantage of Amazon’s initial e-commerce model was fleeting (anyone can build an e-commerce store), but it was their supply chain capability and Fulfillment by Amazon (FBA) services that set them apart from other competitors, such as Shopify. 

Consider Facebook. The platform’s value comes from connecting actual people and encouraging them to share stories. But bringing people to share content is hard. That’s the achievement, not the technology behind Facebook. Similarly, YouTube’s success hinges on users’ willingness to upload and watch videos. Airbnb’s competitive advantage arises from people willing to host and stay at other people’s homes, while Uber relies on drivers and passengers engaging in ride-sharing. It’s not the platform or the technology. It’s the ability to mobilize people to make decisions and help a firm use those decisions to deliver a more competitive product or service to the next set of customers. 

It’s all about people and infrastructure. Not technology. Technology is merely the enabler. Not the differentiator. 

For Netflix, in its early days, the primary constraint was its supply chain rather than its content. However, when the focus shifted to content quality, their moat was no longer as strong. Although it may be difficult to cancel a subscription, customers still do so.

Lasting competitive moats are built on human interactions (network effects), behaviors (high switching costs), and long-term dependencies (significant investment in infrastructure building scale-based advantages), rather than technology alone. 

The issue is that technology becomes easier to build and replicate. Midjurney, the Chat-GPT like model for text-to-image, has only 11 employees. The next text-to-image service is already building a better product. 

As technology continues to become a thinner and thinner layer, it will be increasingly easier to replicate unless companies can incorporate people’s friction and interactions, logistics, or care (as in health care delivery) into their services. This will make it harder for competitors to replicate, solidifying the differentiating aspects of the business model.

The last few months have seen one of the most significant leaps in AI and machine learning, putting many firms, business models, and jobs at risk. If you don’t want your firm or job to become obsolete, make sure you include an operational component or friction to your business model.