The Physics of Competitive Advantage

The forces reshaping markets are not new. They have just become impossible to ignore.

The media business runs on a simple equation. Minutes drives revenue and quality content drives minutes. Content investments are measured by the minutes produced by that title. Every streaming platform’s marketing and content acquisition strategy traces back to some version of that formula.

Turns out, for streaming media, that math is wrong.

At Plex, we started digging into how users actually behaved once they arrived. What we learned was the opposite of what we expected. All of our efforts and spending were optimizing for the wrong outcome.

The force we weren’t looking for.

The titles that drove a user to Plex were rarely the ones that drove their total minutes. Premium titles, the ones with the strongest brand recognition and the most marketing support, often produced significantly lower total engagement per user. Users came in, watched what they came for, and left. Their goal was to watch that specific title. The search or ad campaign they clicked on solved their problem. Plex was an invisible delivery mechanism.

Lower-profile titles produced something entirely different. Users came in, stayed, browsed, and discovered things they didn’t know they were looking for. They came back. Over and over. Their goal was to find something to watch. Plex solved their discovery problem.

We started measuring what we called “halo minutes”: the total minutes a user accumulated beyond the title they entered on, expressed as a multiple of that initial title's direct minutes. The titles with the biggest halo were rarely the ones we predicted. They were often ones that weren't quite what the user was looking for, leaving them to look for something else to watch.

That discovery changed the equation. Minutes drive revenue, engagement drives minutes, and discovery drives engagement. Halo minutes were the measure of a title’s impact on discovery, the force multiplier. Efforts applied to high-halo titles produced nonlinear outcomes across all other titles. What drove the halo was not always predictable, so our systems had to sense it early and respond quickly.

The halo effect impacts all businesses, industries, and markets. It’s a fundamental part of the physics of dynamic markets. It's the ability to recognize, measure, and leverage those forces that sets you apart.

Three compounding forces.

There are three fundamental forces at play in competitive markets, each well understood and key components of successful growth strategies. Individually.

Feedback loops are a foundational mechanism. A force acts on an input. That action produces an output which feeds into the same input, amplifying the external force. The next cycle is stronger than the last. The result is nonlinear growth even when the external force remains constant.

Network effects are a connective force. The value doesn't loop back through a single input. It lives in the connections between participants, so each new participant raises the value of the system for every other one at once. The amplification is distributed across the whole network rather than routed through one channel. This is why winner-take-most dynamics are so common in networked markets. The network's value accelerates as it grows, and the gap between the leaders and the followers widens faster than any challenger can close it.

Halo effects are orthogonal forces. The input force produces outputs impacting entirely different systems, amplifying those systems' own inputs. The output doesn’t loop back, it radiates outward, compounding the impact downstream.

Typically we think of and measure these forces as independent systems. In reality when they interact, intentional or not, they produce something categorically different. Something far more powerful.

Halos designed and discovered.

In the early 2000s, Apple provided a clear illustration of the impact of halo effects across both modes.

iTunes was architecture, not serendipity. The iPod was compelling hardware but getting music to the device created real friction. Ripping CDs was labor-intensive. Napster taught us how to download what we wanted. Neither were accessible to the masses, and not particularly legal. Apple created iTunes to remove this friction. A legal, low-cost, and seamless access to music designed specifically to drive hardware adoption. The halo effect was the point.

What Apple did not expect was the impact on Mac sales. iPod owners who had never owned a Mac started interacting with Apple software regularly. The brand relationship deepened. When those users considered their next computer purchase, Apple was no longer a stranger. The iPod was generating outputs in a system Apple had not aimed at. That is a discovered halo effect, and it contributed as much to Apple's resurgence as the hardware itself.

One action. Two halo effects. One designed, one discovered. Both real. Both compounding.

What happens when the forces interact.

Amazon is where the argument becomes fully visible, because the story happened twice. Each wave was larger than the one before. Neither was planned. Both changed the architecture of their respective markets, not just the competitive position of the company that triggered them.

Amazon started as a bookstore. The feedback loop was straightforward: more customers meant more selection, lower prices, and faster delivery, which attracted more customers. The loop compounded. But the more consequential outcome was not the growth of the bookstore itself. It was what the bookstore required Amazon to build.

To fulfill orders at scale, Amazon built logistics and distribution infrastructure that no physical retailer could match. That infrastructure did not just improve Amazon's economics. It radiated outward. Marketplace sellers got access to fulfillment capability that had previously been available only to large retailers. Consumer expectations around delivery speed shifted across the entire retail category. Physical retail, which had already been under structural pressure from Walmart's logistics innovations and Dell's supply chain compression, could not absorb the new expectation. The impact on local retail was not Amazon's goal. It was an emergent outcome produced by the interaction of Amazon's logistics halo and feedback loops in an already stressed market.

Then the second wave arrived.

To run the commerce platform, Amazon built one of the most sophisticated technology infrastructures ever assembled. Engineering teams were spending the majority of their time on what Jeff Bezos called undifferentiated heavy lifting: the basic plumbing every project required before any actual work could begin. The solution was internal. Build a shared infrastructure layer that any team could access on demand. At an executive meeting in 2003, the leadership team recognized they might have something that extended beyond Amazon's own needs.

AWS launched publicly in 2006. Today it generates more revenue than the retail business that created it.

The halo effect from the commerce infrastructure investment became the input to an entirely new feedback loop in cloud computing, which then generated its own network effects as more developers built on it, attracting more enterprise customers, attracting more developers. The emergent outcome was not a better Amazon. It was the reorganization of global IT infrastructure. On-premises enterprise computing did not decline because cloud was a superior product decision. It declined because the feedback loops and network effects running through the AWS ecosystem made the switching costs in one direction asymmetrically higher than the other, and that gap compounded every year.

Andy Jassy, who built AWS, said they did not have the audacity to predict it. That is not modesty. That is emergence, stated plainly by the person who lived it.

Emergent behavior in markets.

Emergence is not a feedback loop, network effect, or a halo effect. It is what happens when all three interact recursively across independent systems, producing outcomes that none of the individual forces could have generated alone and nobody could have fully predicted.

The retail disruption was not caused by Amazon's logistics. It was caused by Amazon landing in a market already under structural stress, compounding the impact of Walmart’s hub-and-spoke logistics, Dell’s JIT manufacturing, and shifting consumer expectations on cost and speed. Together they amplified the impact and accelerated the timeline. Recursion produced the outcome. The depth of the recursion determined the scale.

You already know the music industry story. You know what happened to the album model, to physical retail, to major label distribution power, to independent artist access, to the behavioral foundation that made streaming possible. That is emergence. The inputs were iTunes, a generation trained on Napster, and digital device adoption curves. Nobody planned what came out the other side.

This isn’t business strategy, it’s physics. These forces are impacting your business, today, in real-time. They have been running through your market since before you entered it. The organizations that have compounded the most advantage over the past two decades were not the ones that predicted emergence. They were the ones with systems capable of sensing it early and responding before anyone else.

Emergent strategy.

Strategies built on feedback-loops, network effects, or halo effects aren’t new, or even rare. Social networks and platform business models are built on the principles of network effects. Feedback loops are the basis of Product Led Growth and growth marketing in general. Integrated advertising derives value from expected halo effects across campaigns. The “or” is doing a lot of work in that first sentence.

These strategies are designed and measured in isolation. Predictability is the goal. The focus, and definition of success, is based on projections and achieving expected outcomes. Each force in isolation produces real results, so it’s easy to believe that one is enough.

But the interaction of all three, running simultaneously and recursively across multiple systems, is where the compounding becomes truly nonlinear. Not a marginal improvement on building one force well, a different category of outcome entirely. The organizations generating those outcomes right now are not smarter or better resourced. They are more deliberate about building the conditions that let all three interact.

Emergence, by definition, cannot be designed for. The continuous interaction of forces across large numbers of independent systems produces outcomes no planning process can anticipate. That is not a failure of planning. That is the nature of the system. The right response is not prediction, it’s the ability to sense when emergence is beginning and adapt.

At Plex, we recognized a new way to measure our investment and capitalize on the halo effect, even when we couldn’t predict which titles would drive it. Social features, user reviews, and shared lists added network effects and feedback loops to the system. The result was significant growth in engagement while dramatically reducing spend.

The physics are inescapable. Complexity is the market reality. Leveraging those forces is the competitive advantage.

When emergence happens in markets, the math changes. The companies that thrive change the equation when it does.