Network Effects, Neutral Network Effects, and Anti-Network Effects.


TLDR: Most founders are aware of the phenomenon that some products become exponentially more effective as the number of users increases. However, in many situations, such as the development of a contemporary marketing campaign or the birth of a new business model, there’s a decrease in effectiveness as a function of the number of people who copy that approach. I’m calling this phenomenon ‘anti-network effects’.

Network Effects

It works better with more people.

Most of us are familiar with the concept of network externalities.

The idea of demand sided economies of scale: as the product acquires more users, its effectiveness increases.

When Robert (Bob) Metcalfe tried to sell his invention, Ethernet, in the early 80s he made the case that his customers’ Ethernets needed to cross a certain threshold to reap the benefits of what we now call Metcalfe’s law.

‘’3Com sold $1,000 cards that connected desktop computers into a network. Here was the payoff: The cost of installing the cards at, say, a corporation would be proportional to the number of cards installed. The value of the network, though, would be proportional to the square of the number of users. Multiply the number of networked computers by ten and your systemwide cost goes up by a factor of ten but the value goes up a hundredfold.’’ *1 *2

(Metcalfe, 2007).

(It was actually George Gilder in 1993 who coined the term Metcalfe’s Law when he discovered Bob’s 1980’s Ethernet sales presentation slides.)

One of the reasons Snapchat was able to blow up so quickly is because if you’re a high school teen with Snap, it’s pretty boring.. unless your friends have it too.

So there’s a viral coefficient (or basic reproductive number R0 in epidemiological terms) built-in that’s greater than 1.

One user ‘’infecting’’ on average more than one, and so it spreads exponentially.

We looked at flipping the popular funnel approach in Which Side Is Up?

Neutral Network Effects

How well it works is uncorrelated with the number of people.

There are many situations where there’s a neutral relationship between the number of people who adopt it vs. its usefulness. *3

It either works or it doesn’t.

Lifting is one such example.

You can teach a thousand people to increase muscle mass and decrease body fatness in order to improve the overall body composition but Lara’s gains won’t take anything away from Clarke’s.

Physics is another example where neutral network effects come into play.

Nature behaves how nature behaves.

We create mathematical frameworks that we can overlay onto reality and then depending on how well those frameworks correlate with reality, our confidence in them grows.

They either work (have predictive power and internal consistency) or they don’t.

Newtonian physics works perfectly when you’re dealing with non-relativistic speeds and low gravitational fields (sending a rocket to the moon) but it breaks down when you’re dealing with high speeds (photons, you’ll need special relativity), high gravitational fields (supermassive stars, you’ll general relativity), and subatomic matter (electrons, you’ll need quantum mechanics).

There’s no increase in corrosion of those frameworks as the number of people increases.

And while one can argue that innovation in finding frameworks that are better at describing the behavior of the universe, is correlated with how many people have access to those frameworks, ultimately, the laws that govern the behavior of the universe don’t care about whether or not we discover them.

They’ll continue to operate just fine whether 0 people, 10 people or 10M people are aware of them.

Anti-Network Effects

The more people use it, the more its effectiveness declines.

Which brings us to the reason I wrote this essay.

A concept I’d like to introduce which I’m calling anti-network effects.

Anti-network effects describe those situations where effectiveness declines as a function of the number of users.

The reason this concept matters is because so much of the business world is governed by anti-network effects.

Anti-network effects are single-handedly responsible for the absence of marketing formulas.

In a sense, it’s a signal that someone fundamentally doesn’t grasp marketing when she sells a course or book teaching plug-and-play marketing. This is the same category mistake many agencies make with their interchangeable marketing approach. In manufacturing, copying is rewarded by an increase in efficiency and thus effectiveness. In marketing an increase in efficiency destroys effectiveness. Inefficiency is rewarded, so marketing ends up being more similar to sex than it is to manufacturing. (Youngling, 2019).

No self-respecting marketer will ever have the audacity to write a book with formulas for marketing in the way one can write a book with fundamental mathematical theorems (made possible by neutral network effects).

This is because good marketing is marketing that works.

Marketing that works is marketing that stands out.

Marketing that stands out can only be copied so often until it doesn’t stand out anymore.

So there’s an inverse relationship between effectiveness and adoption.

Unfortunately, perhaps due to loss aversion, our obedience orientated educational system or simply because of capitalistic forces, people prefer to copy rather than invent.

So when the founder of TOMS comes up with his philosophy ‘’One for One’’, buy one and we’ll donate one to a child in need, lazy ‘marketers’ copy the best ‘tactic of the day’ and implement it in an effort to maximize revenue, completely missing the forest for the trees and overlooking what made it work… sincere generosity.

This is what has happened to pretty much all effective marketing.

Kendrick comes up with something new, risky, and generous.

Not because he’s seeking to maximize revenue, obviously, (there’s no proof that it’s good business yet), but because he believes it matters.

It starts to work, ‘’ marketers’’ see that and now copy-paste it into their business.

Consumers used to love email marketing. Why? Because there was no money in it yet, so the people who used email marketing didn’t do it for the money, but to genuinely add value.

As more and more people started to use it as a tactic, it became less about the business serving the user and more about how the user can serve the business.

Zooming out, you see anti-network effects occur in disruptive innovation as well

A business is an exchange between Sara who has something and Blake who wants it and gives her money for it.

There’s a market inefficiency, meaning, the asset price doesn’t reflect all available information yet because her product is so new.

Other entrepreneurs want to make money as well, so capitalistic forces rush to fill this ‘’low-pressure void’’ of market inefficiency.

There’s money to be made and people want a slice of the pie.

As more and more entrepreneurs start to copy Sara, prices drop.

Eventually, they reach a (stable by definition) Nash equilibrium, where a further decrease is impossible (you’ll go bankrupt) and an increase is impossible (consumers will just by from the competition).

We looked at this extensively in The Art Of Business, Where Science And Business Depart.

The only way to break this Nash equilibrium is when an existing (or almost always new) player makes a new move not currently accounted for and completely changes the game, aka disruptive innovation.

James Dyson didn’t make a cheaper, better, bag vacuum (more of the same). He made a totally different, bagless, incredibly expensive vacuum (completely changing the game).

As my mentor, Ogilvy’s Rory Sutherland so aptly puts it: ‘’The opposite of a good idea, can be another good idea.’’

It’s important to realize that this is not possible in mathematics.

You can prove that pi is an irrational number by proof of contradiction. You assume it’s rational (can be expressed as p/q) then show that that will yield a contradiction, leaving you with the only possible conclusion that it, therefore, has to be irrational.

The opposite of ‘’pi is irrational’’ is demonstrably false.

But in business, you can win by being the cheapest (H&M) and by being the most expensive (Gucci). 

You can win by treating your workers horribly in an effort to maximize revenue (Amazon) or by treating them like they’re family (Zappos) or by treating your customers like they’re family and fighting with vendors for lower prices so you can make the product even cheaper for your customers (Costco).

The Golden Age Of Marketing

Marketing is not a modern concept.

It can be traced back to the Egyptians using papyrus to make sales posters, commercial messages and political campaign in the ruins of Pompei, or images on signs associated with a trade (i.e. boot for a cobbler) used in the Middle Ages when the general populace was unable to read.

But modern marketing started with the rise of newspapers and advertising agencies.

The late half of the 19th century and most of the 20th century (esp. 1950–2000) was an incredible time for marketing.

Early on, marketing equaled having enough money to afford running an ad informing people about your product and then using that profit to buy more ads.

But as competition started rising, the need for creativity grew, resulting in some of the most beautiful, marketing campaigns designed to alter human behavior.

My favorite being N.W. Ayers who were approached by De Beers corporation to make people want diamonds (post-Great Depression of 1930). The solved it by inventing the concept of the diamond engagement ring.

I wrote about this in Conjuring Up Value: Why You Want An Engagement Ring.

Anti Network Effects In The Post-Modern Marketing Era

But as we started shifting into the post-modern marketing era, where we moved off of advertising in newspapers and TV (where ‘’shelf space’’ is limited), and onto the web (where there’s no physical limit anymore to how many people can market), creativity surprisingly disappeared.

And in flooded: SEO best practices, Guaranteed results Adword campaigns, and How to do Facebook ads for dummies.

With the cost of experimentation being so low, you could’ve offered the highly plausible hypothesis that creativity in companies’ marketing would massively increase.

But instead, the opposite has happened. People are more afraid than ever and simply copy what everyone else is doing.

This is made even worse by the fact that most creative you see online doesn’t come from trained professionals but is thrown in as a freebie by the vendor. This, coupled with the fact that the vast majority of agencies copy everyone else and are thus interchangeable, explains why so much of post-modern marketing is absolute shit (in scientific terminology).

We now live in an age where there aren’t any good marketers left.

And I now believe marketing is completely dead. The word marketing has so many negative connotations that it’s beyond repair.

Which is why those of you familiar with Youngling & Feynman’s thesis know I’ve started referring to marketing as pragmatic, behavioral psychology:

How you can influence human behavior in the real (practical, not in a lab) world in order to achieve a certain goal (including but not limited to sales).

This will be the future of marketing.

Pragmatic, behavioral psychologists who use the guidelines of human behavior and blend that with unique art (not repeatable science) in order to maximize results and minimize the negative results that come from anti-network effects.

I wrote about this in Advertisers Are Clueless About Advertising…

Thank you for reading. I hope you were able to find some value in it.

*1 N-squared because 1 user can connect to all other users. If there are n users, that means n users can connect to n-1 users, which yields a ‘’total value’’ of n*(n-1) but for simplicity’s sake we just call it n squared since value isn’t cardinal but ordinal at best, meaning you could perhaps argue X has more value than Y but it would be laughable to express value in a decimal notation (unless you’re a Keynesian economist.. just kidding).

*2 It’s important to realize that the concept of network effects was useful early on not so much because of the exponential increase in usefulness but specifically because of the juxtaposition between linearly rising costs and exponentially rising value. N times more networked computers makes the cost go up by n*cost, but the value by (n-1)*n.

*3 Neutral network effects isn’t that useful of a concept. You’re basically talking about the absence of network effects. The only reason I introduced it is so I could juxtapose network externalities against anti-network effects in order to demonstrate that the absence of network effects is materially different from the concept of anti-network effects.


Metcalfe, R. (2007). It’s All In Your Head. Retrieved 31 July 2019, from

Youngling, R. (2019). Which Side Is Up? — Youngling & Feynman. Retrieved 15 August 2019, from

Youngling, R. (2019). Marketing Is Sex, Not Manufacturing — Youngling & Feynman. Retrieved 15 August 2019, from

RJ Youngling