Outline for this post:

  1. Back-to-Basics: Economies of scale in leveraging a fixed cost
  2. What happens when a fixed cost becomes variable?
  3. Areas for further thought & research

1: Back-to-Basics: Economies of scale in leveraging a fixed cost

Economies of scale, as defined by 7 Powers, produce “a business in which per unit cost declines as production volume increases.” This can result in highly attractive margins for the leading company in an industry. For example, Netflix turned the variable cost of licensing content (i.e., an annually recurring expense) into a fixed, upfront cost of buying rights to shows (and even producing them outright), thereby leveraging a fixed cost to lower the per-subscriber cost of content creation.

Let’s use a simplified example to demonstrate this: When evaluating the cost of buying a show for $10m, a SVOD (Streaming Video On Demand) provider with only 10m subs faces $1/sub in content cost, whereas a leading SVOD with 100m subs faces just $0.10/sub cost. That leads to much more attractive profit margins for the leader. Furthermore, in this situation, the leader could actually outbid the smaller SVOD provider by 5x and still earn 2x the margins on a per-subscriber basis than the #2 player. Economies of scale in leveraging a fixed cost can be extremely powerful!

While the numbers for Netflix vs peers are nowhere near this extreme, this is essentially the core of the bull thesis for NFLX (alongside SVOD being a far better value proposition than linear tv in the USA and other markets e.g. paying ~$10/month for Netflix vs ~$80/month for cable…).

The same principle historically applied to local advertising. Companies advertising in the pre-internet era primarily relied on advertising in newspapers, TV, and billboards. Here, the advert might cost $100. This would be the cost regardless of whether the advertiser had 10 customers that consumed the ad or 10m. This heavily favored larger companies who had lower fixed costs of advertising per customer. This enabled them to further outspend the competition on advertising while simultaneously generating more attractive margins.

But how has the internet changed this? What happens when a fixed cost becomes variable?

2: What happens when a fixed cost becomes variable?

It is impossible to develop economies of scale over a strictly variable cost, given the definition of economies of scale above. When a cost goes from being fixed in nature to variable, existing business models are disrupted as companies lose the potential to obtain economies of scale over that line item.

I think this has happened in digital advertising, due to how Facebook & Google charge advertisers:

Facebook & Google only charge advertisers when a potential customer clicks through the advertising link provided, instead of charging a fixed cost per advertisement shown. The advertiser then has a certain ‘conversion rate’ once customers click on the Google/Facebook ad. For example, if a shampoo advertiser pays Facebook $1 per click, and then converts one-in-ten of those Facebook users into a paying customer, then their effective advertising cost is a variable $10 per-customer-acquired, regardless of the quantity of advertising spend from the advertiser (assuming constant conversion rate).

The mechanics of Facebook & Google’s pricing models more closely align ad spend with revenues and turn what used to be a fixed cost for advertisers into something more variable. This has been a huge boon for smaller businesses who can now afford to advertise & reach consumers in ways they never could before, and indeed on a (much more) level playing field with large incumbents. It has simultaneously reduced economies of scale for those incumbents and generated tremendous wealth for Facebook and Google.

Many advertising/S&M costs are still fixed. For example, most medical devices require local, relatively fixed marketing costs in the form of sophisticated sales teams meeting with and selling to doctors, hospitals, and pharmacies. Thus, there remain substantial economies of scale in advertising for many enterprise sales & marketing efforts. However, that is less the case for advertising efforts performed substantially over the internet.

Another well-known example of turning a fixed-cost into a variable cost is what is happening with the cloud. The high cost of having on-premise servers and an internal IT department has been turned from a fixed operating cost into a variable cost-of-revenue as the cloud has enabled companies to outsource this part of their business.

Here’s Ben Thompson on why companies are moving to the cloud:

There is a secular shift in enterprise computing moving to the cloud, not because it is necessarily cheaper (although costs are more closely aligned to usage), but because performance, scalability, and availability are hard problems that have little to do with the core competency and point of differentiation of most companies.

IT costs are today much more variable than they were a decade ago. This puts smaller businesses at more of a level-playing field with incumbents than was historically the case.

Note that digital advertising, too, solved many problems for advertisers that have little to do with the core competency and point of differentiation of those businesses, while simultaneously increasing the alignment between advertising costs and revenues (CAC is the new rent, and is often now variable).

3: Areas for further thought & research

What other industries does this apply to?

  • Electric Power: companies don’t have to generate their own electricity but instead pay power utilities for on-demand access
  • Water: companies pay for on-demand access from central water utilities
  • Shipping/distribution: from being a fixed cost at every company -> UPS/FedEx/Fulfillment by Amazon etc
  • Newspapers: distribution went from being a fixed cost to, well, free
  • Credit ratings: from companies having to maintain constant dialogue with the market to relying on Moody’s for that (particularly around the issuance of new debt/equity)
  • Credit scoring: from companies’ internal efforts -> FICO/the credit bureaus
  • Recruiting: from large internal teams -> headhunting firms
  • Substack: paying a fixed cost for domain name + website template -> 10% rev take-rate
  • Apple’s App store/other platform biz’s
  • Lots of others…

How does this relate to The Conservation of Attractive Profits?

  • If a company emerges as the chief beneficiary of a technological, business model, or industry structure shift that turns their customers’ fixed costs into a variable cost, then that company can build economies of scale within the niche of providing that ‘industry infrastructure’.
  • For example, as fixed costs in digital advertising have diminished economies of scale for the consumer packaged goods incumbents this has simultaneously put smaller CPG companies on a more level-playing field while also leading to the rise of what is essentially a digital advertising duopoly in Facebook and Google (which can be viewed as infrastructure for advertising).
  • Ben Thompson has written about this here
  • Do these forces perfectly counteract one another? For example, do the profits of the mega-scaled cloud companies (primarily AWS & Azure) perfectly match profits lost from the reduction in economies of scale at other points in the value chain?

Has the internet made it harder to have economies of scale in multiple areas, but possible to have larger economies of scale over one area?

  • Obviously Amazon has multiple different economies of scale across its business… But this is not normal!
  • Look at, say, the New York Times: it used to have economies of scale in: (1) content creation, (2) printing presses, and (3) physical distribution of its newspapers. Vs today content is digital so printing presses are irrelevant and distribution of newspapers is free, so no.’s 2-3 are irrelevant. However, the NYT arguably has even larger economies of scale in content creation today than ever!
  • Does the internet tilt competitive advantages towards fewer, larger companies?