on

# Marginal costs

This post is part of a series on product design patterns.

I get asked a recurring question on product design that is always a variation of **“will it scale?”**.

The answer often comes down to going through an analyses of **marginal costs**, and it normally boils down to trying to find a business model where the **marginal costs** are *virtually zero* even at the cost of high **fixed costs**.

If I sent you this link as a result of giving feedback to your business model, chances are that your business incurs into a cost profile that has a non-zero marginal cost, putting a ceiling to its growth.

**Marginal cost** is the cost added by producing one additional unit of a product or service. As opposed to **fixed costs** that are constant independently of the quantity of goods or services you produce, **marginal costs** is the direct cost associated with growing your product user base.

Take restaurants as an example. For every new customer, the restaurant has to buy the ingredients to prepare their meal and pay extra staff. For restaurants, as an approximation, the **marginal costs** of running your business grows **linearly** with each new customer: `x`

new customers incurs a cost of `A * x`

, where A is a **constant** that represents the monetary value of serving that number of customers. Restaurants also have **fixed costs** (e.g. paying rent or their licenses, which needs to be paid independently of how many customers they serve), so their total cost is closer to `cost(x) = A * x + B`

.

Digital services, on the other hand, have a massively different **cost profile**. Take a mobile game, published on an app store as an example. For digital games, as an approximation, the **marginal costs** stays virtually **zero** with each new customer: there aren’t any distribution costs and duplicating software is free. On the other hand, developing a successfull game has a massive **fixed cost**: hiring engineers, UX designers, story tellers, etc is expensive (in comparison to, say, rent for a restaurant). So, even while the **fixed costs** are huge, the cost profile of selling games is closer to `cost(x) = C`

where `C`

is the **fixed cost**.

Even if `C`

is fundamentally larger than `B`

(e.g. say `C ~= US$ 5M`

whereas `B ~= US$ 1M`

), what makes game development **scale** is that the `cost(x)`

function doesn’t change as `x`

customers grows. That is, for small values of `x`

, `cost(x) = A * x + B`

is smaller for restaurants, but for large numbers of `x`

, `cost(x) = C`

is significantly lower than `cost(x) = A * x + B`

.

Here are a few examples of business models with small **fixed costs** but large **marginal costs**:

- restaurants
- coffee shops
- doctor offices
- schools

And here are a few examples of business models with a large **fixed cost** but small **marginal costs**:

- digital services (e.g. a social network, a search engine, a game)
- consumer electronics (e.g. an ipad, a laptop, a phone)
- media content (e.g. movies, books, videos, podcasts, articles, etc)

Along with breaking the chicken and egg problem, this is is one of the two most frequent pieces of feedback I give on product design.

Find a business model where the cost of growing your business is independent of the number of customers you serve. Find a cost function `cost(x)`

that doesn’t have `x`

on the right side (or is associated with a constant `A`

that is *virtually* zero).