What it means
Most products deliver roughly the same value to the thousandth customer as to the first. Network effects break that pattern, because each new participant adds something other participants can use, whether that is another person to message, another seller to buy from or another data point that sharpens a recommendation.
Economists separate direct from indirect effects. Direct effects come from users of the same type, such as members of a messaging app; indirect effects come from users of different types on a platform, such as drivers and riders, or merchants and cardholders, where each side attracts the other.
The commercial consequence is that early growth is painfully slow and later growth can be self-sustaining. Below a certain size the product is not useful enough to hold anyone, which is the cold-start problem, and above it the same product begins to sell itself and acquisition costs fall.
Businesses with genuine network effects tend to hold their market position longer, because a competitor with a better product still has to persuade users to abandon everyone they are connected to. That is a defence built on other customers rather than on features, which is why investors pay close attention to whether an effect is real.
Two cautions are worth keeping in mind. Effects are often local rather than global, so a delivery platform dominant in one city has no advantage in the next one, and networks can suffer congestion or quality decline that makes extra users a negative rather than a positive.
In practice
Real-world examples.
Example
A freelance marketplace struggles for a year with 40 clients and 60 writers, because clients see too little choice and writers see too little work. Once it passes a few hundred on each side, jobs fill within hours, and both sides start arriving through word of mouth rather than paid advertising.
Example
A card payments network grows because merchants accept the card when enough consumers carry it, and consumers carry it because enough merchants accept it. Neither side would join for the card itself; they join for the other side of the network.
Example
An accounting software firm builds an app marketplace around its product. Each new integration makes the core product more useful to accountants, and each new accountant makes the platform more attractive to developers building the next integration.
Formula
Calculation
There is no accounting formula, but the scale of the effect is often illustrated with the number of possible connections in a network, sometimes called Metcalfe's law:
Potential connections = n x (n - 1) / 2, where n is the number of users
Take a professional community platform. With 100 members, potential connections are 100 x 99 / 2 = 4,950. Double the membership to 200 and connections become 200 x 199 / 2 = 19,900, roughly four times as many from twice the users. At 400 members, connections reach 400 x 399 / 2 = 79,800.
If the platform earns an average of $2 of annual value per potential connection, that is $9,900 at 100 members, $39,800 at 200 and $159,600 at 400. The formula overstates reality, because most users never interact with most others, but it captures why value can grow faster than headcount.Case study
Seen in the real world.
Tidewater Table is a fictional restaurant booking app used here as an illustrative example. It launches across three cities at once with 12 restaurants in each, and the results are dismal: diners open the app, find almost nothing near them, and never return. Customer acquisition cost sits at $22 per active diner and monthly bookings average 0.4 per diner.
The founders make an unpopular decision and withdraw from two cities entirely, concentrating every dollar of sales effort on one. Within six months that single city has 60 participating restaurants, coverage is dense enough that most searches return something walkable, and bookings rise to 1.5 per diner per month. Acquisition cost falls to $9 as diners begin recommending the app to friends.
Only then do they reopen the second city, and this time they sign restaurants street by street rather than spreading thinly. The illustrative lesson is that network effects are usually local and have a threshold: below it, marketing spend leaks away, and above it, the same spend compounds.
Watch out
Common mistakes.
- Calling any fast-growing product a network business, when growth is actually coming from advertising spend rather than from users making the product better for each other.
- Confusing network effects with economies of scale, which lower the cost of production rather than raising the value delivered to each customer.
- Assuming a network advantage in one region or segment automatically carries into the next one, when most effects are local and have to be rebuilt.
Questions
People also ask.
How do you tell whether a network effect is real?
Check whether engagement or retention per user improves as the user base grows; if the numbers stay flat, growth is coming from somewhere else.
Can a network effect work against a business?
Yes, if quality falls as the network grows, through spam, congestion or poor moderation, extra users reduce value and the effect reverses.
Do network effects guarantee a monopoly?
No, users often participate in several networks at once, and low switching costs let competitors coexist even in networked markets.
From the founder's library
Take it further with the book.
Build your financial confidence beyond this definition. Shihan's full-length guide, Accounting Fundamentals, takes the same plain-English approach and turns it into a complete, practical playbook for non-finance managers, business owners and students - with chapter-end quiz answers and presentation slides included.
25% off with code MMHQ25, applied at checkout. Priced in USD - checkout may show the equivalent in your local currency.
View the book and save 25%