Cross-side network effects exist when the value one group of users gets from a platform depends on the size and activity of another group, such as buyers and sellers. They are also called indirect network effects.
Table of Contents
Types of cross-side network effects
- Positive: more of one side makes the platform more valuable to the other, as when more sellers bring more choice for buyers.
- Negative: more of one side makes it less valuable to the other, as when more advertisers mean more adverts for viewers.
Key factors
- Balance, not just scale. Growing one side helps only if the other side grows with it. Gains from size appear mainly where the sides are well matched.
- Which segments matter. Some users create far more value for the other side than others, so recruitment should target them.
- Gaining and keeping. A cross-side effect can attract new users or keep existing ones active, and the two need measuring separately.
- Slow reactions. Users adjust over months, so short tests understate how strongly one side responds to changes on the other.
- Rivals. Cross-side effects tie the sides together but do not stop users leaving for a better new product.
Examples
Pandora. Pandora ran a 21-month experiment on about 35 million listeners, varying how many adverts each heard. Each extra advert per hour cut listening by about 2%, and the effect grew over time. Advertisers bring revenue, but more of them drives away the listeners they pay to reach, a negative cross-side effect.
eBay. In 2024 eBay stopped charging selling fees to private sellers in the UK, first for clothing and then, from October 2024, for all categories except vehicles. The aim was to bring more listings, competing with Vinted, so that a wider choice of items would draw more buyers. It traded fee income on one side for growth on the other.
Frequently asked questions
What is the difference between cross-side and same-side network effects?
Cross-side effects run between different groups, such as drivers and riders. Same-side effects run within one group, such as users of a messaging app gaining from other users.
How do you measure cross-side network effects?
Find a change that affects only one side, such as a price change or a supply shock, and track how the other side responds. Simple growth correlations usually mislead.
Why do platforms subsidise one side?
Because that side’s presence draws the other. If sellers attract buyers more than buyers attract sellers, a platform may charge sellers little and earn from buyers instead.
Can cross-side network effects run in one direction only?
Yes. Viewers may dislike advertisers while advertisers value viewers. The effect need not be equal or even positive on both sides.
Full guide: Network effects: why they matter
Related concepts
Same-side Network Effects
Multi-sided Platform
Platform Pricing