Concept · Platforms

Same-side network effects

When members of the same group change one another’s usefulness or competition

More peers can make a group more useful—or more crowded.

A same-side network effect links members of the same participant group. The boundary is the interaction being studied. People collaborating on a shared file may be peers even when one pays for the account and another does not. A marketplace’s sellers competing for a buyer are also on the same side for that transaction. Their effect on one another can be beneficial, harmful or mixed.

In one sentence

A same-side network effect occurs when participation or activity within a defined group changes the service’s value to another member of that group.

The practical decision is whether to recruit peers, improve how they interact or control an adverse interaction. A participant might need a colleague’s contribution rather than a larger public membership. Conversely, another seller could increase competition for the same request. The operator should specify the benefit and burden separately rather than treating growth in the group as a uniform improvement.

The CMA’s distinction between direct and indirect network effects provides the classification. It does not establish the sign or size of the effect in a particular product. The decision analysis must follow the task and evidence, including the possibility that the product is useful without peer participation.

Decisions and participation patterns

Define the task before choosing a participation measure.

Coordinating peers

Relevant members make shared work or communication possible.

Contributing peers

Other members supply information or contributions that can help a task.

Competing peers

Members seek the same opportunity, attention or capacity and may impose a burden on one another.

A continuum, not a switch

The comparison moves from unrelated membership to relevant peer interaction; its effect can remain adverse and has no automatic positive endpoint.

LowUsers treated independentlyHighValue depends on peers in the same group

“The relevant peer is the one who changes the task, not simply another registered member.”

Why it matters

Group membership is not sufficient. The operator needs a useful relation: access to the same work, compatible interests or a contribution the recipient can use. Recruiting people without that relation may enlarge the count while leaving the task unchanged. A permission problem may require a better handoff instead of an acquisition campaign.

The operating conflict can involve reach and control. Easier invitations can help a team bring in the necessary collaborator, while broader access may expose information or create administrative work. A product should distinguish the intended collaborator from everyone who could be invited. The relevant boundary follows the actual task and authority to participate.

There is also an economic conflict within some groups. Additional sellers may broaden a market for buyers while intensifying competition among sellers. The cross-group benefit does not imply a positive same-group effect. An operator evaluating recruitment needs to ask whose outcome improves and whether another participant bears the cost.

Real-world examples

The same concept shows up in different ways across industries.

DropboxRelevant collaboration differs from aggregate reach

→

Dropbox’s registration statement describes sharing and invitations alongside a self-service acquisition model. The choice supports a handoff from an existing user to another participant. For a shared-file task, the relevant peer is the person who can supply, receive or work with the file. The filing does not isolate the value created when that peer joins, nor show that every invitation produces sustained collaboration. Referral acquisition and the usefulness of shared work therefore remain separate claims. A useful product decision would examine a failed sharing task before assuming that more invitations solve it.

AirbnbReviews can help a later participant, but accuracy matters

→

Airbnb’s FY2020 disclosure identifies reviews as information used in decisions and quality standards, and warns about unreliable reviews. A guest’s contribution may help another guest assess a stay, which is a plausible same-group information mechanism. The actual policy choice creates a place for those contributions; the filing does not prove their accuracy or isolate their effect on booking. The rival account is that price, availability or other information determines the decision. The operator must evaluate whether the contribution improves a relevant judgment rather than counting reviews as value by definition.

When it breaks

A group definition can conceal a cross-group effect. A developer attracting a customer and a customer attracting another customer are different relationships. State the role each person occupies in the task before assigning a network label.

Shared infrastructure and peer participation can be confused. A hypothetical team tool may work better after the operator improves synchronization, even with the same collaborators. That is product improvement unless the evidence also identifies a change caused by peer activity.

Contribution quantity can obscure quality. Repetitive, inaccurate or irrelevant information can make a task harder. A same-side effect can reverse when additional participation adds burden rather than a usable contribution.

Key takeaways

01
Define the participant group through a specific interaction.
02
Identify the relevant peer and the work that their participation changes.
03
Separate coordination benefits, information quality and competition within the group.

Sources

Merger Assessment Guidelines: platforms and network effects · Competition and Markets Authority. Updated September 3, 2026; Two-sided platforms and network effects section
Dropbox registration statement on Form S-1 · Dropbox / SEC. Our Business Model: signup, acquisition and paid conversion paragraphs; paying-user definition printed p. 13; registration/conversion risk p. 15; referral and enterprise selling risk; Sales and Marketing
Airbnb 2020 Form 10-K · Airbnb / SEC. Our System of Trust, printed p. 6; reviews and reputation risk, p. 21; Host Protections, p. 117