SC

Collaboration & exchange network effects

What are Collaboration & exchange network effects?

Participation by relevant colleagues or counterparties makes communication, shared work or information exchange more useful.

How Collaboration & exchange network effects work

Participation by relevant colleagues or counterparties makes communication, shared work or information exchange more useful.

How companies won with Collaboration & exchange network effects

From the win chains in Strategy Canon case studies: what the winner did, and what its rival did at the same step.

vs. HipChat

Slack

Collaboration network effect

What it did
Shared channels across companies from 2017; more than four per paying customer by 2019
HipChat
Guest links, the option Slack declined; no cross-company network
Read the Slack case study →
vs. InVision

Figma

Network effects: colleagues and the talent pool

What it did
Designers, PMs and engineers in the same files
InVision
Figma overtook it for prototyping by 2019
Read the Figma case study →
vs. Sermo

Doximity

Colleague network tipping by hospital, city, specialty

What it did
Tangney: tipping came hospital by hospital, city, specialty (2014)
Sermo
U.S. membership plateaued near 130,000 by 2012, growing about 4% a year
Read the Doximity case study →
vs. Digg · 2011–24

Reddit

Network effects inside each community

What it did
Over 100,000 active subreddits, 500+ with a million subscribers
Digg
Users followed topics, not people, so leaving Digg was easy (Gottheil, 2010)
Read the Reddit case study →

Also tagged: Bloomberg · Facebook / Meta · LinkedIn · Procore Technologies · Snowflake · Xero

The question to ask of a company

Which additional colleague or counterparty improves existing participants’ work, and can they exchange through competing tools?

Who names it

AuthorWhat they call it / where it appears
NFXPersonal utility (adapted to collaboration and exchange)
Note. This label adapts NFX’s personal-utility category to working groups and counterparties. Information exchange does not establish learning from pooled usage data. Unrelated users may add no value.