Every Transaction You Process Is Starting from Scratch
Think about what happens when you close a new client today.
You find them. You diagnose them. You deliver the engagement. You close it out. Maybe they refer someone. Maybe they don't. Then you start again.
The work from that engagement -- the patterns you noticed, the benchmarks you built -- mostly lives in your head. Or in a file no one else can use. Or in a case study that gets published once and forgotten.
Parker, Van Alstyne, and Choudary draw a clean line in Platform Revolution: a pipeline business creates value through a linear sequence. A platform business creates value by facilitating exchanges between producers and consumers -- and becomes more valuable with each exchange.
Three Types of Network Effects You Need to Engineer
You don't just build a network. You engineer the effects that make the network self-reinforcing. There are three types.
The Tipping Point: What It Looks Like When the Network Pulls
Below a certain threshold, you are pushing. Above the tipping point, the network pulls.
The Governance Trap: Growing Fast and Dying Slowly
Andrew Chen calls it "Eternal September" -- what happens when a network built by high-quality early adopters gets flooded with mainstream participants who dilute the quality that made it valuable.
The deadly decision is lowering certification standards to hit growth targets. Every book on platform strategy warns against it: if you lower the bar, you grow faster and die sooner.
Governance evolves with scale -- founder-led at 1-25 partners, council-assisted at 25-100, community-governed at 100-500, self-governing at 500+. The transition from direct control to distributed governance is one of the hardest psychological shifts a founder must make. But it is the only path to a business that operates without you at the centre of every decision.
The INTEGRATE pillar is where most methodology businesses stall. They build the methodology. They train the partners. They build a sales system. And then they assume the network will naturally compound. It won't. Not without deliberate engineering. Same-side effects require designed peer learning infrastructure. Data network effects require structured collection from Day 1 -- not as an afterthought in Year 2.
- If different partners use different formats, ask different questions, or deliver different outputs -- your data is not comparable. No benchmarks means no data flywheel. You're collecting information but building no moat.
- Be specific. Count the deals. If the answer is below 30%, your network is not yet pulling. The cross-side effects are not yet active.
- Name the specific things. If the list is long, you have not yet made the pipeline-to-platform transition.