EI-301f · Module 1

Network Graph Construction

3 min read

While the layer model shows vertical structure, the network graph shows horizontal relationships. In a network graph, each actor is a node and each relationship (partnership, integration, investment, competition) is an edge. The resulting visualization reveals clusters (groups of densely connected actors), hubs (actors with many connections), bridges (actors connecting otherwise separate clusters), and isolates (actors with few connections). Each structural pattern has strategic implications.

Do This

  • Color-code edges by relationship type: green for partnerships, blue for integrations, gold for investments, red for competitive overlap
  • Size nodes by market relevance (revenue, user base, or strategic importance) — visual hierarchy helps readers navigate the graph
  • Include only relationships you can verify from public sources — speculative relationships contaminate the analysis
  • Update the graph quarterly — relationships form and dissolve, and a stale graph is misleading

Avoid This

  • Include every company in the ecosystem — a graph with 200 nodes is unreadable; limit to 40-60 strategically relevant actors
  • Map only partnerships — competitive relationships and investment ties are equally informative for structural analysis
  • Create the graph once and treat it as permanent — the AI ecosystem restructures quarterly

The structural patterns in a network graph are the strategic intelligence. A densely connected cluster of companies around a major cloud provider suggests an ecosystem that is consolidating around that platform. A bridge actor connected to multiple clusters has unique strategic value and negotiating power. An isolated actor may be pursuing a go-it-alone strategy that will either produce unique independence or leave them outside the ecosystem's value flow. The structure tells the strategy story.