The Source Layer Model

A source network has three layers, each serving a different function. Layer 1 is primary sources — the original publishers of information: vendor blogs, GitHub repositories, regulatory filings, patent databases, earnings transcripts. These are the closest you can get to ground truth. Layer 2 is aggregation sources — curated feeds, newsletters, analyst reports, and industry publications that synthesize primary sources. They save time but introduce editorial bias. Layer 3 is human sources — conference conversations, community discussions, and professional networks where unwritten knowledge circulates before it reaches any published source.

1. Layer 1: Primary Sources Build a list of 20-30 primary sources: vendor engineering blogs, official changelogs, GitHub release pages, regulatory agency publications, and patent databases. These sources publish the raw facts before anyone interprets them. Your competitive advantage starts here — most analysts skip primary sources and rely on aggregators.
2. Layer 2: Aggregation Sources Curate 8-12 aggregation sources: industry newsletters, analyst firms, tech publications, and curated RSS feeds. These sources save time by filtering and summarizing. But always verify the most important claims against Layer 1. Aggregators can misinterpret, selectively quote, or lag behind primary sources by hours or days.
3. Layer 3: Human Sources Develop 5-10 human source relationships: peers at industry events, community members in relevant Discord or Slack channels, contacts at vendor organizations (ethical and public-facing), and academic researchers. Human sources provide context, interpretation, and early warnings that never appear in published sources. They are the hardest layer to build and the most valuable.