a builder's codex
codex · operators · Aaron Ross · ins_seeds-nets-spears-lead-portfolio

Treat leads as a three-asset portfolio, not a single funnel

By Aaron Ross · Architect of the Salesforce.com outbound model; author of Predictable Revenue · 2026-03-03 · essay · Seeds, Nets, and Spears: The Three Lead Types

Tier A · TL;DR
Treat leads as a three-asset portfolio, not a single funnel

Claim

Pipeline planning fails when all leads are treated as equivalent. Leads come in three distinct asset classes — Seeds (referral/organic), Nets (inbound marketing), Spears (outbound) — each with its own conversion rate, time-to-develop, scalability profile, and cost curve. A mature B2B company should rely on at least two; single-source dependency is an existential risk.

Mechanism

Each lead type has different physics. Seeds compound but cannot be bought; conversion rates run 2-5x higher than Spears but they require months-to-years of relationship work. Nets scale with budget but hit diminishing returns and a market-wide ceiling on inbound demand. Spears have the lowest conversion rate but are uniquely predictable — output is directly proportional to SDR input, so they are the only lever you can pull on a quarterly horizon. Mixing them in one pipeline metric obscures which engine is broken; running the three formulas separately reveals where to invest next.

Conditions

Holds when:

Fails when:

Evidence

"No mature B2B company should rely on fewer than two types. Single-source dependency creates existential pipeline risk."

"Spears [are the] foundation of predictable revenue because output is directly proportional to input. The only lead type where you can directly control output by controlling input."

— Aaron Ross, Seeds, Nets, and Spears (predictablerevenue.com)

Signals

Counter-evidence

PLG and bottom-up products (Figma, Linear) intentionally collapse the taxonomy — there is no "lead", just usage. For those motions the three-portfolio frame can over-engineer pipeline reporting where product telemetry is the better instrument. Chris Walker's "create demand vs. capture demand" frame also reframes the same data into a two-bucket system that treats Seeds and Nets as a single demand-creation engine.

Cross-references

Open the interactive view → View original source → Markdown source →