# Playbook: Predictable Revenue Pod
**School:** Aaron Ross, *Predictable Revenue* (2011), built from scaling outbound at Salesforce (~$100M in added recurring revenue). Named framework: "Cold Calling 2.0" (specialize the rep role) and "Seeds, Nets, Spears" (classify lead sources). The core move is organizational, not technical — but the CRM is what makes the specialization real instead of aspirational.
**Deliverable type:** `journey-map` with per-stage hand-off contracts, plus a `pipeline-audit` carrying the coverage-ratio math.
---
## When to use this
- More than one person touches a deal end-to-end and they're currently doing all of prospecting, closing, and account management themselves (the classic "one rep does everything" failure mode Ross wrote the book to fix).
- You have (or want) distinct roles: someone who sources/qualifies, someone who closes, someone who retains/expands.
- You need the CRM to **enforce** the hand-off rather than trust it to memory or Slack messages.
## When NOT to use this
- Solo-founder or single-operator sales motion — there's no one to specialize into roles yet. Build the scoring/SLA plumbing first; add pod structure when headcount justifies it.
- Deal cycle is so short and transactional that role handoffs would add latency without adding quality (a $9 self-serve purchase doesn't need a prospector).
---
## Step-by-step build
1. **Classify every lead source using Seeds / Nets / Spears.** Seeds = organic/referral/word-of-mouth. Nets = inbound/marketing-generated. Spears = targeted outbound. This classification determines which role owns the lead first.
2. **Define the three roles as CRM-enforced stages, not job titles.** Prospector (sources + qualifies — owns Inquiry→SAL), Closer (owns SAL→Won), Farmer/Account-manager (owns Won→expansion/renewal). Map real 19Keys agents: prospector = outreach-operator, closer = funnel-closer, farmer = owner-action-chaser.
3. **Write the hand-off contract for each transition.** Not "the prospector passes it to the closer" — state the objective, system-verified criteria: e.g., "SAL requires `lifecycle_stage = SAL`, a populated next-step field, and a scheduled call on the calendar. The CRM will not let a contact enter the closer's queue without all three."
4. **Build the pipeline-coverage math per role.** Ross's arithmetic: pipeline needed = quota ÷ average deal size ÷ win rate, at some target coverage ratio (commonly 3–4x). State the actual ratio you're targeting, not the generic number.
5. **Instrument per-role conversion.** Prospector: qualification rate. Closer: win rate. Farmer: expansion/renewal rate. Each role gets its own dashboard number, not one blended metric that hides which stage is actually broken.
6. **Write the conflict rule.** What happens if a contact re-enters at a different source (a Spear who then also fills out a form, becoming a Net)? State the dedupe/ownership rule explicitly — this is a required field in the `journey-map` rubric.
## Worked mini-example
Lead classified as a Spear (targeted outbound list). Owned by outreach-operator through the enrichment waterfall and initial qualification. Hand-off contract to funnel-closer: `lifecycle_stage = SQO`, next-step field populated with a specific date, and a discovery call already completed and logged. Once those three system-verified conditions are true, the contact appears in funnel-closer's queue automatically — no Slack ping required. Post-close, ownership transfers to owner-action-chaser for onboarding and renewal tracking. Pipeline math: quota of $50K/month, average deal $2K, win rate 20% → need $250K in qualified pipeline; target coverage 3.5x means outreach-operator must keep $875K of Spear+Net pipeline moving at all times.
## How the verify gate applies
The `journey-map` linter requires named, ordered stages with entry criteria per stage, a trigger inventory (every automation referenced by spec-file), conflict rules for contacts targeted by two automations/roles at once, and per-transition instrumentation. A hand-off stated as prose ("prospector hands off when ready") without the three-condition system check fails — it isn't machine-evaluable. The companion `pipeline-audit` fails if the coverage ratio and per-role conversion numbers aren't tagged with a source query.