Skip to main content

What does success look like — in terms anyone can explain without reading the spec?

OUTCOME MAP: SALES DEV AGENT
════════════════════════════════════════════════════════════

  DESIRED OUTCOME
  Agent produces 10+ scored prospects/week so the
  founder converts 5 into discovery calls/month
       │
       ├── Contributing Factors
       │   ├── CRM, email, LinkedIn adapters production-ready
       │   ├── 9 algorithms exist (sales-forecasting, explore-exploit, SPCL, etc.)
       │   ├── Sneakers Media independently identified same problem from buy side
       │   ├── 5 test clients already identified (HRV, Kitchen Studio, Asset Factory, BizCover, MAS)
       │   └── NZ small market = quality over volume = our advantage
       │
       ├── Obstacles
       │   ├── LinkedIn app approval delay (days-weeks) → Register immediately, mock fallback
       │   ├── AI outreach feels spammy → Human reviews every message, SPCL gate
       │   ├── Construction ICP may not respond to cold → Test solar EPC in parallel
       │   ├── Agent build time > manual outreach → Start manual immediately, agent in parallel
       │   └── 0 paying customers, 28 days to kill date → Compress to essentials
       │
       ├── Investigations (answer before committing)
       │   ├── Does the construction/solar ICP respond to cold outreach? [Owner: Wik, Track 3]
       │   ├── What CPL does Sneakers accept for pre-qualified vs unqualified? [Owner: Wik + Eddy]
       │   └── Can SPCL gate reliably distinguish "worth reading" from "sounds like AI"? [Owner: Eng, S3]
       │
       ├── Success Measures (binary)
       │   ├── ≥5 discovery calls booked in first month → YES / NO
       │   ├── ≥15% reply rate on agent-drafted outreach → YES / NO
       │   ├── ≥20 scored contacts in CRM pipeline → YES / NO
       │   └── ≥1 pilot signed with Sneakers Media client → YES / NO
       │
       ├── Roles (RACI)
       │   ├── Accountable: Wik (go/no-go on every outreach message)
       │   ├── Responsible: Sales Dev Agent (research, score, draft, sequence)
       │   ├── Consulted: Eddy Whatt / Sneakers Media (buy-side validation)
       │   └── Informed: Engineering team (commissioning state)
       │
       └── Next Actions (momentum within 48 hours)
           ├── Register LinkedIn Developer app (Sprint -1, day 1)
           ├── Set 3 Resend env vars in Vercel (Sprint -1, day 1)
           └── Add first 20 prospects to CRM (Track 3, immediate)

════════════════════════════════════════════════════════════

Gate​

Before moving to Value Stream Map:

  • Desired outcome is crisp: "Agent produces 10+ scored prospects/week → founder converts 5 into calls/month" — YES
  • Contributing factors have evidence: each references existing infrastructure or named people
  • Obstacles have mitigations assigned: each has a parallel strategy
  • Investigation questions have owners: Wik (Track 3), Wik+Eddy (pricing), Engineering (S3)
  • Success measures are binary: all four are YES/NO with thresholds
  • RACI is complete: one Accountable per decision domain
  • Next actions are assigned: three tasks with owners and timing

Context​