Skip to main content

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

OUTCOME MAP: SALES CRM & RFP
════════════════════════════════════════════════════════════

DESIRED OUTCOME
Construction sales teams win >30% of bids (up from 22%)
using a CRM that learns from every bid they write

├── Contributing Factors
│ ├── 41 features mapped, 20 live (49%), 6 built, 3 dormant algorithms
│ ├── Domain model covers: Contacts, Companies, Ventures, Deals, Properties
│ ├── RFP auto-fill reporting 70% fill rate — needs first real data
│ ├── Construction vertical gap: no competitor offers native RFP+CRM+Property
│ ├── 23 real contacts, 3 ventures, 1 deal — internal dogfood in progress
│ └── Agency lib has 3 dormant algos: Sales Forecasting, Compound Rate, RFP Detection

├── Obstacles
│ ├── Auth REGRESSION: infinite redirect (PostgreSQL 22P02) — app inaccessible
│ ├── Answer library empty — auto-fill flywheel can't spin without seed data
│ ├── 3 algorithms dormant — built in agency lib, not wired to UI
│ ├── Company entity missing — can't answer "show me everything with ABC Corp"
│ ├── Activity logging empty — timeline UI exists, 0 entries across entire system
│ └── Convex permissions: rfp_answer read blocked, ventures query fails

├── Investigations (answer before committing)
│ ├── Will construction teams adopt a vertical CRM over configured Salesforce? [Owner: Wik, Pilot]
│ ├── Does 70% auto-fill hold with real RFP data? [Owner: Eng, Sprint 1]
│ ├── Will sales teams create deals (not just contacts)? [Owner: Wik, Activation metric]
│ └── Can answer library seed fast enough to prove compound value? [Owner: Eng, Sprint 1]

├── Success Measures (binary)
│ ├── Win rate >30% within 6 months → YES / NO
│ ├── First answer approved to library within 7 days of pilot start → YES / NO
│ ├── Pilot user creates a deal within 48 hours of onboarding → YES / NO
│ └── Answer library has 10+ approved answers with confidence scores → YES / NO

├── Roles (RACI)
│ ├── Accountable: Wik (product direction, pilot recruitment, kill decisions)
│ ├── Responsible: Engineering team (algorithm wiring, auth fix, entity work)
│ ├── Consulted: Construction/solar EPC sales teams (validation, friction feedback)
│ └── Informed: Sneakers Media (RFP workflow patterns applicable to agency lead gen)

└── Next Actions (momentum within 48 hours)
├── Fix Identity & Access auth (unblocks everything) — Eng
├── Fix Convex rfp_answer permissions (unblocks 3 pages) — Eng
└── Wire Sales Forecasting algo to forecast page (Sprint 0, day 1) — Eng

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

Gate

Before moving to Value Stream Map:

  • Desired outcome is crisp: "Win >30% of bids using a CRM that learns from every bid" — YES
  • Contributing factors have evidence: 41 features audited, 49% live, named algorithms and entities
  • Obstacles have specificity: auth regression has error code (22P02), permissions name the table
  • Investigation questions have owners: Wik (adoption), Engineering (auto-fill, seeding)
  • Success measures are binary: all four are YES/NO with thresholds
  • RACI is complete: one Accountable, clear Responsible, named Consulted
  • Next actions are assigned: three tasks with owners

Context