Prepared 2026-07-22. Figures verified as of this date unless marked otherwise. All numeric targets are editable assumptions, not commitments — tune each at the noted checkpoint. Day 1 = [WEEK_1_START_DATE].
Revision note (red-team findings C2/M2, applied 2026-07-22): all targets below are re-based to the corrected financial model (file 22 §6: base ≈ 5–6 funded loans year 1, first funded loan month 4 at the earliest). The earlier targets (8 scenarios / 2 apps by day 30; 2–3 closings by day 90) described a business the model's own funnel cannot produce. Closings are excluded from gating before month 4 — a first funded loan before month 3 would be luck, not plan (file 22 §3). Files 22, 23, and this file now describe the same business.
Phase 1 — Days 1–30: "Become Useful" (Foundation + Pilot Launch)
Theme: Earn the right to be in agents' phones. Product mastery + the Scenario Desk working in public, on a 30-agent pilot list. Volume of useful touches, not volume of touches.
Objectives
Every first-72-hours item closed or on a dated open list (file 10).
Scenario Desk live with the documented SLA (ack ≤1 business hr; written, guideline-checked brief ≤4 business hrs, Mon–Fri — the v0 SLA; "human-reviewed"/Saturday are v1 claims gated on written verification, file 09 §2.3/§2.5) — and the SLA hit on every request received.
Full 14-day pilot sequence (email → manual call → short voicemail → LinkedIn → useful asset → in-person where appropriate) run against all 30 pilot agents.
Five confirmed product lanes mastered to "explain to an agent in 60 seconds" standard; Ready's actual IL menu verified (kills all [UNVERIFIED] flags from the manager meeting).
First-Home Lab: naming approved by Ready (or renamed), first two cornerstone pieces drafted and in compliance review — DPA stack (IHDA Access + Cook County DPA + FHLB DPP) and Cook-vs-Will tax-payment math.
Key actions
Weeks 1–2: ramp calendar (file 12) weeks 1–2 — learning-heavy. Warm-first entry order (red-team M3): the first 6–8 pilot slots are warm — Crosstown Hybrid/non-Flex agents via the family connection plus family-network introductions to agents at other brokerages (conflict-hygiene rules in file 15 §1 apply) — before any cold-corridor wave starts. Week 2 runs at half-dose (waves of 5, not 10); week 3 scales to full waves only if week-2 debrief energy ratings average ≥6/10.
Approval-starved contingency track (red-team C5): if compliance approvals stall, weeks 1–2 switch to the no-approval-needed holding pattern defined in file 10 (drills, sample-brief portfolio, pilot-list research, facts library, community presence) with day-7/14/21 escalations. Day 21 with no approved outreach copy = employer-viability signal → file 26 §Plan B, not just a delay.
Daily: 20-min ChatGPT Voice role-play; 3×/wk written scenario drills; every real inbound scenario logged and answered inside SLA.
Pilot-sequence completion (touches on time, for agents entered so far)
90%
Day-30 gate (go/adjust)
≥3 scenario requests AND ≥10 meaningful conversations → proceed to Phase 2 as planned. (This is base case per file 22 — it is a pass, not a disappointment.)
<3 scenario requests → the offer may not be landing, but the sample is tiny at day 30: re-interview 3 friendly agents and revise the ask; the formal offer-redesign stop rule fires at day 45 if <6 total scenario requests despite ≥25 completed sequences (file 09 D2 item 11 — the single source of truth for that rule).
<7 meaningful conversations → diagnose energy vs. message first (scorecard rule 1: volume is not the fix for a message problem). Review the week-2 half-dose debriefs: if energy was red, retune the dose (rule 10); if energy was fine, audit call windows and scripts with the manager. Adding a sprint/wk is the last lever, never the first.
KPI checkpoint: full scorecard review in the Day-30 after-action; every assumption above gets re-set for Phase 2.
Phase 2 — Days 31–60: "Convert Usefulness into Pipeline" (Deepen + Content Launch)
Theme: Fewer, deeper agent relationships; the First-Home Lab goes public; first closings-in-motion.
Objectives
Tier the pilot list: A (engaged, ≥2 real interactions) / B (warm) / C (silent). A-tier gets biweekly bespoke value; C-tier gets replaced from adjacent offices.
First-Home Lab launched: cornerstone DPA-stack piece + tax-math piece published (compliance-approved, NMLS IDs per 1050.940); condo-triage piece framed post-change from the start — the Limited Review sunset hits/hit 2026-08-03 at plan start, so the asset is "the rules just changed — here's the new triage" (files 09 §3.3, 18), not a race already lost.
First live files worked cleanly: buffer blocks in the calendar absorb them without dropping the SLA.
First buyer-side education event: one small-room First-Home workshop (library/chamber venue, ≤15 seats — small-room suits Mitchell; no big-stage events).
IHDA path status from Ready (U.S. Bank MRBP question) resolved to a written yes/no/timeline; positioning updated accordingly.
Key actions
Steady-state calendar (file 13) replaces ramp calendar at week 5.
Outreach shifts ~40% new / 60% follow-up-and-deepen; open-house Saturdays with A-tier agents (own branded flyer only — never fund the agent's kit; RESPA).
Content batching: one 2-hr batch block/wk produces 1 cornerstone or 2 short derivatives; everything through compliance before posting.
Model F-lite drills continue at 2–3/wk as they have since week 1 (file 09 §4 — drills are cheap and run from the start; what waits for day 60 is scaling to a named content pillar); additionally collect 3 real self-employed/bank-statement scenario patterns from chamber contacts — no content spend yet.
Numeric targets (all [assumption — reset at Day-30 gate]; re-based to file 22 base case)
Metric
Day-60 cumulative
Meaningful agent conversations
28–32
A-tier agents
5–6
Scenario requests
8–10
Applications
3
Preapprovals
1–2
First-Home Lab pieces live
3
Workshop attendees
8–12
Day-60 gate — Model F-lite go/no-go
GO if: bank-statement product confirmed on Ready's IL menu, ≥3 organic self-employed scenarios have appeared, and core KPIs are ≥80% of target. F-lite = one content franchise ("Investor/Self-Employed File Lab") + 2–3 scenario drills/wk. Nothing more.
NO-GO if any of the three fails → park to day 90+, reallocate the hours to A-tier agent depth.
Also at day 60: pull-through review of every application; response-time audit; pricing/turn-time reality check vs what Mitchell has been telling agents.
Theme: Make the machine run without heroics. Referrals from closed/near-closed files, documented SOPs, and a defensible decision about what to scale.
Objectives
First closings done cleanly; after-action on each file within 48 hrs of CTC; testimonial/review ask built into the close (compliance-approved wording).
SOP library complete: scenario-brief SOP, application handoff SOP, AI SOP / Never-Automate List v2 (the Fannie LL-2026-04 governance angle formalized as an agent-facing differentiator).
Pilot list evolved into a stable book: 8–10 A-tier agents producing repeat scenario flow; second 30-agent wave launched only if A-tier is saturated with attention.
First-Home Lab cadence locked: 2 pieces/mo + 1 small-room workshop/mo; measure by scenario requests and workshop-to-preapproval conversion, not views.
Phase-2 geography decision prepped (Frankfort–Mokena–New Lenox vs Homewood–Flossmoor–Matteson) with data, for day-90+ execution only.
Numeric targets (all [assumption — reset at Day-60 gate]; re-based to file 22 base case)
Metric
Day-90 cumulative
Meaningful agent conversations
50–55
A-tier agents
8–10
Scenario requests
14–16
Applications
5–6
Preapprovals
3–4
Closings (funded or CTC)
0–1 (informational — NOT a gate metric; first funded loan is month 4 base case)
Agent-referred borrowers
3–4
Day-90 gate — leading indicators only (closings excluded from gating before month 4)
≥8 A-tier agents + ≥3 agents who have sent a second scenario (repeat senders) + SLA record ≥95% + ≥5 applications → scale: second agent wave + Phase-2 geography pilot + (if day-60 GO) F-lite continues.
Below that → do not expand geography; diagnose the funnel stage with the worst conversion and spend month 4 fixing only that. UIUC alumni idea stays deferred unless everything else is green.
A funded loan by day 90 is a bonus, not a criterion — gating on closings this early would gate on luck (file 22 §3).
Weekly: Friday pipeline review (files by stage, stuck >5 days flagged) + Saturday 45-min after-action (what worked / what to stop / next week's single focus).
Days 30/60/90: gate reviews above, with manager, decisions written down.