Mo's Files Digest — /Users/mo/Downloads/military/

Produced by Explore agents on June 4, 2026 as input to the master audit. Captures everything Mo has documented across 20 files of iterative audit/strategy work.

SECTION 1 — Audit-flavored files (digested)

1. LG_Military_DEFINITIVE_AUDIT.md (CANONICAL narrative audit)

~195 lines / ~2,400 words. Window Mar 1 – Jun 3, 2026. Self-labeled "Definitive."

Top findings: - $2.45M spend → $3.35M revenue → ~$891K net, 1.36x blended ROAS (internal QW dashboard) - Auto 1.38x / 27.3% margin vs Home 1.28x / 22.1% — Auto is the better business - Scaled 4.5x April→May with no ROAS degradation - 98% of Purchase event price data has formatting issues — Meta Events Manager flag; all events send the same flat price - fbc/fbclid truncation on 15 ad sets, $67,174 spend, 27% of CompleteRegistration events affected - Auto BAU pacing gap: $19,800/day budget vs ~$11,400/day actual - Odometer_05/15 running at $34.63 CPL on ~$178K — highest-cost big spender - "Mil FB Customer Match 4.12.21 (Full List)" — audience list reads as a 2021 list, stale

Conclusions: Two tracks — Track 1 (cost levers for CPL goal posts), Track 2 (fix value pipe for real optimization). GP1 (15% CPL reduction) high confidence from cost levers alone; GP3 (25%) only achievable the right way (value up, not junk in).

Gaps flagged: What value is wired into Purchase event (flat vs real payout)? Why is Auto BAU pacing only ~$11.4K/day? What happened May 23 (0.90x loss day)? Can real QW per-lead payout be sent server-side?

2. LG_Military_Integrated_Audit.md (post-media-buyer-strategy version)

~103 lines / ~1,500 words.

Top findings: - SAC (Special Ad Category) removes the standard CPL levers — no lookalikes, no age/geo targeting, no exclusions. This invalidates the entire "shift budget to 25–44" and "build value-seeded LALs" plans from FULL_AUDIT.md v3. - FunnelFlux offline conversion upload fires every ~15 minutes (single-fire postback) — value tracking exists in principle, but the browser Purchase event's flat value may be diluting the algorithm's signal - Campaign structure (BAU/ALC/MaxVal trio) is intentional, not cannibalization — media buyer tested more and saw performance dip from overlap - Auto > Home is a carrier story: Progressive bids strong on auto; Allstate (home) is reactive - The "auto-updating" customer-match list was last confirmed refreshed May 2025 (13+ months ago, per the Master Task List/Learnings PDF) - Two narrow Custom ad sets (IDs 120240808864440498, 120227142278180498) delivering to ~1,000-person audiences — essentially broken

Conclusions: Measure on RPL−CPL spread, not CPL alone — the single most important strategic ask. Drop the lookalike plan. Creative is the primary CPL lever remaining under SAC.

3. LG_Military_Optimization_Audit.md (LEADS campaigns optimization trace)

~72 lines / ~900 words.

Top findings: - Custom conversion total (~33,277 / 30d) reconciles with internal QW dashboard (34,158 May leads) and CompleteRegistration (31.3K) — all within ~2.5% - Native Meta Lead event is never populated ("Not available") — leads flow through custom conversions "click to quotes" — this is not a malfunction - Max Value campaigns pay $36.24 CPL vs $21.91 on BAU — a ~65% premium — to optimize for value against a flat value signal (98% malformed Purchase price data) - A paused SALES campaign optimizes toward fb_pixel_purchase (standard Purchase event) — different/older setup, currently inactive - CAPI alongside Pixel reported 17.5% more Purchase conversions in the last 7 days

4. LG_Military_Master_Audit_TaskList_Learnings.pdf (5-page June 2026 deck — CANONICAL client-facing)

Top findings (identical economics, new framing): - Meta Opportunity Score: 99/100 — no structural fixes to make; leverage is in measurement, creative, and the success metric - Key Learning #4: Max Value pays a ~65% CPL premium to chase a value signal that doesn't exist - Key Learning #5: SAC removes standard CPL levers — bypassing SAC is "Evasion violation that damages account health" - Customer-match list last refreshed May 2025 (13+ months ago, not auto-updating) - fbc truncation: 15 ad sets, $67,174 spend, 27% of CompleteRegistration (confirmed) - CAPI reported 17.5% more conversions (last 7d) — signal health is real - Dataset C (Military.com's content pixel) CAPI Gateway 0% set up; Add a Signals Gateway Pixel impact estimated at ~23% cost/result

Conclusions: 5 workstreams — W1 Measurement Plumbing (highest leverage), W2 Creative (primary SAC lever), W3 Efficiency (learning-phase safe), W4 Audiences (SAC-compliant), W5 Strategic/Metric (most important ask: reframe to RPL−CPL spread).

5. FULL_AUDIT.md (v3 ROAS-verified — superseded but contains the inversions)

Key inversions vs v1: - Veteran-VO: 9.37% CTR → 0.90 ROAS, loses money (v1 said "revive it") - Instagram Stories: 0.34% CTR → 1.42 ROAS — top placement (v1 said "cut it") - Desktop: 3.6x CPC → 1.45 ROAS — best device (v1 said "exclude") - 65+: cheapest CPL → 1.18 ROAS — lowest of all ages (v1 said "shift budget here") - Rewarded video: "fraud" → real leads, $22.65 CPL, break-even at 0.96 ROAS - TYFYS video: "CTR mirage" → 1.57 ROAS — best high-spend ad - Lower-Your-Auto_04/28: $421K spend, 1.45 ROAS, +$191K net — the workhorse winner - Odometer_05/15: $222K spend, 1.21 ROAS — net positive but least efficient big-spend ad - Meta reported value ($1.62M) vs internal dashboard ($3.35M) — Meta captures ~half of real revenue

6. lg_military_full_audit.md (v2 — superseded)

Blended: ~$842K spend → ~33,366 leads → ~$25.23 CPL. Odometer at $38.92 CPL (later revised to $34.63). Drive-On_05/30: $5,137 at $18.48 CPL — best auto CPL, newer creative.

7. lg_military_audit.md (v1 — origin doc, mostly self-corrected)

$841,952.78 spend / 30d. Account frequency 6.45. Estimated recoverable value: $140–200K/month conservative, $335–565K/month full execution (based on CTR assumptions later invalidated). Most v1 claims were inverted by v2/v3.

8. Master prompts (procedural artifacts, not audits)

full_audit_master_prompt.md, lg_military_full_audit_master_prompt.md, ad_library_research_prompt.md — execution prompts for prior audit pulls. Contain the 8 epistemic rules Mo developed (never trust CTR alone, tag every finding, custom conversions don't break by placement, etc.) — worth preserving.

Audit Synthesis

Canonical now: DEFINITIVE_AUDIT (narrative) + Master_Audit_TaskList_Learnings.pdf (deck). Everything else superseded but historically useful.

Consolidated picture of what Mo found: A healthy $850K/month lead engine with three compounding problems, not the six "critical leaks" the v1 audit identified:

1. Broken conversion value (the root cause): 98% of Purchase price data is malformed/flat. Max Value campaigns can't optimize for value. Meta estimates +7% ROAS from fixing this alone.
2. fbc/fbclid truncation on 15 ad sets ($67,174 spend, 27% of CompleteRegistration events): degrades attribution and optimization signal.
3. Odometer_05/15 concentration: ~$178K/30d at $34.63 CPL vs Lower-Your-Auto at ~$22 CPL. Reallocation is the largest single pure cost lever today.

The SAC constraint removes the standard CPL levers (no demo reallocation, no exclusion audiences, no lookalikes). What remains: creative (UGC test batch — primary lever), budget/pacing efficiency, value-signal quality, and the strategic ask to reframe success metric from CPL to RPL−CPL spread.

SECTION 2 — Plan / strategy / pitch files (digested)

1. LG_Military_30Day_Action_Plan.md

~1,400 words. Task-level execution plan with workstream B (audiences), C (tracking), D (cost): - B1: Define Engaged audience from FB/IG page engagers - B2: Replace stale 2021 customer-match list with 180-day sold-leads export from Lauren - B3: Build suppression/exclusion audiences - B4: Delete ~30 dead placeholder lookalikes; rebuild on sold-leads + engagement seeds - C1: Fix Purchase event value — "#1 priority" — real QW payout server-side - C2: Fix fbc/Click ID truncation - D1: Reallocate Odometer_05/15 budget → Lower-Your-Auto_04/28 - Close Auto BAU pacing gap

2. LG_Military_Master_Plan.md

5-workstream strategic plan. Thesis: Meta Opportunity Score 99/100 → structural wins don't exist → leverage in measurement plumbing, creative, and success metric reframe.

Most important strategic ask (W5-T1): Reframe success metric to RPL − CPL spread (or ROAS), or pair CPL with RPL/acceptance floor.

HELOC expansion thesis (W5-T2) introduced here.

SAC mechanics extensively documented: 15-mile radius minimum, no age/geo targeting, LAL eligibility contested for financial category.

3. LG_Military_Audit_June_Priorities.md

Most complete economics table (verified Meta $25.23 vs internal $24.91, sub-2% reconciliation). 4.5x April→May spend scale-up without ROAS degradation as evidence of headroom. Conversion-value fix called #1 June priority (longest lead time). 7% ROAS improvement estimate from Meta modeling on value-fix alone.

4. LG_Military_Deck_Source.md

16-slide deck. Frames GTW as "sharpening, not fixing." RPL−CPL reframe needs leadership signoff, not just media buyer.

5. military_meeting_package.md

Pre-meeting walk-in pitch. Three problems / three threads (Performance/Compliance/Partnerships). Stage 1 $20K diagnostic + cross-portfolio. Stage 2 $75K pilot. Stage 3 $60K/mo + 5% net-new revenue + 10% partnerships. Stage 4 $1.5–4M Cross-Valnet.

6–8. military_pitch_v2/v3.md, military_proposal.md

Earlier pitch iterations. Stage 3 $50–100K/month. Thread C (recruitment partnerships): "10 partners × $100K = $1M/year at 60–80% margin." military_proposal.md is the formal $75K 90-day pilot doc.

9. claude_code_master_prompt.md

Multi-agent orchestration prompt for original meeting prep. Contains the IP clause language (GTW retains War Room; client owns campaign data + creative assets). 8-vector risk assessment including "Hassan needs Rony signoff," "Valnet pays Net-60+ in practice."

Strategy Synthesis

Canonical strategy doc now: /Users/mo/Desktop/military.com/deliverables/june_2026/06_june_strategy_master.md. None of the Downloads strategy files reach that level of post-access operational specificity.

What's NEW in Downloads files not in June 2026 deliverables:

1. HELOC expansion thesis — not in June plan at all
2. RPL − CPL spread metric reframe — Mo's "most important strategic ask"; June plan operates within CPL goal-post framework without proposing change
3. Thread C (recruitment partnerships) — present in pitch files; absent from June 2026 deliverables scope
4. Customer-list lookalike eligibility test (W4-T3) — specific in-product test for the contested SAC/LAL question; not in June 2026 weekly roadmap
5. Value-based lookalike seeded from payout-weighted customer list — not in June 2026 deliverables
6. IP clause language — GTW retains War Room; client owns campaign data
7. SAC mechanics documentation — more precise in Downloads files than in June 2026 deliverables

Tensions / contradictions with renegotiated deal terms

Cross-references to the existing deliverables structure

• LG_Military_DEFINITIVE_AUDIT.md + LG_Military_Master_Audit_TaskList_Learnings.pdf → connect directly to /Users/mo/Desktop/military.com/deliverables/05_audit_visual_spec.md (the reliability framework / two-track plan source)
• LG_Military_Audit_June_Priorities.md → connects to /Users/mo/Desktop/military.com/deliverables/june_2026/04_baseline_vetting_framework.md (economics table)
• LG_Military_30Day_Action_Plan.md → maps onto Track 1 (cost levers) in /Users/mo/Desktop/military.com/deliverables/june_2026/01_first_week_playbook.md and Track 2 (value pipe) in 03_tracking_capi_rebuild.md
• LG_Military_Master_Plan.md W5 (metric reframe) → NOT YET reflected in any June 2026 deliverable — surface in master audit
• military_meeting_package.md, pitch v2/v3, military_proposal.md → fully superseded by the renegotiated deal; historical reference only

End of digest.