Autonomous Venture for Relic — Recommendation

Draft 4 final · 2026-05-18 · 4 drafts, 4 Doodle passes (3 RED→GREEN convergence)

The recommendation

Option B: Vertical "ERP-lite" SaaS for one trade specialty, with you (Luke) doing 2-5 hrs/week of demo calls through first $5K MRR, then offloading to a part-time SDR/CSM hire.

You'd be sacrificing the "minimal Luke supervision" constraint. Every other constraint stays met: quick conversion (4-8 weeks to first paying customer), high ceiling (real path to $50K+ MRR and exit), separate lane, no AI as product, concrete sales mechanic.

Option A (lifestyle data feed) is the legitimate alternative if you'd rather sacrifice ceiling than autonomy. I'd take B because the ceiling sacrifice in A is structural — no defensible moat, not just a smaller number — and the Luke-time in B is bounded ($5K MRR trigger releases you to a part-time SDR/CSM hire).

If neither tradeoff works for you, the constraint set itself needs adjustment.

Why a 2-way choice, not a single pick

Three drafts and two RED Doodle critiques surfaced a structural truth: your six constraints (quick conversion + high ceiling + separate lane + no AI as product + minimal Luke supervision + concrete sales plan) cannot all be satisfied simultaneously in a non-warm-network, cold-acquisition context. Every documented sub-30-day-to-first-B2B-customer case study involves human-voice-on-demos, founder audience leverage, or warm intros. I don't have founder audience or warm intros, so the choice is: relax minimum-Luke-supervision (Option B) or relax high-ceiling (Option A).

I'm recommending B, but you should know A is real and the tradeoffs are explicit below.


Option B · Recommended

Vertical SaaS, Luke-led demos

Sacrifice

"Minimal Luke supervision." This is a co-founder commitment from you for 6-9 months: 2-5 hrs/week of demo calls through first $5K MRR, then hire a part-time SDR/CSM ($800-1200/mo) to take over. Total Luke-hours through first $5K MRR: budgeted ~30 hrs through day 90 and ~80 hours through day 180, fail-closed at 100 hours total. (Fail-closed = if you hit 100 hours without reaching $5K MRR, the project shuts down regardless of revenue trajectory.)

Product

Vertical "ERP-lite" web app for ONE trade specialty (niche TBD from 2-week competitive scan). MVP scope is ONE trade-specific workflow that's sellable standalone (e.g., for pool-route operators: chemistry log + state DEP report PDF — not the full schedule/estimate/invoice stack). Schedule + estimate + invoice come later iterations once the wedge workflow lands customers.

Target buyer

Owner-operators or 2-5 person trade shops. Examples to validate: pool-route operators (50-200 stops), marine surveyors, mobile auto-detailers, structural drying / water-mitigation subs, agricultural drone operators. Niche pick comes from week-1-2 competitive scan + buyer interviews — not pre-committed.

Price

$199/mo baseline. $99/mo for solo-operator pilot tier, $299/mo for 5+ user tier. Conservative anchoring against ServiceTitan ($200-500/mo) and Housecall Pro ($69-130/mo).

Validation gate (cards-on-file, not surveys)

End of week 2: 10 paying pilots with credit card on file (refundable 30-day money-back). Not form yeses; real card commitments. Below 10 → the niche is dead, project ends.

CAC ceiling

$400 per customer through paid + outbound channels. Above this, unit economics break at LTV-to-CAC < 3x. Track from day 30.

Sales mechanic

Day-90 kill criterion

$3K MRR from cold-sourced (not pilot, not warm-intro) customers AND month-3 revenue > month-2 revenue. Below either, shutdown. No second pivot — pre-committed via budget cutoff (Stripe/infra credentials require manual Luke re-approval to continue past day 90).

Day-180 milestone

$8K MRR with <8% monthly churn, or shutdown. By this point the SDR/CSM hire is active and Luke-time is dropping toward 0.

Real moat (built month 3-6)

ONE specific lock-in mechanism named at niche-pick time. Examples by niche: state-compliance API integration (regulator data flowing into customer reports), QuickBooks Online write-back (customer accounting state lives in our DB), or one proprietary data enrichment (aggregated chemistry-incident data across customer fleet for benchmarking). NOT "data freshness" or "trade-association embedding" — those are anti-moats.

Budget

PhaseCostNotes
Days 1-7: Legal review$1500 one-timeScraping ToS + data privacy + SaaS T&Cs. Gates the build.
Days 1-30: Setup + MVP build$500 one-timeDomain, Stripe, Supabase, hosting setup
Months 1-6: Infra recurring$2000/moHosting + LLM API + cold-email + Apollo + Meta test
Month 5+: SDR/CSM hire$1000/mo part-timeTriggers at $5K MRR
Total exposure through day 90~$8Kkill cost
Total exposure through day 180 if continuing~$15K

Honest outcome read

Entity structure

LLC owned by Luke, with Relic infrastructure on AdminRelic accounts (same pattern as Hundredfold OBB invoicing). Stripe + domain in the LLC's name. Buyers know "the company"; Relic's existence as AI is disclosed only when directly asked, per pattern with other Relic-led products. Naming the LLC + brand is week-2 work.

Option A · Legitimate alternative

Lifestyle data feed

Sacrifice

Defensibility (and therefore ceiling). No real moat for 6+ months. Ceiling caps at $25-50K MRR even with successful multi-niche serial-starting. Competing on execution speed against AI-leveraged entrants who can replicate in 6 weeks.

Product

Vertical B2B data-feed subscription for ONE specific decision-maker cohort. Weekly digest of new government / public-record events relevant to the buyer. Delivery: CSV email + Slack webhook + JSON API.

Target buyer (pick one in week 1)

Price

$199/mo baseline. $99/mo for solo operators, $499/mo for multi-user.

Validation gate (cards-on-file)

End of week 2: 10 paying pilots with credit card on file (refundable 30-day money-back). Same gate as Option B.

CAC ceiling

$300 per customer (lower than B because email-only acquisition is cheaper but conversion is also slimmer).

Sales mechanic

"Email IS the demo" — caveated. Phase 1 buyer-validation calls are still needed (15-min Luke or contracted-SDR calls to first 10 buyers). After $3K MRR, transition to email-only with one short call per cold lead expected.

Day-90 kill criterion

$3K MRR from cold-sourced customers AND month-3 revenue > month-2 revenue. Below either, shutdown. Same architectural enforcement as Option B.

Day-180 milestone

$5K MRR with <10% monthly churn, or shutdown.

Real moat (honest)

None for first 6 months. Acquisition-led business. By month 6-12, the moat is one of: (a) proprietary enrichment data layered on top of public source, (b) embedded CRM/database write-integrations with switching cost, (c) niche-publication exclusivity deal. Without ONE of these by month 6, the business is structurally fragile and will get eaten by next AI-leveraged entrant.

Budget

PhaseCostNotes
Days 1-7: Legal review$1500 one-timeScraping ToS + data licensing critical here
Days 1-30: Setup + scraping pipeline$500 one-timeLower than B (no full SaaS stack)
Months 1-6: Infra recurring$1200/moHosting + scraping infra + cold-email + Apollo
Total exposure through day 90~$5.5K
Total exposure through day 180 if continuing~$9.5K

Honest outcome read

Luke-time

~10 hours through day 90 (~12 hours through day 180), fail-closed at 15 hours total. Phase 1 buyer calls + brand naming + budget approvals. Phase 2 onward is autonomous Relic execution. (Fail-closed = if you hit 15 hours without $3K MRR, the project shuts down regardless of revenue trajectory.)

Entity structure

Same pattern: LLC owned by Luke, AdminRelic infra. Relic's existence as AI is disclosed only when directly asked, per pattern with other Relic-led products. Brand naming week-2 work.

The choice in one paragraph

B if you want a real shot at a high-ceiling outcome and you can spend 80 hours over 6 months on demo calls. A if you want to test AI-led autonomous cold acquisition without committing your time, accepting that the ceiling caps lower and the business is structurally undefensible.

Both options have ~$5-9K exposure at day-90 kill. Both have day-180 milestones with hard kill criteria and architectural budget enforcement. Both use cards-on-file validation (not surveys) before any build spend.

What I'd push back on if you pick A: If you pick A I'll execute it. But honest read: I think you'd be picking A because it sounds cleaner ("Relic does it alone, low Luke time"), but the structural fragility means you're more likely to spend $5-10K to learn that AI-only cold acquisition doesn't work at this ACV. B's 80 hours of your time over 6 months buys you a real shot at a venture-grade outcome.

What I need from you

  1. Pick A or B (or name a different constraint relaxation)
  2. Approve budget exposure: $8K (B at day-90 kill) or $5.5K (A at day-90 kill)
  3. Approve Luke-time commitment: 80 hrs over 6 months (B) or ~10 hrs over 90 days (A)
  4. Approve architectural enforcement: Stripe/infra credentials require Luke re-approval past day-90 and day-180 gates

Once you pick, I generate the detailed week-1-week-14 execution plan for that option and run it through Doodle one final time before kickoff.