Final decision: Blueprint
RBSClear — California RBS Server Roster Completeness Pack Engine
Done-for-you inspection-ready reconciliation of active alcohol servers vs. ABC RBS portal roster status, expiration calendar, and portal add/remove action list — so California on-premises licensees stop gambling a 10-day alcohol privilege suspension on a spreadsheet.
Executive summary
Decision: Blueprint. Business: RBSClear. Sell: a per-license / per-cycle RBS Roster Completeness Pack — not training software, not a chatbot, not TTB/COLA for producers.
California ABC requires on-premises licensees to ensure alcohol servers (and their managers) hold valid RBS certifications within 60 days of hire and to maintain certification records for inspection (BPC §§25682–25684). Regulations set a 10-day suspension guideline for first-offense RBS violations (4 CCR §173). ~48,690 on-sale licenses (ABC FY2024–25) and ~56,000 on-premises-privilege licensees sit under this regime. Training vendors sell courses and dashboards; almost nobody sells a done-for-you pack that reconciles payroll + schedule roles against the ABC portal roster and produces a release-ready gap scorecard and portal action list.
~48,690
CA on-sale ABC licenses (FY24–25)
~56,000
On-premises privilege licensees (ABC)
10-day
First-offense RBS suspension guideline
$249–$899
Per Completeness Pack (flat)
Six-gate score: 26/30. Evidence threshold cleared. Not a duplicate of beverage-alcohol-market-ready (producer COLA/TTB) or AllergenPack / TipPoolClear / FOG ManifestClear.
Thesis
High-turnover California restaurants, bars, and multi-outlet hotels already buy RBS training. What they still fail at — and what ABC actually audits for — is the operating system: who is allowed to pour tonight, which Server IDs sit on which license roster, who expires in 90 days, and whether managers who supervise servers are certified. RBSClear turns that reconciliation into a productized, AI-accelerated service with a human compliance specialist as the trust interface, priced per pack / per month — never hourly.
Discovery rationale
This run steered away from saturated healthcare filing engines and from OFF-LIMITS hospitality clones (BEOClear, AttritionClear, ManifestClear, GPLClear, SlipLienClear). Research across retainage, make-ready backcharges, pool logs, STR TOT, and CA RBS roster ops showed the strongest CODE + narrow wedge + non-SaaS whitespace in California RBS server-roster completeness: statutory duty is clear, penalty is sharp, buyer already spends on training, and incumbents sell courses/dashboards rather than DFY portal reconciliation packs.
Candidate comparison
| Candidate | Score /100 | Verdict | Why |
| CA RBS Roster Completeness Pack (RBSClear) | 86 | Winner | Verified statute + penalty; existing training budget; DFY whitespace vs dashboards; narrow MVP |
| Specialty-sub Retainage Release Completeness Pack | 78 | Runner-up | Billd 56-day DSO strong; CloseoutDesk adjacency + Siteline SaaS gravity |
| Hotel/Apartment Pool Inspection Log Completeness Pack | 68 | Reject | Jurisdiction fragmentation; physical-service gravity; weaker centralized WTP |
| Multifamily Make-Ready Vendor Backcharge Evidence Pack | 64 | Reject | DepositPack adjacency; dispute/UPL risk |
| CA STR TOT Gap Remittance Desk | 61 | Reject | Avalara/Monterra SaaS; tax-practice risk; SB 346 is facilitator-side |
CODE validation
- Consumer/buyer trend: Post-2022 RBS mandate is now operationally mature; hotels with multiple license numbers (47/48/70/portable bar) struggle with roster sync; high F&B turnover continuously restarts the 60-day clock.
- Opportunity: The underserved problem is not “get training” — it is “prove every person who can pour is certified, on the correct license roster, and not expired,” including managers.
- Demand: ABC FAQ + portal tooling; training providers publish employer how-tos; hotel F&B content markets multi-license audits; licensees already pay $8–$15+/server for courses plus paid training time.
- Economic sizing: Beachhead: independent + small-group CA Type 41/47/48 operators (~tens of thousands of licenses). If 5,000 packs/year at $399 blended average ≈ $2.0M revenue; 2% of ~48.7k on-sale licenses buying a $99–$249/mo monitor ≈ multi-million ARR path. Ranges uncertain; not all on-sale licenses are equal RBS intensity.
Rubric scorecard
| Dimension (1–5) | Score | Note |
| Low trust burden | 4 | Already outsourced training; Designee role exists in portal |
| Low task-level judgment | 4 | Mostly matching + deadlines; exceptions for role classification |
| High intelligence threshold | 4 | Cross-system synthesis: payroll/POS/schedules/portal exports |
| Regulation as moat | 5 | ABC discipline + consecutive suspension rules |
| No physical labor | 5 | Fully remote documentation/ops |
| Sam Altman test | 4 | Better extraction/matching as models improve; ops still compound |
| Outcome pricing | 5 | Per pack / per license / monitor — never hourly |
| Gross margin potential | 5 | Document workflow; minimal licensed review |
| Buyer urgency | 4 | Suspension / OIC cost / liability exposure |
| Competitive whitespace | 4 | Training SaaS ≠ DFY completeness pack |
| Novelty vs manifest | 5 | No prior RBS roster pack; producer COLA is different |
| AI capability fit | 5 | OCR, entity resolution, calendar, checklist drafting |
| Active demand evidence | 4 | Strong regulatory + vendor how-to; weaker public complaint volume |
| Budget/competitor proof | 4 | Training spend + manager time already allocated |
| Waitlist/lead-magnet | 5 | Free Roster Gap Scan maps cleanly |
| Narrow MVP clarity | 5 | One pack, one license, one cycle |
| Distribution clarity | 4 | LinkedIn GM/HR, CalRest, training partners |
| Licensing feasibility | 4 | Ops docs; not legal advice; Designee auth required |
| Operational repeatability | 5 | SOP-ready reconciliation |
| Speed to first revenue | 5 | Manual portal + sheets day one |
Weighted takeaway: Clears evidence threshold. Main residual risk is portal access friction and confusion between training SaaS and DFY service.
Target buyer
- ICP: California independent or 2–12 unit restaurant/bar groups and boutique hotels holding Type 41/47/48 (and related on-premises) licenses, 15–120 FOH staff who can serve alcohol, no dedicated compliance coordinator.
- Economic buyer: Owner / GM / multi-unit operator.
- Champion: AGM, HR/people lead, F&B director, or office manager who already owns training enrollments.
- Trigger: New hire cohort, banquet season, ABC inquiry/warning, multi-license hotel audit, certification mass-expiration wave, insurance questionnaire, or ownership change.
Jobs-to-be-Done
- When I hire a bartender or banquet server, I need them certified inside 60 days without me chasing portal IDs.
- When ABC or my insurer asks, I need a defensible roster proof per license number — tonight.
- When someone works across restaurant + banquet + portable bar licenses, I need them on every relevant roster.
- When certifications approach the 3-year expiry, I need recertification queued before the license is exposed.
Painful problem
RBS compliance fails as an ops problem, not a knowledge problem. Spreadsheets drift from 7shifts/HotSchedules roles; terminated employees linger on portal rosters; managers who supervise servers are missed; multi-license hotels maintain separate rosters that do not auto-sync. First-offense guideline is a 10-day license suspension (4 CCR §173), which can compound consecutively with other alcohol-service violations. Alcohol often drives 20–30% of casual-dining revenue — so even a short suspension (or Offer-in-Compromise cash-out) is materially expensive versus a few hundred dollars for a completeness pack.
The outcome we sell
“Inspection-ready RBS Roster Completeness Pack” for a named ABC license number: gap-scored reconciliation of eligible servers/managers vs portal roster + certification status + expiration calendar + ordered portal add/remove checklist + SB/Lab-cost tracker for training fees and paid training time — delivered done-for-you. Optional monthly Monitor keeps the pack current.
First one-feature MVP wedge
| Element | Definition |
| ICP | CA Type 41/47 independent restaurant or bar, 20–80 FOH, one primary license |
| Trigger | Owner/GM cannot produce a clean roster proof in <15 minutes |
| Pain | Unknown gaps between who serves and who is certified/rostered |
| One-feature MVP | Single-license Roster Completeness Pack |
| Input | Payroll export + current schedule role tags + ABC portal roster export/screenshots + Designee auth |
| Output | Scorecard + exception list + portal action checklist + 90-day expiry calendar + evidence binder |
| Human chokepoint | Compliance specialist confirms role→server classification and releases pack |
| Success metric | ≥95% of flagged gaps closed or explicitly owner-accepted within 10 business days; zero “surprise expired” in next 30 days |
| Next asks | Multi-license hotel packs; monthly Monitor; new-hire 60-day SLA; training-provider enrollment orchestration |
Evidence summary
- Verified BPC §§25680–25684: certification duty, 60-day hire window, recordkeeping, disciplinary (not criminal) liability.
- Verified 4 CCR §173: 10-day suspension guideline for BPC §25683 violations; aggravating factors include % uncertified and duration.
- Verified ABC reports ~48,690 on-sale licenses (FY2024–25) and states ~56,000 on-premises-privilege licensees.
- Verified NRA CA fact sheet: 86,779 restaurant locations; $220.1B sales (2024 data basis).
- Verified Training vendors document portal roster management, multi-license hotel pain, and employer dashboards.
- Inferred Material WTP at $249–$899/pack vs suspension / OIC economics.
- Unverified Industry blogs claiming SB 476 directly amends RBS employer-pay rules — primary SB 476 text amends food-handler HSC §113948; treat RBS employer-cost obligations via Lab §2802 / ABC practice, not as a verified SB 476 RBS amendment.
Claim table (Verified / Inferred / Unverified)
| Claim | Label | Confidence |
| Licensees must not employ alcohol servers without valid certification (post windows) | Verified | High |
| New servers have 60 calendar days from employment to certify | Verified | High |
| Certification valid 3 years | Verified | High |
| First-offense guideline = 10-day suspension (4 CCR 173) | Verified | High |
| ~48,690 CA on-sale licenses FY24–25 | Verified | High |
| ~56,000 on-premises privilege licensees | Verified | Med-High |
| Multi-license hotels must manage separate rosters; portal does not auto-sync | Verified | Med (vendor/primary FAQ consistent) |
| Buyers will pay $249–$899 for DFY pack | Inferred | Med |
| SB 476 itself creates RBS employer-pay duties | Unverified | Low — do not rely |
| Alcohol = 20–30% casual dining revenue in CA | Inferred | Med (industry stats) |
Market and demand evidence
Beachhead is California on-premises alcohol service compliance operations. ABC’s FY2024–25 state totals list 48,690 on-sale licenses (Type 41 alone >25k). ABC’s RBS page cites ~56,000 licensees with on-premises privileges. NRA’s 2025 CA fact sheet reports 86,779 restaurant locations and $220.1B sales. Demand signal is regulatory compulsion + continuous hire/turnover + multi-license complexity, evidenced by agency FAQs and vendor employer content — not a soft “AI interest” trend.
Active buyer conversations
- ABC FAQ language answering “How do I manage my RBS server roster?” and Designee invitations — proof licensees are expected to operate the portal continuously.
- Hotel F&B educational content emphasizing quarterly roster audits and multi-outlet sync failures.
- Training providers selling bulk enrollment / employer dashboards — buyers already shopping solutions adjacent to this problem.
- Insurance and ownership diligence routinely request training/compliance evidence (industry practice; treat as inferred for RBS specifically).
Competitive landscape
| Player | What they sell | Gap vs RBSClear |
| ABC RBS Portal | Official roster + alerts | DIY tool; no payroll reconciliation |
| ServingAlcohol / AAA / other providers | Training + employer dashboards | Course completion ≠ portal completeness pack |
| HRIS / 7shifts / HotSchedules | Scheduling/HR | No ABC roster truth / Server ID matching |
| Hospitality compliance consultants | Broad ABC/health consulting | Hourly, not productized pack unit |
| beverage-alcohol-market-ready (prior blueprint) | Producer COLA/TTB | Different buyer + workflow entirely |
Competitor and budget validation
Budget already exists: per-server training fees ($~8–$15), ABC registration fees, paid training hours, and manager time spent chasing certificates. RBSClear redirects a slice of that ops budget (and risk budget) into a DFY outcome. Winning argument vs dashboards: “We don’t ask your GM to become a portal power user — we return a signed completeness pack and action list.”
Pricing evidence and proposed pricing
- Free Roster Gap Scan (lead magnet): 1 license, lightweight scorecard.
- Roster Completeness Pack: $249–$499 single-license independent restaurant/bar; $499–$899 multi-outlet hotel license cluster (up to 3 related license numbers).
- Monitor: $99–$249/license/month for monthly delta packs + 60-day new-hire SLA tracking.
- New-hire surge add-on: $15–$25/head for bulk onboarding cohorts (orchestration, not legal advice).
- Never hourly. Never contingency on avoided fines.
Anchor: 10-day suspension or OIC cash-out commonly dwarfs pack price (industry guides cite hundreds–thousands per day economics; treat as inferred ranges).
Regulatory and compliance considerations
- Core: BPC §§25680–25686; 4 CCR Article 25 §§160–173.
- Licensee remains the duty holder; RBSClear is an operations vendor / License Administrator Designee assistant.
- Do not claim criminal exposure for servers (BPC §25684).
- Drink-spiking curriculum additions commence on/before Jan 1, 2027 (BPC §25681(f)) — future recert content change, not MVP blocker.
- Separately track food-handler SB 476 / HSC §113948 if upselling adjacent — do not conflate statutes in customer materials.
Licensing boundary
| Allowed | Not allowed / must not claim |
| Extract, match, classify, calendar, draft portal checklists, assemble evidence binders | Legal opinions on ABC defense strategy; representing licensee in ABC hearings |
| Operate as authorized Designee when customer grants portal access | Unauthorized practice of law; guaranteeing “no ABC action” |
| Coordinate enrollment with approved training providers | Issuing RBS certifications (only approved providers/ABC exam) |
| Track Lab §2802-style cost reimbursements as ops bookkeeping | Wage-hour legal advice; HR counsel substitute |
Disclaimers: documentation/ops service; licensee remains responsible for compliance; not legal advice; portal actions executed only under written authorization.
AI-native advantage
AI collapses the expensive middle: OCR of portal screenshots/exports, fuzzy name/Server-ID entity resolution across payroll and schedules, role tagging (server vs host vs manager-of-servers), exception ranking, and customer-ready scorecards. Humans only resolve role ambiguity and release. As models improve, match accuracy and multi-license synthesis improve — unit COGS falls without rewriting the offer.
Internal AI engine architecture (10 layers)
- Intake: secure upload + Designee auth + license number(s).
- Normalization: canonicalize employee identities, roles, hire dates.
- Retrieval/knowledge: BPC/CCR checklists, license-type RBS applicability notes, SOP templates.
- AI workbench: extraction, matching, gap narrative drafts.
- Deterministic rules: 60-day clock, 3-year expiry, manager-included rule, multi-license presence checks.
- Human chokepoint: specialist release on role classification + high-risk exceptions.
- QA: dual-check counts, spot audit 10% of matches, red-team “who can pour tonight?”
- Delivery: PDF/HTML pack + portal action CSV + calendar ICS.
- Learning loop: missed matches → alias dictionary / role rules.
- Model-portability: prompts + schemas model-agnostic; swap frontier models without changing customer outcome.
AI-vs-human operations pipeline
1. IntakeHuman collects auth + files
2. ExtractAI OCR/normalize
3. MatchAI + rules entity resolve
4. ScoreDeterministic gap gates
5. DraftAI pack + checklist
6. ReviewHuman specialist
7. QAChecklist audit
8. DeliverCustomer pack
9. Optional executeDesignee portal actions
10. LearnAlias/SOP update
Dynasty translation layer
- Buyer: CA on-premises owner/GM who fears ABC suspension more than training cost.
- Service: DFY completeness pack + optional monitor; AI inside, human face outside.
- Workflow: intake → reconcile → pack → portal actions → renew.
- Tooling: Sheets/Notion + LLM + secure drive day one; later light portal connectors.
- Sales: “We’ll tell you who can legally pour tonight — and fix the roster gaps.”
- Delivery: Manual Designee clicks OK for first 20 customers.
- Expansion: Multi-license hotel packs, franchise rollups, insurer-ready attestations, adjacent food-handler roster (separate statute).
Anti-duplication analysis
Checked full manifest.json (317 runs) + blueprint filenames. No RBS roster completeness pack. Nearest: beverage-alcohol-market-ready-compliance-engine (producer COLA/TTB — different ICP/workflow/outcome); TipPoolClear / AllergenPack / FOG ManifestClear (different regs). Not a CredentiAlert clone (healthcare credentialing). Differentiator: ABC portal roster truth × payroll/schedule reconciliation as DFY pack for on-premises licensees.
Anti-commoditization analysis
If future models make DIY matching easy, RBSClear still wins on (1) Designee execution + accountability, (2) multi-license hotel SOPs, (3) audit-ready binder + attestations insurers/buyers accept, (4) continuous Monitor SLA, (5) compounding alias dictionaries per brand. The product is the operating system and trust interface, not the chat prompt.
Service delivery workflow
- Sales → Gap Scan → paid Pack SOW.
- Collect Designee authorization + payroll + schedule + portal export.
- AI reconcile → exception queue.
- Specialist review → release pack.
- Customer (or Designee) executes portal adds/removes; training enrollments for gaps.
- Confirm closure; optional Monitor subscription.
Operations as product
SOPs for intake evidence, role taxonomy, match confidence thresholds, exception codes (MISSING_CERT, EXPIRED, NOT_ON_ROSTER, ON_WRONG_LICENSE, MANAGER_GAP, TERMINATED_STILL_ROSTERED), QA sampling, versioned pack templates, postmortems on any post-delivery ABC issue.
No-holes quality engine
- Completeness gates block release if hire-date nulls exceed threshold.
- Confidence scoring on name matches; low-confidence → human.
- “Who can pour tonight?” red-team: sample 5 scheduled FOH names must appear certified + rostered.
- Audit trail of source files hashes + reviewer sign-off.
What the human expert actually does
| Task | License? | Min @launch | Min @D90 | Automation path | Quality risk | Cannot automate |
| Collect auth / explain Designee | No | 25 | 12 | Templates | Access denial | Trust conversation |
| Role classification exceptions | No | 20 | 8 | Rules library | Missed manager | Edge job titles |
| Pack release sign-off | No | 15 | 6 | Checklist | False clear | Accountability |
| Portal execution (optional) | No (Designee) | 30 | 15 | RPA later | Wrong license | Portal MFA/edge UI |
| Customer coaching on gaps | No | 20 | 10 | Playbooks | Scope creep | Change mgmt |
Minimum viable offer
“RBSClear Completeness Pack — Single License” for $349 founding price: Gap Scan free → Pack in 5 business days → 30-day email support to close gaps. Upsell Monitor at $129/mo after first pack.
Fulfillment process (first 3 customers)
- Zoom intake; Designee invite; upload payroll CSV + schedule PDF + portal screenshots.
- Founder + AI notebook reconcile in Sheets.
- Deliver branded PDF pack + action checklist.
- Optional screenshare to execute portal updates together.
- Capture corrections into SOP v0.1.
Tools day one: Google Drive, Sheets, Claude/GPT, Calendly, Stripe, Notion. Automate matching later — not portal MFA.
Human-in-the-loop quality control
No pack ships without specialist sign-off. Auto-fail if >3% low-confidence matches unresolved. Monthly calibration set of gold-standard anonymized packs.
Nonlinear scaling and unit economics
| Metric | Launch | Day 90 | Year 1 target |
| Specialist minutes / pack | 90–120 | 45–60 | 25–40 |
| Automation % | 40% | 65% | 80% |
| COGS (model+tools+labor) | $70–$110 | $40–$65 | $25–$45 |
| Price | $349 | $399 | $399–$599 |
| Gross margin | ~65–75% | ~75–85% | 80%+ on Monitor mix |
| Throughput / specialist / day | 3–4 packs | 6–8 | 10–12 |
| Rework rate | <15% | <8% | <5% |
| CAC payback | ≤2 packs or 2 Monitor months | ≤1.5 | ≤1 |
Path to 50%+ GM is immediate; path to 80% is Monitor + alias compounding. Revenue/FTE scales via packs/day, not headcount-linear consulting.
Distribution proof table
| Channel | Why ICP reachable | First angle | Conv. assumption | Proof | Measure | Follow-up |
| LinkedIn GM/HR | Operators searchable by title+CA | Gap Scan offer | 2–4% to scan | Vendor how-to demand | Scan→pack % | 7-day cadence |
| CalRest / local chapters | Trade association density | Suspension math teardown | Warm leads | NRA/CA restaurant density | Event→scan | Workshop |
| Training-provider partners | Already enroll staff | White-label pack | Referral 10% | Competitor adjacency | Referral packs | Rev share |
| SEO/AEO | Query intent “RBS roster” | Checklist magnet | Long-cycle | ABC FAQ traffic proxies | Organic scans | Email nurture |
| Hotel F&B groups | Multi-license pain | Quarterly audit offer | Higher ACV | Hotel guides | Cluster packs | Monitor |
Sales and outreach plan
Lead with diagnosis, not demo: “Send payroll + last week’s schedule — we’ll return who is uncertified/not rostered in 48 hours.” Convert Gap Scan → Pack → Monitor. Offer founding 10 customers $249 packs + 60 days Monitor at $99.
Founder-led content plan
Teach the exact failure modes: manager gaps, multi-license banquet servers, terminated-still-rostered, 60-day clock math, consecutive suspension stacking, insurer questionnaires. Show redacted scorecards. Avoid generic “AI for restaurants.”
First 30 days of content
10 educational posts: (1) 60-day clock explainer, (2) managers need certs too, (3) multi-license sync myth, (4) expired-but-scheduled, (5) Designee setup, (6) consecutive suspensions, (7) Lab 2802 training-cost hygiene, (8) banquet seasonal hiring spike, (9) what ABC records duty means, (10) portal alerts aren’t reconciliation.
3 diagnostic teardowns: redacted Gap Scan examples for Type 41, Type 47, hotel 47+68.
2 lead-magnet angles: Roster Gap Checklist PDF; “Who Can Pour Tonight?” one-pager.
1 webinar: live Gap Scan for 3 volunteer licensees.
1 outbound template: personalized note citing their license type + hire season + offer free scan.
Lead magnet and waitlist plan
Free Roster Gap Scan → waitlist for Monitor → Pack purchase. Buyer receives scorecard before paying; trust from specificity; pain signal = # of critical gaps; sales-ready if ≥1 critical gap or multi-license complexity.
Warm GTM plan
Founder network of CA operators; training-provider affiliate intros; prior Gap Scan users; local hospitality meetups; accountant/bookkeeper referrals who already see payroll chaos.
Targeted outbound plan
List build: CA Type 41/47 independents via public license data + Google Maps. Message: diagnosis memo with 3 likely failure modes for their license type. CTA: 48-hour Gap Scan. Cap 30 personalized outs/day.
Answer-engine / search visibility plan
Publish canonical pages answering ABC FAQ intents: “How to manage RBS roster,” “RBS manager certification,” “multi-license hotel RBS,” “RBS 10-day suspension.” Structure with citations to LegInfo/ABC. Target AI-overview citation via clear definitions + tables.
Pilot design and early-demand-trap mitigation
- Pilot cap: 8 licensees (max 12 licenses total).
- Incentive: founding pricing + public case study optional.
- Success: ≥90% gap closure in 10 business days; NPS ≥40; specialist minutes trending down.
- Mitigate demand trap: refuse custom ABC hearing defense; park feature requests into monthly roadmap; do not hire ops bodies to hide SOP holes.
Early-access feedback flywheel
Every miss → exception code → SOP/prompt/rule update. Product feedback = recurring match errors. Custom work = unique legal strategy — out of scope. Corrections become retrieval notes and QA checks within 48 hours.
Build-before-scale checkpoints
- After 5 pilots: harden intake evidence list + completeness gates.
- After 10: harden SOPs, exception taxonomy, reviewer checklist, delivery templates.
- After 20: pause new logos until COGS, rework, escalation, cycle time measured for 2 weeks.
7-day / 30-day / 90-day launch plans
7-day: landing page, Gap Scan form, SOPs v0, 20 outbound, 3 content posts, Stripe.
30-day: 8 Gap Scans, 3 paid packs, 1 Monitor, webinar, partner conversation with 1 training provider.
90-day: 25 packs, 10 Monitor logos, hotel multi-license playbook, alias dictionary ≥500, measured COGS <$65/pack.
Metrics and KPIs
- Gap Scan→Pack conversion; Pack→Monitor conversion; cycle time; critical gaps/pack; minutes/pack; rework %; portal execution error rate; churn; referral rate; CAC; gross margin.
Risks and mitigations
Portal access friction, statute conflation (SB 476), SaaS training vendors bundling DFY later, role misclassification, ABC UI changes, seasonality, insurance overclaim risk — each mitigated via Designee auth SOPs, precise legal labeling, partner not compete-head-on on courses, human chokepoint, model-portable extractors, Monitor retainers, and careful marketing claims.
Exhaustive risk register
1. Portal access denied / MFA friction — L:Med I:High
Mitigation: written Designee auth kit; screenshare fallback; never block pack delivery on optional execution.
2. Misclassify host/busser as alcohol server — L:Med I:High
Mitigation: role taxonomy + human review on ambiguous titles.
3. Training SaaS launches DFY service — L:Med I:Med
Mitigation: speed, hotel multi-license specialty, insurer-ready binders, partner channel.
4. Customer confuses pack with legal immunity — L:Med I:High
Mitigation: disclaimers; no “guarantee no ABC action.”
5. Statute conflation SB 476 / RBS in marketing — L:Med I:Med
Mitigation: cite BPC/CCR primary; Lab 2802 for costs; legal review of copy.
6. ABC portal UI/API change breaks ops — L:Med I:Med
Mitigation: screenshot+export resilient intake; quarterly portal drills.
7. Seasonal banquet surge overwhelms capacity — L:High I:Med
Mitigation: pilot caps; surge pricing; waitlist.
8. False-clear pack ships — L:Low I:High
Mitigation: red-team sampling; dual count checks; release checklist.
9. Multi-state expansion dilutes focus — L:Med I:Med
Mitigation: CA-only until 50 Monitor logos.
10. Wage-hour claims if advising pay practices — L:Low I:High
Mitigation: ops tracking only; refer employment counsel.
11. Data privacy (employee PII) incident — L:Low I:High
Mitigation: DPA, minimal retention, encrypted vault, access logs.
12. Low public complaint volume → slow inbound — L:Med I:Med
Mitigation: outbound + partner + Gap Scan lead magnet; don’t rely on SEO alone.
What could kill this
ABC eliminates roster concept / auto-syncs payroll statewide; major POS+HRIS ships perfect ABC connector free; or we overclaim legal outcomes and lose trust. Also fatal: becoming a generic hospitality compliance consultancy on hourly billing.
Go/no-go reasoning
Go. Clears evidence threshold: clear buyer, painful specific problem, verified regulatory spend/labor, competitor/budget proof (training), narrow MVP, service-first delivery, 50%+ GM path, differentiation from producer COLA blueprint and training SaaS. No fatal disqualifier. Honest about SB 476 conflation risk.
Final recommendation
Build RBSClear. Ship Gap Scan this week, sell Completeness Packs to CA Type 41/47 independents, expand to multi-license hotels, productize Monitor. Keep licensing boundary tight. Do not pivot into TTB/COLA or broad ABC defense counsel.