StillChase — TTB cockpit for solo craft distilleries

1. One-liner

Photo-and-voice cockpit that turns daily distillery readings into reconciled TTB monthly reports for solo craft distillers.

2. Trend signal — why now?

US craft distilling is now a 2,282-active-DSP industry (ACSA’s 2025 census, methodology cleaned to drop dormant permits) doing $7.6B in retail value but margin-thin. The category got permanent federal excise relief in 2024 ($2.70/proof gallon for the first 100,000 PG), so micro-DSPs that survived the 2022-24 shakeout are staying open — but TTB enforcement against record-keeping sloppiness is sharper than ever.

The pain is well-documented:

• 78% of TTB violations against craft distilleries are record-keeping errors (FX5 / industry reporting), and 73% of craft distilleries get hit with at least one violation in their first three years.
• Average violation cost > $25K when you add fines, legal, and remediation.
• A real example: Mountain View Distillery (Colorado) — $12K fine for three months of missing daily operating reports plus mismatched transfer records, plus quarterly TTB reviews for two years.
• Failure-to-file penalty: $1,000 per occurrence + up to 1 year imprisonment under 27 CFR Part 19.
• Existing software (Whiskey Systems, FIVE x 5, DISTILL x 5, OnBatch, TTB Tamer) starts at $169/mo + per-bottle royalty (Whiskey Systems) or annual contracts with 20% surcharge for month-to-month. ADI Forum has open complaint threads about price increases and “on it a few months and not liking it.”

What changed in the last 12 months that makes a fresh entrant viable:

• Vision models cheap enough ($0.001/image) to read a hydrometer/thermometer photo and auto-apply the TTB Gauging Manual Table 1 (temperature correction → true proof) in the field, on a phone, in 8 seconds.
• Whisper-class voice transcription so a distiller in PPE can say “240 wine gallons in still A, hearts cut at 9:14, reading 158.4 at 72.6” while pumping, and the cockpit fills the production form without typing.
• TTB e-filing is now mandatory for tax returns ≥ $5M; the rest still file paper or PDF, but APIs and Pay.gov integration are mature enough to push semimonthly 5000.24 returns programmatically.

Provenance:
  - Signal 1: 78% of TTB violations are record-keeping; 73% of new craft distilleries get one in 3 years; avg $25K cost — https://barrelclarity.com/blog/ttb-compliance-mistakes/ — 2026-05-10 — and https://fx5.com/18-mistakes-ttb-audit/
  - Signal 2: ACSA 2024 Craft Spirits Data Project shows 2,282 active US craft DSPs; CBMA permanent reduced rate ($2.70/PG) keeps small DSPs viable post-2024 reform — https://americancraftspirits.org/wp-content/uploads/2017/02/2024-Craft-Spirits-Data-Project-Final.pdf — 2026-05-10 — and https://distilling.com/resources/government-affairs/temporary-tax-parity/
  - Signal 3: Whiskey Systems pricing $169/mo + per-bottle royalty + 20% surcharge for month-to-month; ADI forum thread on price increases — https://whiskeysystems.com/features-pricing/ — 2026-05-10 — and https://adiforums.com/topic/13536-whiskey-systems-price-increases/
  Category: Underserved niche (with workflow-automation overlap — daily logbook → TTB form is the workflow, and the existing tools price out the bottom 60% of the market)

3. The opportunity

The bottom 60% of the US craft distillery market — solo and 2-person DSPs producing under 10,000 proof gallons/year — is stuck in a bad place. They can’t afford Whiskey Systems / FIVE x 5 at $169/mo plus per-bottle royalties on top of $12K/yr ERP overhead. So they keep daily records on paper logbooks, Excel sheets, and handwritten dipstick notes (confirmed pattern on ADI Forum). When TTB or the state ABC shows up for an audit, the paper trail doesn’t reconcile to physical inventory, and they eat the $12-25K hit.

The incumbents are real ERPs that try to be cost-accounting + production tracking + TTB filing all at once. They were designed for distilleries with a controller. A solo distiller who’s also the master distiller, sales rep, and forklift driver doesn’t need ERP — they need the daily operating record to write itself, and the monthly TTB report to fall out the back of it.

That’s the wedge: cheap, voice-first, photo-first daily logbook that auto-generates the four TTB monthly forms (5110.40 Production, 5110.11 Storage, 5110.28 Processing, 5000.24 Excise) and flags inventory mismatches before the auditor finds them.

4. Target market

• Primary customer: Owner-operator and 2-person US craft distilleries (DSPs) producing < 10,000 proof gallons/year. ~1,400-1,700 of the 2,282 active DSPs fall in this band. Annual revenue $150K-$1.5M.
• Why they buy: They already had one TTB letter, or they know the next one will end them. They’ve been told by their state ACSA chapter / industry consultant that “you have to log daily” but they’re a one-person shop with no time to do data entry after a 12-hour distilling shift. Existing software is too expensive and too complex.
• Rough TAM: ~1,500 US solo/duo DSPs at $99-149/mo = $1.8M-$2.7M ARR US-only ceiling. With micro-cideries, micro-wineries (5120 series TTB forms — same shape), and small breweries (BROP), TAM expands to ~7,000 producers. Add Canada, UK, Australia, NZ versions of the same regulator-by-volume model and the ceiling moves to $5-8M ARR.
• Why now for them: TTB resumed normal audit cadence post-COVID; states are tightening their own beverage tax filings (e.g., new direct-to-consumer rules in 2024-2025 across NY, CT, KY); CBMA permanent reduced rate means they have to prove their proof-gallon numbers to claim it.

5. Product sketch (MVP)

• Voice batch logger — distiller speaks “Mash bill 70 corn 20 rye 10 malted barley, 800 lbs, 250 wine gallons mash water, fermenter F2, started 9:42” while moving; the system parses and writes a fermentation record entry with timestamp.
• Photo-of-hydrometer proofing — phone snap of the hydrometer + thermometer in the proofing parrot; vision model reads both, applies TTB Gauging Manual Table 1 temperature correction, returns true proof with audit-trail (the photo itself is the source document).
• Photo-of-dipstick / scale — for tank gauging and cooperage receipts; vision reads the scale, cross-checks the wine-gallon figure against the prior reading and the still output.
• Auto-fill 4 TTB monthly forms — TTB F 5110.40 (Production), 5110.11 (Storage), 5110.28 (Processing), and 5000.24 (semimonthly Excise Tax Return). Pre-flight check: every line ties back to a daily record source, and inventory at month-end = inventory at month-start + receipts - transfers - removals - losses.
• Reconciliation flag — the moment a daily record creates an inventory mismatch (e.g., 215 PG produced but only 208 PG into bond), the cockpit highlights it and prompts the distiller to either correct or annotate-the-loss before it shows up on the TTB report.
• State ABC filings shim — Kentucky, Tennessee, Texas, New York, California first; same daily records → state monthly beverage tax form.
• 3-year archive — every photo, voice clip, and form export held in the format TTB expects on audit (3-year retention is the federal rule).
• “Audit Ready” weekly digest — Monday-morning email: what’s missing, what’s mismatched, what TTB or the state would flag if they walked in today.

6. AI angle — what’s load-bearing

Without AI, this is a clipboard. With AI it collapses 90 minutes of daily paperwork into 8 minutes of pointing the phone at things and talking. Specifically:

• Vision model does the gauging — reading hydrometer + thermometer photos to within TTB’s 0.1°F / 0.05 proof tolerance, applying the gauging tables, and producing an audit-quality record (with the photo as evidence). This is the manual step distillers most often get wrong, and the most-fined error category.
• Voice transcription + structured extraction — the distiller doesn’t type. They talk. The model parses ingredients, batch IDs, timestamps, weights, and routes them into the right CFR Part 19 record category.
• Reconciliation reasoner — when a daily record creates a mass-balance violation, the model explains what’s wrong in distiller-language and proposes the fix (operational loss vs. theft vs. data-entry typo). This is the work an outside compliance consultant charges $150/hr for.

Strip out the AI and you have Excel. The whole product is the AI doing the data entry.

7. Localization angle (if any)

US-first by design — TTB’s regulatory regime is the wedge. No international v1.

After 18 months, the same shape ports to:

• Canada — CRA Excise Act distillery licensees (~250 distilleries), monthly K50B excise return.
• UK — HMRC W5 Distillers’ Return for ~600 distilleries, plus AWRS.
• Australia — ATO Excise Returns for ~700 distilleries (PSP licence).

All three regulate by volume of pure alcohol and require similar daily records. Same vision model + same reconciliation engine + different output forms. That’s the path from $2-3M US-only ARR to $5-8M Anglosphere.

8. Business model — path to $1M–$5M ARR

• Pricing tier 1 (Solo): $99/mo — single user, up to 5,000 PG/year, all 4 TTB forms, voice + photo capture, 3-year archive.
• Pricing tier 2 (Duo): $149/mo — 2-3 users, up to 15,000 PG/year, plus state ABC filings for one state, plus barrel/cooperage tracking.
• Pricing tier 3 (Tasting Room): $249/mo — adds DTC shipping compliance (per-state direct-shipping permits), club fulfillment receipts, COLA prep.
• No per-bottle royalty. Flat. This is the wedge against Whiskey Systems.
• ACV: $1,500-$2,000 average across mix.
• Path to $1M ARR: 600 customers × $140 ACV/mo × 12 = $1.0M. ~600 / 1,500 addressable solo DSPs = 40% penetration of bottom segment within 24 months.
• Path to $5M ARR: Add Anglosphere + micro-cideries/wineries (TTB 5120) + small breweries (BROP). ~3,500 customers × $120 ACV/mo × 12 = $5.0M.
• Expansion path: state ABC filings (each state = $25/mo add-on); DTC shipping compliance ($50/mo add-on); COLA assistant; trademark watch.

9. Go-to-market wedge — first 100 customers

• Day 1 channel: ADI + ACSA forums. The American Distilling Institute forum (adiforums.com) has an active “Distillery Software Pros & Cons” thread with named complaints about Whiskey Systems pricing. Post a side-by-side pricing comparison and a free “TTB audit pre-flight” tool. Expect ~30 leads in 60 days.
• Day 2 channel: 2026 ACSA Convention (Sacramento, April 17-18). Cheap booth or sponsor the compliance track. ACSA has the high-intent buyer who already pays for memberships and consultants. Expect 50-80 demo bookings, 15-25 conversions.
• Day 3 channel: TTB compliance consultants as resellers. ~30 named compliance consultants (NiekampLaw, ZahnLaw, BarrelClarity, FX5 itself) make money cleaning up violations. Offer 25% recurring rev share for referrals. They send the distillers who can’t afford Whiskey Systems.
• Day 4 channel: TTB Pay.gov submission scrapes — public DSP permit database. TTB publishes the active DSP permit list (Permit Online). Filter to DSPs that registered in last 24 months (the 73%-violation cohort). ~1,200 emails. Personalized cold outreach: “You opened your DSP in 2024. 73% of distilleries in your cohort already have a record-keeping violation. Here’s the 7 most common.” Expect 4-6% conversion to demo.
• Day 5 channel: Master Distiller workshop sponsorships. Moonshine University, Distillery University, Siebel Institute. The new operators graduate, then need software within 90 days of permit issuance. Sponsor the compliance module.

If first-90-days target is 100 paying customers at $99-149/mo, that’s $11K MRR — credible based on the ADI thread + ACSA convention math.

10. Build complexity — justification

Medium. Off-the-shelf tech everywhere — phone PWA, vision API, Whisper, Pay.gov form templates. The hard parts are: (1) every TTB form line item correctly mapped to a daily-record source so audits reconcile, (2) the reconciliation engine that detects mass-balance breaks across a fermentation→distillation→bottling chain, (3) state ABC form variants. None of this is novel R&D — it’s domain-specific care. Solo founder with a distillery advisor can ship v1 (one TTB form + voice + photo) in 12 weeks; full 4-form + reconciliation in 5 months.

11. Gating checklist

Gate	Pass?	Note
Legal in target market	✅	TTB record-keeping is a customer obligation, not a regulated activity for the SaaS vendor; we’re a tool, not a permit-holder
Ethical — no harm / dark patterns	✅	Helping operators stay legal
Market exists (evidence above)	✅	2,282 active DSPs, 4 named software competitors, ADI forum complaint threads
1–5 person team can build this	✅	Standard PWA + vision/voice APIs + form generators
Launchable with <$50K / ₹40L	✅	$20K covers AI API credits + LLC + ACSA booth

All five clear.

12. Feasibility score

Axis	Weight	Score	Notes
Problem intensity	20	17/20	$25K avg violation cost; $1,000+ federal failure-to-file penalty per occurrence; permit suspension is an existential threat. Hair-on-fire for the 73% cohort.
Demand evidence	15	12/15	Multiple incumbents charging real money, public ADI forum complaints about pricing, ACSA census of 2,282 active DSPs. Direct user voice on solo-DSP pain still mostly secondhand — a few primary interviews would push this to 14.
Build feasibility	15	11/15	Vision + voice are commodity; TTB form mapping is meticulous but documented; state ABC variants add weeks. Solo can do it in 5 months.
Distribution clarity	15	12/15	ADI forum + ACSA convention + TTB permit list + compliance-consultant resellers — four named, costed channels.
Revenue mechanics	15	11/15	$99-149/mo benchmarks under Whiskey Systems’ floor; ACV believable; $1M-$5M ARR math holds. Churn risk if a distillery closes — need to model.
Time to first revenue	10	9/10	Pre-sell at ACSA April 2026; ship beta-paid in 8-10 weeks.
Defensibility	10	8/10	Workflow lock-in (3 years of source documents in your system) + TTB form-mapping correctness as quality moat + state ABC integrations compound over time. Not patent-grade, but real.
Total	100	80/100

13. Qualitative modifiers

Founder-fit tags

domain-expertise-required · technical-heavy

The founder needs to have either run a DSP or worked with one closely (former TTB compliance consultant, ex-Whiskey Systems engineer, distillery controller). Without that, the form-mapping fidelity that customers actually pay for won’t be there. Pair with a competent full-stack engineer for the build.

Key assumptions to validate (3–5)

1. Assumption: Solo / 2-person DSPs will pay $99-149/mo flat for a daily-records-first tool, not full ERP. How to test: 25 in-person interviews at the April 2026 ACSA Convention with a printed mock landing page; ask “$99/mo, no per-bottle royalty, voice + photo, auto TTB forms — would you switch?” Need 40%+ “yes, today” rate.
2. Assumption: Vision model can read a hydrometer + thermometer in real distillery lighting (steam, cluttered parrot, gloved hands) within TTB’s 0.05-proof tolerance. How to test: 200-image bench test against ground-truth manual gauging, in three real distilleries; need 95%+ within tolerance.
3. Assumption: Compliance consultants will refer business at 25% revshare. How to test: Direct outreach to 10 named consultants (NiekampLaw, ZahnLaw, BarrelClarity); need 4+ to commit to refer in writing.
4. Assumption: TTB Pay.gov / Permit Online endpoints are stable enough to programmatically push 5000.24 returns. How to test: TTB sandbox + 5 hand-tested submissions across permit types in first 60 days. If TTB blocks API push, fall back to PDF-export-and-manual-upload (still saves 90% of the work).
5. Assumption: Churn ≤ 18% annual once installed (workflow lock-in works). How to test: Track first 50 customers’ retention at month 12; abort if churn > 25% — means the product is replaceable.

Risk flags

1. Regulator dependency: TTB could change form formats or push its own e-filing portal that captures daily records directly. Mitigation: focus on state ABC + Anglosphere expansion early so no single regulator is >50% of value.
2. Incumbent retaliation: Whiskey Systems / FIVE x 5 cuts price to match $99/mo. Mitigation: ship faster, own the solo segment, and build state-ABC depth they don’t bother with.
3. Cohort timing: The 2024-2025 craft-distillery shakeout removed ~800 DSPs. If consolidation continues, the addressable solo market shrinks. Mitigation: the per-customer LTV stays the same; just expand to micro-cideries and small breweries to widen the funnel.
4. Vision model brittleness: Hydrometer-reading accuracy in real distillery lighting could be the make-or-break edge case. Mitigation: photo always saved as source-document, manual override always available, and a “human in the loop” review tier for the first 90 days of customer life.

14. Structured verdict

Score:                  80/100
Verdict:                STRONG GO
Confidence:             Medium
Best-fit builder:       Ex-distillery operator or TTB compliance consultant paired with one full-stack engineer comfortable with vision/voice APIs
Time to revenue:        8–12 weeks (pre-sell at ACSA April 2026, beta-paid by end of June 2026)
Capital to launch:      $20–35K (AI credits, ACSA booth, 6-month runway for one engineer)
Top 3 assumptions to validate first:
  1. Hydrometer + thermometer photo gauging within TTB tolerance — 200-image bench test in 3 real distilleries
  2. 40%+ "yes, today" intent at $99-149/mo flat from 25 ACSA Convention interviews
  3. 4+ compliance consultants signing letters of intent on 25% revshare referral
Kill criteria:
  - Abandon if vision gauging accuracy < 90% within tolerance after 200-image test
  - Abandon if < 25% intent rate at $99-149/mo from ACSA in-person interviews
  - Abandon if Whiskey Systems or FIVE x 5 launches a "$99 starter" tier within 90 days of validation sprint

15. Next step — 1-week validation sprint

• Day 1–2: Build the bench-test rig — 200 photos of hydrometer/thermometer pairs across 3 friendly distilleries, ground-truth proof from manual gauging. Run them through Claude vision / GPT-4o vision / Gemini vision and measure tolerance. Pick the cheapest model that hits 95%+.
• Day 3: Cold-call / DM 30 solo DSPs from the ADI forum. Ask: “What’s your current TTB record-keeping setup and what does it cost you per month?” Score for matching pattern (paper / Excel / Whiskey Systems / FIVE x 5).
• Day 4: Mock up landing page — “$99/mo, voice + photo, auto-fills your 5110.40 / 5110.11 / 5110.28 / 5000.24, no per-bottle royalty, no annual contract.” Buy 200 clicks of LinkedIn ads targeting ACSA / ADI members; measure email-capture conversion.
• Day 5: Decide. Go condition: vision tolerance ≥ 95%, ≥ 8 of 30 solo DSPs say “yes I’d pay $99 today,” landing page email-capture ≥ 12%. Anything less = back to Stage 2 with a different cluster.

Falsifiable. Numbers, not vibes.