SB StartupBasket
All ideas
76 /100 GO Medium complexity

SpeciesDossier — Lacey Act due-care dossier for importers

Chases your overseas factory for the wood species and harvest country your Lacey declaration legally cannot fake.

TradeTech SaaS US Compliance-drivenSMBAI-agentSolo-builderSupplier-collaboration
— views
Evaluation Scores
76/100

GO

Overall Score

16
Problem
11
Demand
12
Build
12
Distrib.
11
Revenue
8
Time
6
Defense

SpeciesDossier

1. One-liner

Chases your overseas factory for the wood species and harvest country your Lacey declaration legally cannot fake.

2. Trend signal — why now?

Three things converged in the last 21 months, and the third one lit the fuse.

Phase VII took the Lacey Act from “lumber people” to “everybody.” On 1 December 2024 APHIS turned on declaration requirements for 250 additional HTS codes — furniture, cork, sporting goods, footwear, handbags, hand tools, matches with natural wood stems, bamboo and rattan products, certain essential oils, industrial and medicinal plants. Phase VII covers “all remaining plant product HTS codes that are not 100-percent composite materials.” Roughly 450 HTS codes now require a declaration. A company importing desk chairs or handbags — who had never heard of PPQ Form 505 — became a Lacey filer overnight.

Then the paper escape hatch closed. Effective 1 January 2026, APHIS no longer accepts paper PPQ 505 / 505B forms. Everything goes through CBP’s Automated Commercial Environment (ACE) or APHIS’s Lacey Act Web Governance System (LAWGS). Paper is now permitted only during system outages, and only with prior approval from Program Lacey Staff. Guidance is blunt that unauthorised paper submissions after the deadline violate the Lacey Act. The importers who had been coping by mailing a form are now in a structured-data regime with validation.

And APHIS refused to slow down. When industry asked for a 12-month extension on a broad essential-oils code, APHIS declined outright — “a Lacey Act Import Declaration will be required on December 1, 2024” — noting four years had passed since the Phase VI delay and nearly two years of targeted outreach. The SBA Office of Advocacy had formally urged APHIS to clarify requirements and delay compliance dates to relieve burden on small entities. The relief did not come. The duty is firm, dated, and self-executing.

Now the part that makes this a product rather than a newsletter. The declaration requires the scientific name (genus and species) and the name of the country from which the plant was taken — country of harvest, not country of manufacture. APHIS created Special Use Designations (SUDs) as an escape hatch for species you genuinely cannot identify after due care: SPECIAL/COMPOSITE, SPECIAL/RECYCLED, SPECIAL/RECLAIMED, SPECIAL/PREAMENDMENT, SPECIAL/SPF. But:

  • Solid wood cannot use a composite SUD. For solid wood items you provide real scientific names or you don’t import.
  • There is no SUD for country of harvest. None. Where the exact country is unknown, the declaration “shall contain the name of each country from which the plant may have been taken” — you must enumerate every candidate country, which is itself a research task.
  • As one industry summary of the standard puts it, the declaration “has to be precise and not ball park,” and “‘I don’t know’ is also unacceptable.”

Meanwhile the factories don’t speak in genus and species. Trade guidance is explicit that “many factories in Vietnam and China identify wood by common trade names rather than by genus and species,” that furniture “may contain many species within a single finished product,” and that “transshipment may further obscure the origins of those species.”

So: a newly-in-scope population, a hard electronic filing regime, a field with no legal escape hatch, and a supply chain that structurally does not volunteer the answer.

Provenance:

3. The opportunity

The gap here is not “filing.” It’s getting the answer out of the factory, and proving you tried.

Look at who owns which piece today:

The customs broker files but does not know. Only the importer of record or their agent may file a declaration, and — critically — both parties “assume responsibility for the accuracy and truthfulness of the information.” The broker types what you send them. Brokers charge $150–$400 per formal entry in 2026, with entries carrying 5+ line items or agency checks at the $250–$400 end, plus PGA filing surcharges of roughly $15–$75 per agency and $3–$10 per line. Your broker will happily transmit “Acacia mangium / Vietnam” to ACE. They have no idea whether it’s true, and they will tell you so. The data problem is upstream of them and they do not solve it.

The enterprise platform sells to the wrong tier. Altana is the name that comes up for Lacey supply-chain intelligence — verifying material origins, genus/species and country of origin. Altana has raised $383.6M, sells to UPS, Maersk and CBP itself, and quotes enterprise custom pricing with no public price. A 12-person furniture importer bringing in 40 containers a year is not buying that. This is the vendor floor vs legal floor shape: the legal duty now reaches a band that sits entirely below every incumbent’s minimum.

The consultants sell audits, not answers. Double Helix runs supply-chain scoping, risk assessment, supplier tiering, on-site assessment and validation — helping importers “develop, improve and demonstrate quality due care.” Valuable, human-delivered, project-priced, and explicitly not PPQ 505 filing or routine supplier data collection. You cannot buy 400 SKUs’ worth of species data as a subscription.

The trade associations are hand-maintaining the gap. This is the tell. Alliance-International is distributing a PDF to shops and makers with a literal four-question script to ask your distributor: What is the scientific name of this species? What is the country of harvest? Did it come from a sustainable source? Do you have documentation to prove it? Their recommendation is a slogan — “NO DOCUMENTATION? NO SALE!” A dealer publicly describes “developing a complete database of material sources” by hand as part of their ordering process. NAMM has had to arrange free APHIS compliance training for luthiers and small shops. When trade bodies are shipping question scripts and members are hand-rolling species databases, the workflow is unbuilt.

And there is a documented statement of exactly who is excluded: a working luthier’s framing of the statute is that “the crux of the problem with the Lacey Act is that it is OK for the big boys, but not for the smaller business/luthier.”

The product is the supplier-chase loop plus the due-care record. Due care is the whole ballgame legally — and note that “the Lacey Act does not define nor mandate any requirements to constitute due care.” There is no spec. What survives scrutiny is a contemporaneous record showing you asked, what you asked, what came back, and what you concluded. This is my new-duty-with-no-guidance and duty-attaches-to-signature pattern: the statute regulates the diligence, nobody sells the diligence trail.

Concretely, SpeciesDossier:

  1. Reads the SKU and decides whether a declaration is even required. The de minimis exemption is a computable dual test that nobody productises: plant material ≤5% of the total weight of the individual product unit AND total plant material in the entry, at a single 10-digit HTS provision, not exceeding 2.9 kg. Both conditions, and it does not apply to protected species. Half the panic is people filing when they needn’t, or skipping when they must.
  2. Runs the chase. Emails the factory in its own language, asks the four questions in a structured form, parses whatever comes back — a WeChat screenshot, a scanned mill certificate, a PDF packing list, “it’s acacia” — and maps trade names to genus/species candidates, flagging where a trade name is ambiguous across genera.
  3. Adjudicates the escape hatch. Tells you whether this item legally qualifies for a SUD and which one, or whether it’s solid wood and you have no hatch. Where harvest country is unknown, it assembles the defensible list of candidate countries rather than letting someone type one guess.
  4. Emits the broker-ready line data and keeps the dossier. Your broker gets a clean per-line file. You keep a timestamped record of every ask, chase and conclusion — the thing a due-care defence is actually made of.

The AI is load-bearing at exactly the point that was uneconomic two years ago: turning unstructured, multilingual, badly-scanned supplier replies into a per-line species/origin assertion with a confidence level and a citation, at 400 SKUs instead of 4.

4. Target market

  • Primary customer: US importers of record in the newly-in-scope Phase VII categories — wood furniture, case goods, cabinetry, flooring accessories, musical instruments, bamboo/rattan goods, handbags with wood components, sporting goods, hand tools. Sweet spot: $3M–$60M annual import value, 5–80 employees, 50–800 active SKUs, sourcing from Vietnam, China, Indonesia, India or Malaysia. The buyer is whoever owns compliance — typically an ops/logistics manager, a compliance-titled person of one, or the owner. Secondary: customs brokers and licensed customs consultancies who want to stop chasing clients for species data by email, and who can resell this to their book.
  • Why they buy, in their words: the declaration “has to be precise and not ball park,” and “‘I don’t know’ is also unacceptable” — but the factory answers in trade names, the product has six species in it, and transshipment obscures origin. They are personally on the hook: importer and broker both assume responsibility for accuracy. And APHIS has been issuing letters of non-compliance for declaration errors since 2018 — misidentified species, unlikely country/species pairings, incomplete declarations, late filings — with repeated failure to correct potentially referred for investigation or enforcement.
  • Rough TAM reasoning: CBP manages over 300,000 active importer-of-record numbers (2016 Congressional testimony). US furniture imports alone topped $72 billion in 2024 with China, Vietnam and Mexico leading, and the US imported 197.54 million pieces of wooden furniture in FY2023-24. Phase VII’s ~250 new codes span furniture through footwear. I am not going to invent a filer count — APHIS does not publish one. But a defensible serviceable slice is 8,000–20,000 US importers who now file Lacey declarations, have real SKU counts, and sit below Altana’s floor. At $500/mo average, capturing 1% of the low end is ~$500K ARR.
  • Why now for them: Phase VII landed Dec 2024, paper died Jan 2026, and APHIS refused to extend. The next Phase VII additions come with at least 6 months’ Federal Register notice — meaning there is a known cadence of future scope expansion, each one a fresh cohort of first-time filers.

5. Product sketch (MVP)

  • Declaration-required check — upload or sync your SKU list with HTS codes and bills of materials; get a per-SKU verdict on whether a Lacey declaration is required, including the de minimis dual test (5% of unit weight AND ≤2.9 kg per 10-digit HTS provision in the entry) computed rather than guessed.
  • Supplier chase agent — sends each factory a structured request in its own language for scientific name, country of harvest, and supporting documentation; follows up on a schedule; escalates after N ignored asks. No portal for the supplier to log into — they reply by email or WhatsApp like they already do.
  • Reply parser — ingests scanned mill certificates, packing lists, chat screenshots and free-text replies; extracts species and origin claims; maps common trade names (“acacia”, “meranti”, “rubberwood”) to candidate genus/species and flags ambiguity across genera.
  • SUD adjudicator — per line, says whether a Special Use Designation legally applies and which one, or states plainly that this is solid wood and no SUD is available.
  • Unknown-origin builder — where harvest country can’t be pinned, assembles the enumerated candidate-country list the regulation requires instead of a single guess.
  • Error pre-flight — flags the specific error classes APHIS writes letters about, especially implausible country/species pairings, before the entry transmits.
  • Broker handoff file — clean per-line species/origin/quantity/value output your broker can transmit to ACE, or that you enter in LAWGS.
  • Due-care dossier — timestamped, exportable record per SKU and per entry: every ask, every reply, every conclusion and who signed off. Built before the shipment, not reconstructed after a letter arrives.

6. AI angle — what’s load-bearing

Remove the AI and this is a spreadsheet plus a nag email — which is precisely the status quo that trade associations are papering over with question scripts.

Three jobs are genuinely model work:

  1. Unstructured multilingual supplier replies → structured per-line assertions. The input is a photographed mill certificate in Vietnamese, a WeChat screenshot, a PDF packing list with a hand-written note, or a one-line email saying “acacia wood, from plantation.” The output must be a genus/species candidate set, an origin claim, a confidence level, and a citation back to the source document. This is the whole product and it was not cheaply doable at this quality two years ago.
  2. Trade name → taxon disambiguation with honest uncertainty. “Acacia” spans many species; “meranti” spans genera; “rubberwood” is Hevea brasiliensis but arrives called ten things. The model must propose candidates and refuse to collapse to one when the evidence doesn’t support it — because a confident wrong species is worse than a flagged unknown. That refusal behaviour is the design.
  3. Plausibility screening. APHIS explicitly writes letters for “unlikely country and plant species pairs.” Judging whether a claimed species plausibly grows in and is exported from a claimed country, across thousands of pairings, is a reasoning task over reference data.

Note what the AI deliberately does not do: it does not visually identify wood species from photos. XyloTron-class computer vision is impressive research — 97% cross-validation, 86.5–92.4% on independent specimens for Peruvian timbers — but the inventors’ own position is that it “doesn’t always identify specimens correctly, so there is still work to be done before it can be fully commercialized,” and it needs prepared end-grain on a physical sample you don’t have. Lab DNA identification runs $500–$1,000 per sample. Selling physical identification would be selling a research project. We sell the paperwork chase, which is where the actual unmet need sits.

7. Localization angle (if any)

N/A as a market — the buyer is a US importer of record, because the Lacey declaration is a US filing.

But supplier-side localization is a core feature, not a nicety. The chase agent has to work in Vietnamese, Simplified Chinese, Bahasa Indonesia and Hindi, over the channels those factories actually use (email and WhatsApp/WeChat, not a Western SaaS portal), and it has to parse replies that arrive as photos of paper. A generic English-only “supplier portal” is exactly the design that has failed this segment — factories don’t log in. Getting reply rates up in-language is the operational moat.

There is a plausible second act selling the same dossier to EU importers under EUDR, but I’d not build for both at once — and my own catalog already has EUDR coverage, so treat that as an expansion, not a wedge.

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

  • Pricing: three tiers, priced against the broker invoice rather than against GRC software.
    • Starter $199/mo — up to 75 active SKUs, 1 user. The small luthier / boutique importer.
    • Growth $549/mo — up to 400 SKUs, unlimited chases, due-care dossier export, broker handoff. The core segment.
    • Broker/Consultancy $1,200–$2,500/mo — multi-client workspaces, white-label supplier outreach, per-client dossiers. Sold to customs brokers and trade consultancies who resell to their book.
  • ACV: ~$6,600 blended on the Growth tier; ~$21K on Broker. Blended target $8,000–9,000.
  • Rough math to $1M ARR: 105 Growth customers ($693K) + 15 Broker/Consultancy at $1,800/mo ($324K) = $1.017M. That is ~120 paying accounts — reachable from a named, enumerable universe.
  • Rough math to $5M ARR: needs ~450 Growth + 70 Broker accounts, and realistically a second regime (EUDR, or CITES-adjacent species permitting) sold into the same dossier, plus broker channel doing the acquisition. What must be true: broker resale actually works, and each new Phase VII expansion delivers a fresh cohort with 6 months’ notice.
  • Expansion path: SKU-count tiers grow naturally as catalogs grow; per-entry dossier volume; add regimes to the same record (EUDR, CITES, formaldehyde/TSCA Title VI already sits on the same furniture SKUs); charge for the supplier-network effect — once a factory has answered once, that answer is reusable for every customer buying from it, which is the real compounding asset.
  • Gross margin: high but not pure-software. Parsing scanned multilingual documents costs real inference, and hard cases want a human reviewer in year one. Budget 75–82% early.

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

This universe is enumerable, which is why I like it.

  1. Scrape the IWPA member directory, then the customs data. The International Wood Products Association runs a public, searchable member directory by species, product and region. Separately, US import records are commercially available by HTS code and consignee — the exact Phase VII codes are published. Build a list of importers of record filing on Phase VII codes from Vietnam/China/Indonesia, filter to the 5–80 employee band, and you have a targeted list in the low thousands. Send each one a free “Phase VII exposure report”: here are your SKUs likely requiring a declaration, here is which ones can’t use a SUD, here is your de minimis position. That report is the demo, generated from data before they ever talk to me. Expect 3–6% to book a call on a list that specific.
  2. Sell the broker channel, because brokers are pre-aggregated demand. Brokers file the declaration and share accuracy liability, but have no way to get the data — they currently chase clients by email. Pitch 60–100 mid-size customs brokerages a white-label workspace: their clients get the chase, the broker gets clean line data and stops absorbing the back-and-forth. One broker converting brings 5–20 importers. This is the fastest path to the first $250K.
  3. Trade associations that are already hand-servicing this gap. Alliance-International is distributing a question-script PDF; NAMM arranged free APHIS compliance training for luthiers and small shops. Offer the association a co-branded version of the exposure report for their members. They have the distribution and a demonstrated willingness to service exactly this pain for free — which means they’ve already qualified the audience for me.
  4. Ambush the letter. APHIS sends letters of non-compliance for declaration errors. Nobody who receives one has a workflow fix. Publish the single best public resource on responding to an APHIS non-compliance letter and on the country/species plausibility trap, and run narrow paid search on the error language and on “PPQ 505” / “LAWGS” / “Lacey declaration species unknown.” Small volume, extreme intent.
  5. The luthier beachhead for proof, not revenue. Small instrument shops are vocal, networked, and already organised around this pain (NAMM, forums, the RELIEF Act advocacy). They are too small to fund the company, but 15 of them at $199 produce public testimonials and edge-case coverage that convert the furniture importers who actually pay.

10. Build complexity — justification

Medium. Off-the-shelf: LLM document parsing, email/WhatsApp sending, standard web stack, the published HTS and Phase VII code lists, public taxonomic references like GRIN for species validation.

The custom work is real but bounded: a trade-name-to-taxon mapping with honest ambiguity handling, the de minimis and SUD rule engine encoded from the regulation, a country/species plausibility screen, and a chase-scheduler that behaves like a persistent human. Deliberately not in v1: direct ACE transmission (we hand the broker a file — that avoids becoming a filing intermediary and shortens the build by months), and any physical species identification. Realistic v1 for two people: 14–18 weeks.

The genuine risk is not engineering, it’s domain encoding. Getting the SUD eligibility and de minimis logic wrong produces confidently incorrect compliance advice, which is worse than no product. This needs a licensed customs broker or Lacey-experienced trade consultant reviewing the rule engine before anyone relies on it — budget for that as a real line item, not a favour.

11. Gating checklist

GatePass?Note
Legal in target market✅Helping importers assemble accurate declarations and due-care records. No filing licence needed if we hand data to the broker rather than transmit; a customs broker licence would be required to file on a client’s behalf, which v1 deliberately avoids.
Ethical — no harm / dark patterns✅Pushes toward accurate declaration and honest uncertainty. Explicitly refuses to guess a species to make a field green — the design goal is flagging unknowns, not laundering them.
Market exists (evidence above)✅Phase VII (250 codes, Dec 2024), paper ended Jan 2026, $110K+$250K enforcement case, SBA Advocacy burden letter, associations distributing question scripts.
1–5 person team can build this✅Two builders, 14–18 weeks, plus paid domain review.
Launchable with <$50K / ₹40L✅~$25–35K: import-data subscription, domain expert review, inference costs, outbound tooling.

All five pass.

12. Feasibility score

AxisWeightScoreNotes
Problem intensity2016/20Real teeth: shared importer/broker accuracy liability, $110K criminal + $250K administrative in a 2024 case, APHIS letters for error classes, and one field (harvest country) with legally no escape hatch. Held under 17 because the pain is per-shipment and many importers are quietly filing sloppy declarations without consequence yet — enforcement is real but not dense.
Demand evidence1511/15Strong indirect and institutional evidence: scope expansion, dated paper cutover, SBA Advocacy letter, associations hand-distributing scripts, a dealer hand-building a species database, NAMM arranging APHIS training. What I could not find is a body of importers publicly asking for this software — no forum thread of people begging for a Lacey chase tool. That gap is why this isn’t 13+.
Build feasibility1512/15Standard stack plus a bounded rule engine; ~14–18 weeks for two. Docked for the domain-encoding risk and the multilingual scanned-document parsing quality bar.
Distribution clarity1512/15Enumerable universe (IWPA directory, import records by Phase VII HTS code), a pre-aggregated broker channel, and associations already servicing the pain. Docked because broker resale is asserted, not yet proven, and the free-exposure-report conversion rate is an estimate.
Revenue mechanics1511/15$199–$2,500/mo is defensible against $150–$400 per-entry broker fees and against a $383M-funded enterprise alternative with no public price. ~120 accounts to $1M. Docked because I have no observed willingness-to-pay datapoint for this specific product, and margin is softer than pure SaaS.
Time to first revenue88/10The exposure report is a pre-sale instrument — it can sell before the chase engine is finished. Realistically 8–10 weeks to first paid pilot; not ≤4 weeks, because compliance buyers want the rule engine reviewed before they trust it.
Defensibility106/10Soft but compounding: an accumulating factory-answer corpus (each supplier answer is reusable across customers), encoded SUD/de minimis logic, workflow lock-in via the dossier of record, and broker channel relationships. Not 8+ because Altana could step down-market and a competent competitor could rebuild the rule engine in a quarter.
Total10076/100

13. Qualitative modifiers

Founder-fit tags

technical-heavy · domain-expertise-required

Needs someone who can build reliable multilingual document extraction, paired with real customs/Lacey domain access — a licensed broker or trade compliance consultant as co-founder or paid advisor. Without the domain half, the rule engine will be confidently wrong and the first enforcement-adjacent customer will discover it.

Key assumptions to validate (3–5)

  1. Assumption: Small/mid importers in Phase VII categories know they’re in scope and are dissatisfied with how they currently produce species/origin data. How to test: 30 structured calls with importers pulled from import records on Phase VII HTS codes. Ask what they put in the harvest-country field last shipment and how they got it. If most say “our broker handles it” and can’t describe a chase, the pain isn’t owned by my buyer and the wedge shifts entirely to brokers.
  2. Assumption: Factories will answer a structured, in-language, persistent request at a usable rate. How to test: run the chase manually for one friendly importer across 40 suppliers in Vietnam/China/Indonesia. Measure reply rate and how many replies contain a usable genus/species or a document. Below ~35% usable replies, the product’s core promise fails regardless of software quality.
  3. Assumption: Customs brokers will resell rather than build or ignore. How to test: pitch 20 mid-size brokerages the white-label workspace. Look for 3+ signing a paid pilot. If brokers say “we just type what the client sends and that’s the client’s problem,” the channel is dead and CAC roughly doubles.
  4. Assumption: $549/mo clears for a 400-SKU importer. How to test: price-test the exposure report → paid conversion at $349 / $549 / $799 across three cohorts.
  5. Assumption: The SUD and de minimis logic can be encoded to a standard a compliance professional will sign off on. How to test: have a licensed broker adversarially review 50 adjudicated SKUs and count disagreements. More than 5 material disagreements means the rule engine needs a human-in-the-loop tier, which changes the margin structure.

Risk flags

  1. Regulatory reversal risk: the RELIEF Act (Retailers and Entertainers Lacey Implementation and Enforcement Fairness), which NAMM supports, would clarify the Act so instrument manufacturers, retailers and resellers aren’t penalised for unknowingly possessing illegal woods. Broad relief legislation, or a blanket-declaration provision — which stakeholders have already proposed to reduce repeat declarations of identical products — would blunt the pain materially. My own proposal-stage ≠ law and relief-that-creates-exposure notes apply: track it, don’t panic, but a passed blanket-declaration rule is a genuine partial kill.
  2. Wrong-buyer risk: if brokers, not importers, functionally own this field, the product is a broker tool with maybe 400 total customers in the US — a smaller, slower business than the importer-facing version. This is the single most likely way the thesis is wrong, and assumption 1 is designed to find out fast.
  3. Liability exposure: we produce data feeding a declaration that carries criminal exposure. Contracts must be unambiguous that we assemble and evidence supplier claims rather than certify species, and the product must be built to surface uncertainty rather than fill fields. Errors-and-omissions cover from day one, not month twelve.
  4. Enforcement density: the civil administrative penalty for merely violating the declaration requirement is capped at $250 (rising to $10,000 for knowing violations, with separate customs and 18 USC §542 exposure). If an importer’s realistic downside for a sloppy-but-not-knowing declaration is a $250 penalty and a letter, some will rationally choose to keep guessing. The sale then has to lead with the knowing-violation and forfeiture tail, which is a harder, fear-based pitch.
  5. Incumbent step-down: Altana already has the species/origin data assets and $383.6M. A self-serve SMB tier from them would be a direct hit — though large-funded enterprise vendors are historically slow to serve a $549/mo band.

14. Structured verdict

Score:                  76/100
Verdict:                GO
Confidence:             Medium
Best-fit builder:       Technical founder strong on document-extraction AI, paired with a
                        licensed customs broker or Lacey-experienced trade consultant
Time to revenue:        8–10 weeks to first paid pilot
Capital to launch:      $25–35K (₹22–30L)
Top 3 assumptions to validate first:
  1. Importers, not just brokers, own the species/harvest-country field — 30 calls sourced
     from import records on Phase VII HTS codes; ask what they filed last shipment and how
     they got the answer
  2. Overseas factories answer a persistent in-language structured request at ≥35% usable-reply
     rate — run 40 suppliers manually for one friendly importer before writing the agent
  3. Customs brokers will resell a white-label chase rather than shrug — pitch 20 brokerages,
     look for 3 paid pilots
Kill criteria:
  - Abandon if <35% of 40 manually-chased factories return a usable species/origin answer or
    document — the core promise is undeliverable at any software quality
  - Abandon if fewer than 4 of 30 interviewed importers can describe owning this field, AND
    fewer than 2 of 20 brokerages will pilot — no buyer owns the pain
  - Abandon if a blanket-declaration provision or RELIEF-Act-style relief is enacted covering
    the Phase VII categories, removing the per-shipment chase
  - Abandon if price-testing cannot clear $349/mo at 400 SKUs — the unit economics won't carry
    multilingual document parsing plus human review

15. Next step — 1-week validation sprint

  • Day 1–2: Buy one month of US import records. Pull importers of record filing on 12 representative Phase VII HTS codes (wood furniture, bamboo/rattan, footwear with wood components, hand tools) from Vietnam, China and Indonesia. Filter to 5–80 employees. Build the list and hand-generate 5 real exposure reports — SKUs likely in scope, which can’t use a SUD, de minimis position. This doubles as the sales asset and proves the data pipeline exists.
  • Day 3–4: Two parallel tests. (a) Book and run 12–15 calls from that list with one question that matters: what did you put in the country-of-harvest field on your last shipment, and how did you get it? (b) Recruit one friendly importer and manually chase 40 of their suppliers in-language — measure usable-reply rate. Simultaneously pitch 8 mid-size customs brokerages the white-label workspace.
  • Day 5: Decide on three falsifiable numbers: usable factory reply rate ≥35%, ≥4 of 12–15 importers describing personal ownership of the species/origin field with a workaround they dislike, and ≥1 of 8 brokerages agreeing to a paid pilot. Hit all three → build. Hit the reply-rate test but fail the importer-ownership test → rebuild the thesis as a broker-channel product and re-price. Miss the reply-rate test → kill it; no amount of software fixes a supply chain that won’t answer.

Interested in a detailed proposal?

Get a deep-dive with market research, competitive analysis, and implementation roadmap.

Contact us

info@startupbasket.ai