SB StartupBasket
All ideas
77 /100 GO Low complexity

TonneGauge — CBAM headroom gauge for exempt EU importers

Tracks how close an EU importer is to CBAM's 50-tonne cliff, months before crossing it costs them retroactively.

Compliance SaaS EU CBAMImportersCustomsThreshold-monitoringSMBSolo-builderAI-agentCompliance-driven
— views
Evaluation Scores
77/100

GO

Overall Score

16
Problem
12
Demand
13
Build
11
Distrib.
11
Revenue
8
Time
6
Defense

TonneGauge — CBAM headroom gauge for exempt EU importers

1. One-liner

Tracks how close an EU importer is to CBAM’s 50-tonne cliff, months before crossing it costs them retroactively.

2. Trend signal — why now?

On 1 January 2026 the EU’s Carbon Border Adjustment Mechanism entered its definitive regime. The Omnibus simplification that came with it replaced the old €150 per-consignment rule with a single cumulative 50-tonne annual mass threshold across four sectors — iron and steel, aluminium, cement, fertilisers. Import 50 tonnes or less in a calendar year and you owe nothing. Import 50.1 tonnes and you owe everything.

The Commission’s own impact assessment (SWD(2025) 58) put a number on who this creates: a 50-tonne threshold “would lead to the exemption of 91% of importers … representing an estimated 182,000 exempted importers.” The Commission scored the saving at €990m–€1.26bn per year in avoided admin cost. That is the entire justification for the threshold — and it is also, precisely, the population that now has to prove it stayed under.

Here is the part that makes this a product rather than a spreadsheet. Three rules stack into a trap:

  1. It is retroactive and all-or-nothing. “Once the cumulative 50-tonne limit is exceeded at any point during a calendar year, full CBAM obligations apply to all relevant imports from 1 January of that year.” You cannot declare the first 50 tonnes exempt and pay on the excess. Cross by one tonne in November and January’s containers are back in scope.
  2. Authorisation must exist before the crossing tonne lands. Not after, not at year end. And the national competent authority has up to 120 calendar days to assess an application. As of 7 January 2026 more than 12,000 operators had applied and only about 4,100 had actually obtained authorised-declarant status.
  3. Importing unauthorised is punished at a multiple. Roughly €300–500 per tonne CO₂e against a standard rate near €100/tCO₂e — a 3–5× penalty.

So the failure mode is a four-month hole. An importer who notices at 48 tonnes is already too late to be authorised before tonne 51, because the authority needs 120 days and the goods are on the water now.

Meanwhile the regulator explicitly expects a running mass log — “a dated record of each CBAM import, the CN code, country of origin, net mass in tonnes, and the cumulative total for the calendar year” — as the evidence that the importer monitored the threshold actively. Exempted importers must also state the exemption in the customs declaration itself. So the duty to watch is real, dated, and documented. Nobody sells the watch.

The vendors are all standing on the wrong side of the line. Dubrink’s importer pricing starts at €1,990/yr and is built “for importers handling between 50 tonnes and 15,000 tonnes.” CBAMBOO is reported at €9,000/yr basic and €19,000 Pro. Those products begin where my customer’s problem ends. They sell emissions calculation and declaration filing to companies already in scope. The 182,000 exempt importers do not need an emissions engine — they need to know whether they are about to need one.

Provenance:

3. The opportunity

Every CBAM vendor in the market sells the same thing: help computing embedded emissions and filing the declaration. That is the right product for the ~18,000 importers above the line. It is entirely the wrong product for the 182,000 below it, whose only question is binary and arrives once a year: am I going to cross?

The gap is a classic threshold-tripwire with a lead-time twist. Normally a threshold product warns you at the moment of crossing. Here, warning at the moment of crossing is worthless — the 120-day authorisation clock means the only useful alert fires when you are at thirty tonnes and trending, four months out. The product is not a counter. It is a forecast with a procedural deadline attached: given your import cadence to date, your open POs and your seasonal shape, will you cross before 31 December, and is there still time to be authorised first?

The incumbents structurally cannot serve this. Their pricing floor (€1,990–€9,000/yr) is calibrated to companies with real certificate liabilities, and their onboarding assumes you are collecting supplier emissions data — an expensive, months-long exercise that an exempt importer has no reason to start. Selling a €9,000 emissions platform to a company whose correct answer is “you owe nothing, here’s the log that proves it” is not a discount problem, it’s a product-shape problem.

There is a second, sharper wedge underneath: misclassification. The documented industry failure case is an importer whose supplier called the goods “parts,” outside CBAM scope, when their actual customs classification put them squarely inside it. Tonnage you don’t know is CBAM tonnage doesn’t get counted, and the CN code — not the supplier’s opinion, not the commercial invoice description — is what determines scope. Reconciling the broker’s declared CN codes against the CBAM Annex I list is exactly the kind of judgement that is tedious for a human and cheap for a model.

4. Target market

  • Primary customer: The person who owns customs at a small EU importer — typically the operations or finance lead, or a one-person logistics function — at companies importing steel fixings, aluminium profiles, fasteners, castings, tooling, HVAC and structural components. €2M–€50M revenue. Concentrated in Germany, Italy, Poland, Netherlands, Spain, Czechia. They import somewhere between 10 and 60 tonnes of CBAM goods a year, which is to say: they genuinely don’t know which side of the line they’re on.
  • Why they buy: Because the downside is asymmetric and retroactive. A €12,000 unbudgeted certificate liability plus a 3–5× unauthorised-import penalty on a year of shipments, discovered in the following February, lands on a company whose entire annual margin on that product line might be €40,000. They are not buying carbon compliance. They are buying the sentence “you are at 31 of 50 tonnes, and if you keep buying at this rate you cross in October — apply now or stop ordering.”
  • Rough TAM reasoning: The Commission counts 182,000 exempted importers. Most are true occasional importers with 2 tonnes a year who will never care. The live segment is the band plausibly between roughly 20 and 60 tonnes — call it 10–15% of that population, so 18,000–27,000 companies. At €600/yr average, capturing 2% of the mid estimate is ~€270K ARR; 8% is ~€1.1M. That is a real bootstrapped business and far too small to interest the funded CBAM platforms.
  • Why now for them: Calendar year 2026 is the first year the counter runs. There is no prior-year baseline anyone can lean on, no institutional habit, and the retroactive rule means the first genuine reckoning arrives with the 2027 declaration on 2026 volumes. Companies that grew imports in 2026 are discovering the question late, which is exactly when a forecast is worth paying for.

5. Product sketch (MVP)

  • Running mass log — every CBAM-relevant import, dated, with CN code, country of origin, net mass in tonnes, and the year-to-date cumulative total. Formatted as the evidence record regulators expect.
  • Headroom gauge — the single number on the dashboard: tonnes used, tonnes remaining, percentage of the 50-tonne allowance consumed.
  • Crossing forecast — projects the year-end total from cadence to date plus open purchase orders, and names the month you cross. Recalculates on every new declaration.
  • 120-day authorisation alarm — fires when the projected crossing date is less than 120 days out, because that is the last moment applying for authorised-declarant status can still beat the goods.
  • CN-code scope reconciliation — checks the broker’s declared CN codes against the CBAM Annex I list and flags goods your supplier calls “parts” that customs classification puts in scope.
  • Broker document intake — forward the customs declaration PDF or the monthly broker report to a dedicated inbox; the tonnage and CN codes get extracted and added to the log without manual entry.
  • Exemption evidence pack — a dated, exportable file showing you monitored the threshold and stayed under it, for the year the question gets asked.
  • What-if check — “if I place this 9-tonne order, where does it put me?” answered before the PO goes out, not after.

6. AI angle — what’s load-bearing

Two places, both real.

First, document extraction. The input is whatever the broker sends — a customs declaration PDF, a monthly summary spreadsheet with idiosyncratic column names, a forwarder’s emailed statement. Net mass and CN code have to come out of unstructured, per-broker-variable documents reliably. This is the work that otherwise makes the customer maintain the spreadsheet by hand, which is precisely why they don’t maintain it.

Second, and more valuable, scope classification. Deciding whether a given line item is a CBAM good is a genuine judgement call: it turns on the CN code against Annex I, not on the product description, and the documented industry failure is exactly a supplier calling in-scope goods “parts.” A model that reads the description, the declared code and the supplier’s own classification, then flags the disagreements for review, is doing the work a customs consultant charges hourly for.

Strip the AI out and this is a spreadsheet the customer already isn’t keeping. The reason the log doesn’t exist today is that populating it is manual, per-shipment and boring — which is the definition of the work to automate. The forecast maths itself is trivial arithmetic; I’m not going to pretend otherwise. The AI earns its place upstream of the arithmetic, on getting clean tonnage out of messy documents.

7. Localization angle (if any)

EU-first by construction — the regulation is the market. But localisation matters within the EU in two specific ways.

Language is the obvious one: broker documents arrive in German, Italian, Polish, Spanish, Dutch. Extraction has to work on all of them from day one, which is a genuine barrier to a US-built generic tool and a non-issue for a model-based approach.

The subtler one is that the national competent authority differs per member state — Germany’s DEHSt, Ireland’s EPA, and so on — each with its own application portal, guidance and practical timeline. The authorisation alarm is only useful if it points at the right authority with the right lead time. Starting with Germany, Italy and Poland (highest concentration of small metal importers) and adding authorities one at a time is a defensible sequencing, and each one added is accumulated operational knowledge a competitor has to redo.

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

  • Pricing: €39/mo (€390/yr) single-entity, up to 100 import lines a year. €89/mo (€890/yr) for multi-entity groups, higher line volume, and the CN-scope reconciliation on every line. Annual billing pushed hard — the product’s unit of work is the calendar year, so the subscription should be too.
  • ACV: ~€600 blended.
  • Rough math to $1M ARR: ~1,700 customers × €600. Against a live segment of 18,000–27,000, that is 6–9% penetration. Achievable but not a layup — it requires the broker and association channels below to actually convert.
  • Rough math to $5M ARR: Not reachable on this product alone, and I’d rather say so than dress it up. €5M needs either (a) following customers across the line — when someone crosses, they become an authorised declarant who needs declaration filing, so the natural upsell is a €2,000–3,000/yr filing tier that competes directly with Dubrink, or (b) the same threshold-watch shape applied to CBAM’s announced scope expansion into downstream products, which widens the in-scope CN list substantially. (a) is the honest path; it converts a €600 customer into a €2,500 one at exactly the moment they have budget, because they just discovered a real liability.
  • Expansion path: entity count, then line volume, then the post-crossing filing tier. The crossing event is the expansion trigger, which is a genuinely nice property — the product tells you when to upsell, and the customer agrees.

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

  • Customs brokers and freight forwarders are the channel, and they’re motivated. A broker with 300 SME importer clients is being asked “am I affected?” by dozens of them and has no tool to answer with. Build a free broker-facing bulk screen: they upload a year of client declarations, get back a ranked list of which clients are closest to 50 tonnes. That output is a client-retention asset for them and a qualified lead list for me. Target 30 mid-size brokers in Germany, Italy and Poland; each one that adopts brings 5–20 importers.
  • The free public calculator, aimed at the question people actually type. A no-signup page: paste your CN codes and tonnages, get your headroom and projected crossing month. Free calculators already exist for CBAM cost, none for headroom — and the free tool is the honest lead magnet here because the one-shot answer is genuinely different from the year-round watch. Rank for “CBAM 50 tonne threshold,” “am I exempt from CBAM,” “CBAM de minimis.”
  • Sector associations for metals distribution and fasteners. Every member is by definition importing steel or aluminium in exactly the ambiguous volume band. A 40-minute “are you sure you’re under 50 tonnes?” webinar with a national metal-traders association, done three times across DE/IT/PL, puts the specific question in front of a few hundred qualified companies at once.
  • Direct outbound off trade data. Several EU member states publish transaction-level or aggregated import data by importer. Where importer-level data is available, the sub-50-tonne steel and aluminium importers can be identified directly and emailed their own estimated headroom number as the opening line. A cold email that already contains the recipient’s approximate tonnage is a different object from a cold email asking for a meeting.
  • Ride the 2027 declaration deadline. The first surrender of certificates for 2026 goods is due by 30 September 2027. The nine months before that is when every importer who guessed wrong finds out. Content and outbound should be timed to compound into that window.

10. Build complexity — justification

Low. The core is document extraction into a per-entity running total, a projection, and an alert — no emissions engine, no supplier data collection, no verifier workflow, which is precisely what makes the incumbent products heavy and this one light. The genuine work is the CN-code reference data (Annex I, maintained against amendments), multilingual broker-document extraction, and per-member-state authority guidance. A solo builder ships a credible v1 in 6–8 weeks; the free calculator is a weekend.

The thing that would make it Medium is direct customs-system integration, and I’d deliberately not build that for v1 — email-the-broker-document intake covers the same ground at a fraction of the cost and works identically in every member state.

11. Gating checklist

GatePass?Note
Legal in target market✅Monitoring and record-keeping tool. Does not file on the customer’s behalf in v1, so no customs-representative licensing question.
Ethical — no harm / dark patterns✅Helps small importers comply and avoid retroactive penalties. The failure mode of the product is telling someone they owe more compliance, not less.
Market exists (evidence above)✅Commission’s own impact assessment counts 182,000 exempted importers; vendor pricing explicitly starts at the 50-tonne floor.
1–5 person team can build this✅Solo builder, 6–8 weeks to v1.
Launchable with <$50K / ₹40L✅Well under €10K. Main cost is the founder’s time plus modest inference.

All five pass.

12. Feasibility score

AxisWeightScoreNotes
Problem intensity2016/20Retroactive all-or-nothing liability with a 3–5× penalty multiplier is genuinely painful, and the 120-day authorisation lead time means late discovery is unfixable. Docked because for most of the year it’s a background worry, not a daily one — the pain is annual and lumpy, which is a real conversion risk.
Demand evidence1512/15Regulation is live and dated, Commission quantifies the population at 182,000, vendors price from €1,990 explicitly above the line. Docked: I found no verbatim importer complaints — the demand is inferred from regulatory structure and vendor pricing, not from people asking for this out loud.
Build feasibility1513/15Extraction plus arithmetic plus reference data. No emissions engine. 6–8 weeks solo.
Distribution clarity1511/15The broker bulk-screen is a specific, motivated channel with a real artefact. Docked because broker adoption is unproven and the direct-outbound path depends on importer-level trade data availability varying by member state.
Revenue mechanics1511/15€390–890/yr is comfortably affordable against the downside and sits far under incumbent floors. Docked: €1M needs 6–9% penetration of the live band, and €5M requires the upsell into filing, which means competing with funded vendors on their turf.
Time to first revenue108/10Free calculator to paid conversion can start within weeks of launch, but annual-cycle buying and the fact that urgency peaks near year-end and near the Sept 2027 deadline means some of the demand is seasonal.
Defensibility106/10Soft moat: accumulated per-member-state authority knowledge, broker relationships, and the customer’s multi-year running log which is the record they must retain. But the core forecast is copyable, and a CBAM incumbent could ship a free headroom tool as a funnel. That’s the main strategic risk.
Total10077/100

13. Qualitative modifiers

Founder-fit tags

technical-heavy · domain-expertise-required

Needs someone who will actually read the CBAM implementing regulation and track its amendments, and who can run multilingual document extraction reliably. The customs-domain knowledge is learnable in weeks, but it is not optional.

Key assumptions to validate (3–5)

  1. Assumption: A meaningful number of EU importers are genuinely in the ambiguous 20–60 tonne band and know they are. How to test: Get 10 customs brokers to run a year of client declarations through the free bulk screen. Count what fraction of their book lands between 20 and 60 tonnes. If it’s under 5%, the live segment is far smaller than my estimate and the whole thesis shrinks.
  2. Assumption: Importers will pay for a watch rather than checking manually once a year. How to test: 25 direct conversations with importers in the band. The question is not “is this useful” but “would you pay €390 today” — and specifically whether they believe their own tonnage number well enough to feel safe without it.
  3. Assumption: Brokers will distribute rather than absorb — i.e. they’ll pass the tool to clients instead of quietly running it themselves as a service. How to test: Offer the bulk screen to 30 brokers and see whether any ask about a referral or white-label arrangement, versus using it once and going quiet.
  4. Assumption: Broker documents can be extracted accurately enough across 4+ languages that customers trust the running total. How to test: Collect 100 real declaration documents across DE/IT/PL/NL and measure tonnage extraction accuracy. Below ~97% on net mass and the product is worse than useless, because a wrong total is more dangerous than no total.

Risk flags

  1. Regulatory risk (high, and cuts both ways): CBAM is being actively amended. The threshold moved from €150-per-consignment to 50 tonnes once already; it could move again, and announced scope expansion into downstream products would change the CN list. An upward threshold revision would shrink the anxious band and gut the product. Conversely, scope expansion grows it.
  2. Incumbent funnel risk: Nothing stops Dubrink or CBAMBOO shipping a free headroom checker as lead-gen for their paid tiers. They have more reason to give it away than I have to sell it. This is the single most likely way the idea dies, and it’s why the broker channel and the multi-year retained log matter more than the calculator.
  3. Seasonality and low-urgency risk: The pain concentrates near year-end and near the September 2027 surrender deadline. A subscription product with lumpy annual anxiety can see high churn right after the crossing question resolves — customers who confirm they’re safely at 12 tonnes have little reason to renew.
  4. Data-dependency risk: The product is only as good as the broker’s documents. Importers using multiple forwarders, or whose forwarder reports arrive late or incomplete, will have a lagging total — and a lagging total on a 120-day lead-time problem is a partial failure of the core promise.

14. Structured verdict

Score:                  77/100
Verdict:                GO
Confidence:             Medium
Best-fit builder:       Technical solo founder willing to learn EU customs classification;
                        EU-based helps for broker and association relationships
Time to revenue:        6–10 weeks
Capital to launch:      €5–8K
Top 3 assumptions to validate first:
  1. Live segment size — run the broker bulk screen across 10 brokers' client books,
     measure what share sits in the 20–60 tonne band (need >5%)
  2. Willingness to pay for a watch vs an annual manual check — 25 importer conversations,
     asking for €390 today rather than for interest
  3. Extraction accuracy on real multilingual broker documents — 100 documents,
     need >97% on net mass
Kill criteria:
  - Abandon if <5% of screened broker client books fall in the 20–60 tonne band
  - Abandon if a funded CBAM incumbent ships a free headroom/threshold forecaster
    before v1 launches
  - Abandon if fewer than 5 of 25 importer conversations will commit €390 pre-launch
  - Abandon if the de minimis threshold is revised materially upward

15. Next step — 1-week validation sprint

  • Day 1–2: Build the free headroom calculator — CN codes and tonnages in, headroom and projected crossing month out. No signup. Ship it and point it at “CBAM 50 tonne threshold” and “am I exempt from CBAM.”
  • Day 3–4: Email 30 customs brokers and freight forwarders across Germany, Italy and Poland offering the free bulk client screen. The ask is their client declaration data in exchange for a ranked at-risk list. Measure how many accept — that number alone tells me whether the channel is real.
  • Day 5: From whatever broker data comes back, identify importers in the 20–60 tonne band and call 25 of them. Do not ask whether it’s useful. Ask for €390 for the 2026 year, today, pre-launch.

Falsifiable outcome: at least 5 of 25 importers commit €390 pre-launch, and at least 3 of 30 brokers hand over client data for the screen. Fewer than 5 paying commitments means the annual-check workaround is good enough and the watch isn’t worth money — kill it. Brokers refusing to engage while importers still pay means the product survives but the cheap channel doesn’t, and the economics need rebuilding before committing to the build.

Interested in a detailed proposal?

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

Contact us

info@startupbasket.ai