Brussels Kicked 182,000 Importers Out of CBAM. The 18,000 Left Behind Owe Real Money by September 2027.
On January 1, 2026, the EU's Carbon Border Adjustment Mechanism entered its definitive phase: real money, real certificates, real penalties. Then the simplification regulation drew a 50-tonne line that exempted roughly nine in ten importers. What remains is a concentrated market of about 18,000 authorized declarants, each one legally on the hook for installation-level emissions data from suppliers who have no duty to cooperate, certificate purchases that cannot be undone once submitted, and a first annual declaration due September 30, 2027. The compliance industry is selling carbon accounting to sustainability teams. The liability sits with trade teams. That mismatch is the business.
The Problem
Start with the first week of January 2026. In seven days, EU customs validated 10,483 import declarations containing CBAM goods, covering 1.66 million tonnes, with iron and steel making up 98 percent of it. More than 12,000 operators applied for authorized declarant status; 4,100 had obtained it. Turkey, China, India, Canada, Taiwan, and Vietnam led as origins; Belgium, Spain, Romania, the Netherlands, France, and Germany dominated as importers. The carbon border tax is not a proposal anymore. It is a live customs event, running every week, on about 50 billion euros of covered imports a year.
Here is the part that matters for this business: the authorized declarant, not the supplier and not the broker, bears the legal liability. Under Regulation 2023/956 as simplified by Regulation 2025/2083, only an authorized CBAM declarant may release covered goods for free circulation. That declarant must report installation-level embedded emissions for every consignment, have the figures verified by an accredited verifier, and surrender CBAM certificates matching the emissions. A supplier's glossy corporate sustainability report is not usable data; as one CBAM methodology guide puts it, where goods have several production sources, the allocation needs evidence, and a supplier's corporate sustainability report does not automatically provide usable installation-level CBAM data. The Turkish steel mill, the Indian aluminum smelter, the Chinese fastener plant: none of them is subject to EU law, none of them faces a penalty for ignoring a data request, and every one of them holds the numbers the declarant's liability depends on.
The fallback is designed to hurt. When verified actual data is unavailable, the Commission applies default emissions values, and under the simplification these are set punitively: the average emission intensity of the ten highest-emitting exporting countries with reliable datasets. On top of that, a markup applies to importers using defaults instead of actuals, rising from 10 percent in 2026 to 20 percent in 2027 and 30 percent from 2028 for steel, aluminum, cement, and hydrogen, per the draft certificate rules. Defaults are not a safe harbor. They are a fine wearing a methodology costume, and they get more expensive every year by statute.
Then there is the purchasing. CBAM certificates are priced off the EU ETS auction average: 75.36 euros per tonne of CO2 in Q1 2026 and 75.28 in Q2, with recent reporting putting the price around 82.32 euros. Certificate sales through the central platform open February 1, 2027. Each purchase carries a 5-cent administrative fee per certificate, applications run 1 to 99,999 certificates, and once a purchase application is submitted its content cannot be changed, with only one repurchase application permitted per year, under the draft rules now out for consultation. Mistimed or mis-sized buying is locked in for the year. This is a treasury function stapled to a customs function stapled to a supplier-data function, and most declarants have staffed exactly zero of the three.
Picture the actual customer: a German distributor importing 12,000 tonnes of Turkish steel sections a year. It cleared the 50-tonne de minimis easily, so it is in scope. Its three Turkish mills each sent a different spreadsheet, one in Turkish, one with plant-level rather than installation-level figures, one with nothing at all. The distributor's customs broker files entries and wants nothing to do with emissions methodology. Its sustainability consultant sold it a CSRD report last year and has never touched a CBAM declaration. The 2026 imports it is receiving right now are the data for the declaration due September 30, 2027, and every month that passes without installation-level evidence is a month closer to filing on punitive defaults. This company is not confused about whether CBAM applies. It is drowning in the three functions at once, which is a billable problem.
Why Now
The de minimis just concentrated the market. The 50-tonne annual threshold replaced the old 150-euro-per-consignment rule and, by the Commission's own estimate, removes about 182,000 importers, roughly 90 percent, from scope while keeping over 99 percent of embedded emissions covered. The Commission later verified the threshold exempts only 0.87 percent of embedded emissions, under the 1 percent ceiling the regulation allows. Read that as a market-maker's gift: the long tail of tiny importers that would have been terrible software customers is gone by law, and the roughly 18,000 declarants who remain are, by construction, the ones moving real tonnage with real certificate liability. The addressable market got smaller and much richer in the same stroke.
The first declaration has a date, and the data is being created now. The annual CBAM declaration for 2026 imports, with the matching certificate surrender, is due September 30, 2027, after the simplification regulation extended the deadline from May 31. That sounds far away until you do the supply-chain math: installation-level 2026 data must be collected from suppliers, checked against the CBAM methodology, and verified by an accredited verifier before it can go into the declaration. Supplier data campaigns take six to nine months when the suppliers are cooperative. The 2026 import year is already three-quarters gone. Anyone starting the data campaign in Q1 2027 is filing on defaults.
Defaults just moved under everyone's feet. Implementing Regulation 2026/1740, published July 31, 2026, corrected the default emissions values retroactively to January 1, 2026, obliging every company that calculated with the old tables to recalculate. A regulation that rewrites your inputs seven months retroactively is the best possible advertisement for a platform that tracks methodology versions instead of spreadsheets. And the default markup ratchet, 10 percent now, 20 percent next year, 30 percent from 2028, means the cost of not having actual data compounds on a published schedule.
2026 understates everything, which is exactly why procurement decisions made now are dangerous. In 2026 importers surrender certificates for only 2.5 percent of embedded emissions, mirroring the free allocation EU producers still receive. The CBAM factor then climbs: 5 percent in 2027, 10 percent in 2028, 22.5 percent in 2029, 48.5 percent in 2030, 61 percent in 2031, 73.5 percent in 2032, 86 percent in 2033, 100 percent in 2034, per the phase-out schedule in the amended ETS Directive. KPMG's CBAM team puts the warning bluntly: a cost calculation built on 2026 will materially understate the exposure arriving at the end of the decade, precisely the period in which procurement and supplier decisions taken today will still be in force. The single steepest step is 2029 to 2030, a 26-point jump. Every supplier contract signed this year without an emissions-data clause is a liability signed at 2.5 percent prices and payable at 48.5 percent.
The scope is about to double. The Council adopted a general approach on June 12, 2026 to extend CBAM to roughly 180 downstream metal-intensive products, things like metal structures, machinery, and automotive components, from January 1, 2028, while the Parliament's environment committee favors going beyond 400 product codes, with the final list still under negotiation. The expansion's added trade is estimated at 50 to 70 billion euros a year, on top of the current 40 to 50 billion. As KPMG notes, companies that never considered themselves CBAM-relevant will likely receive their first data request from a customer rather than from a regulator. That is a second wave of first-time declarants arriving in 2028, and the platform that owns the first wave's workflow inherits the second.
The trade war is the marketing. Russia filed WTO dispute DS639 in May 2025; India, China, Brazil, South Africa, and Indonesia have lodged objections, and India has threatened retaliatory tariffs. In March 2026, Mozambique's Mozal aluminum smelter entered care and maintenance, partly squeezed by the carbon border tax, putting 4,000 direct and up to 20,000 indirect jobs at risk, with ODI estimating a 1.6 percent hit to Mozambique's GDP. When a regulation is visibly reordering trade flows and triggering WTO disputes, every board asks its trade team for a CBAM plan. The answer most teams have today is a spreadsheet and a consultant's PDF.
The Default-Value Spread: An Original Calculation
Run the arithmetic on the Commission's own first-week data. Week one of the definitive phase saw 1.66 million tonnes of CBAM goods validated, 98 percent of it iron and steel. Annualize that pace: about 86 million tonnes a year. Assume blast-furnace steel at 1.8 tonnes of CO2e per tonne of steel, a standard industry intensity, which gives roughly 155 million tonnes of CO2e. At the Q1 2026 certificate price of 75.36 euros and the full 100 percent CBAM factor, that is about 11.7 billion euros a year in certificate liability. Now check it against the independent estimate: trade analysis puts the iron and steel sector's annual CBAM costs at about 12 billion euros. Two completely different routes, first-week customs tonnage versus sector modeling, land within a few percent of each other. The assumptions are stated: first-week pace may be front-loaded, the intensity is a single representative figure, the price moves. But the convergence says the order of magnitude is right, and the order of magnitude is eleven figures.
At full phase-in, the first week's import pace implies roughly 11.7 billion euros a year in certificate liability, matching independent 12 billion euro estimates for iron and steel.
Now the spread that pays for the software. Take the German distributor importing 10,000 tonnes of steel sections. On punitive defaults at, say, 2.2 tCO2e per tonne versus verified actuals at 1.7, the gap is 0.5 tonnes per tonne of steel, or 5,000 tonnes of CO2e. At 80 euros a certificate that is 400,000 euros a year of avoidable liability at the full factor. In 2026, at the 2.5 percent factor, it is only 10,000 euros, which is why importers are sleepwalking. In 2030, at 48.5 percent, it is 194,000 euros a year for one mid-size importer, every year, for filing on defaults instead of collecting actuals. Add the default markup ratchet, 10 percent now rising to 30 percent, and the retroactive correction precedent, and the default path is the most expensive data strategy available, chosen by everyone who does nothing.
One more number, for the certificate desk. At the first-week pace, 2026's 2.5 percent factor implies roughly 5.9 million certificates to be surrendered for that tonnage, about 445 million euros at current prices. Certificate sales open February 1, 2027 on a central platform, pricing moves from quarterly to weekly, and submitted purchases cannot be amended. The importers buying those certificates are industrial distributors and stockholders, not commodity traders. Somebody will sell them timing discipline. It might as well be the platform that already holds their emissions data.
The Gap in the Market
| Player | What They Do | What's Missing |
|---|---|---|
| CarbonChain | $10M Series A (2023, USV/Voyager); Comply product with customs declaration validation and automated XML generation; customers include thyssenkrupp, Societe Generale, Rabobank, Gunvor. | Built for commodity traders, banks, and metals producers looking upstream. The authorized declarant is usually a mid-size EU importer looking downstream at uncooperative suppliers. Different buyer, different workflow, different data campaign. |
| carbmee | 20M euro Series A (CommerzVentures); enterprise carbon management with a CBAM module; ZF Group as a reference customer. | Enterprise suite sold to sustainability teams. Declarant liability sits with trade and customs teams, who do not log into the sustainability platform. CBAM as a module inside a suite is not a declarant workflow. |
| osapiens + ERM | Partnership announced September 2026; multi-regulation ESG platform covering CSRD, CBAM, CSDDD, and EUDR, paired with ERM advisory. | Breadth is the pitch and the problem: one platform for four regulations means CBAM declarant operations, verifier-ready evidence packs, certificate purchasing, are nobody's dedicated product. Advisory hours do not scale to 18,000 declarants. |
| Greenly and SMB carbon tools | CBAM solution launched February 2025; supplier risk identification, purchasing AI, alternative supplier suggestions. | Aimed at the SME long tail that the 50-tonne de minimis just exempted. Their target customer largely no longer has obligations. The remaining market is bigger importers with harder data problems. |
| Big Four and trade law firms | CBAM readiness assessments, methodology opinions, authorized declarant applications. | Hourly billing on a recurring data problem. A readiness assessment is a PDF; the declarant needs a machine that runs the supplier campaign every quarter and never lets a methodology version slip by unnoticed. |
| Customs brokers and forwarders | File entries, hold AEO status, some offer CBAM reporting as an add-on. | Brokers file what the importer gives them. They do not run installation-level data campaigns at Turkish steel mills, they do not want emissions liability, and their economics are per-entry, not per-tonne-of-CO2e. |
| Accredited verifiers (SGS, Bureau Veritas, TUV) | The mandatory third-party verification of emissions data in the definitive phase. | Verifiers audit; they do not collect. They are the bottleneck, not the pipeline: every declarant needs verifier-ready evidence packs, and nobody is manufacturing them at scale. The platform that feeds verifiers clean packs becomes their preferred channel, which is distribution disguised as compliance. |
| The Commission's own registry | Free, authoritative, integrated with customs and the Single Window. | A filing portal, not a workflow. It does not chase suppliers, track methodology versions, time certificate purchases, or warn you that your defaults were just corrected retroactively. Free is expensive when the declaration is wrong: penalties run 10 to 50 euros per tonne of unreported emissions. |
The pattern: the top of the market is served three times over by enterprise suites, consultants, and verifiers. The declarant, the party with the actual legal liability, is served zero times as a software product. And the de minimis did the segmentation work for free: the 18,000 who remain are concentrated, high-tonnage, and reachable through a few dozen trade associations and forwarder networks. This is the tariff-refund long tail in reverse. There, the wedge was 250,000 small importers the lawyers could not serve. Here, Brussels legislated the small importers away, and the wedge is the concentrated middle that everyone assumed the enterprise suites would cover, and did not.
The Solution
To be clear about what this page is: a free idea, not a company. Nobody paid for it, and Live in the Future is not building it. Take it. A declarant-side CBAM platform with five modules that follow the importer from scope check to certificate surrender:
1. Scope triage (free): The importer uploads a customs extract or connects its broker feed; the engine maps CN codes against the CBAM goods list, annualizes tonnage against the 50-tonne threshold per importer, flags the downstream-expansion codes that arrive in 2028, and produces a one-page scope opinion: in, out, or borderline, with the exact tonnage math shown. Free. This is the acquisition funnel, because the most common question in the market right now is "does this even apply to us," and the honest answer is sometimes no. Prospects told they are out of scope become referrers; prospects told they are in become customers with a deadline.
2. Supplier data campaigns (per-supplier fee): The core machine. For each installation in the supply chain, the platform generates a CBAM-methodology data request in the supplier's language, with worked examples for their production route, a secure upload portal that validates figures against plausible ranges on receipt, and an escalation sequence, nudges, buyer-side pressure templates, and finally a documented good-faith-effort record. That last piece matters: where actual data cannot be obtained despite documented effort, the file shows the work, which is what verifiers and, eventually, auditors ask for. Supplier figures are checked for the classic failure modes on arrival: plant-level numbers submitted where installation-level is required, corporate sustainability reports substituted for process data, precursor emissions omitted. Garbage is rejected at the upload, not discovered at verification.
3. Verification-ready packs (per-declaration fee): The campaign output is compiled into evidence packs structured the way accredited verifiers actually audit: installation data, methodology version pinned, precursor chain, carbon-price-paid documentation for the deduction, and a defaults-versus-actuals comparison showing the money saved by having real data. The platform maintains referral agreements with several verifiers and routes packs to whoever has capacity, which turns the industry's verification bottleneck into the platform's distribution channel. Methodology changes, like the retroactive default correction of July 2026, trigger automatic recalculation across every affected declaration, with the version pinned in the audit trail.
4. Certificate desk (subscription plus execution spread): The treasury module. It forecasts certificate need from the verified emissions data, tracks the ETS auction price that sets the certificate price, and proposes a purchase schedule across the quarterly, then weekly, pricing windows. Because submitted purchases cannot be amended and only one repurchase a year is allowed, the desk optimizes for the asymmetric risk: under-buying into a rising price versus over-buying into a falling one, with the declarant's cash position as a constraint. Execution happens on the central platform from February 1, 2027. The desk does not speculate; it buys the liability on a schedule, and charges for the discipline.
5. Expansion radar (included): The downstream-goods extension lands January 1, 2028 with 180 to 400-plus new CN codes still under negotiation. The radar maps each declarant's current and adjacent product classifications against the draft lists, flags customers who will receive their first supplier data request from downstream buyers, and pre-builds the scope triage for the second wave. When the final list drops, the platform's existing customers already know their 2028 exposure, and the radar becomes the acquisition funnel for wave two.
Revenue Model
| Revenue Stream | Amount | Notes |
|---|---|---|
| Platform subscription | 12,000 to 36,000 euros/yr | Tiered by declarant tonnage and installation count. The anchor product; priced against one avoided default-value filing, which it beats by an order of magnitude. |
| Supplier data campaigns | 150 to 400 euros per supplier per year | The wedge economics: a 40-supplier declarant pays roughly 8,000 to 12,000 euros a year to convert defaults into actuals, against six-figure default spreads at full factor. |
| Verification-ready packs | 2,500 to 8,000 euros per annual declaration | Charged per declaration filed; includes methodology version tracking and retroactive recalculation. Referral margin from verifier partners on top. |
| Certificate desk | 6,000 euros/yr plus execution spread | Subscription for the purchase schedule and price tracking; a small spread on executed volume once the central platform sales open in February 2027. |
| Expansion radar conversions | Platform tier upsell | Wave-two declarants from the 2028 downstream extension enter through the free radar triage and convert to full platform. Zero marginal acquisition cost. |
Unit economics at 900 declarants: 900 platform subscriptions at a 20,000-euro average is 18M euros in ARR. Supplier campaigns: 900 declarants times 25 suppliers average times 250 euros is 5.6M euros. Declaration packs: 900 times 4,000 euros is 3.6M euros. Certificate desk: 500 adopters times 6,000 euros is 3M euros plus execution spreads. Blended year-three revenue near 30M euros on a team of 40 to 50, with the platform and desk lines compounding as the CBAM factor ratchets from 5 percent toward 48.5 percent. The ratchet is the growth engine: every factor step multiplies the value of the data the platform already holds.
Market Size
TAM: Start with the Commission's post-de-minimis estimate of about 18,000 declarants remaining in scope. Assume a 20,000-euro average platform ARPU across tiers, a stated assumption: 18,000 times 20,000 is 360M euros a year in software, roughly $390M. Add the services layer, supplier campaigns, declaration packs, certificate desk, at roughly 1.5 times software, and the full-stack TAM is about 900M euros a year, call it $1B. The honest caveat: ARPU is modeled, not observed, and the declarant count is a Commission estimate that will move as the downstream extension lands.
SAM: The reachable five-year window: iron, steel, and aluminum declarants, the sectors holding 98 percent of current tonnage, perhaps 11,000 of the 18,000. At the same 20,000-euro ARPU: 11,000 times 20,000 is 220M euros a year in software, roughly $240M, acquirable through forwarder partnerships, steel and aluminum trade associations, and the free scope triage. Fertilizers and cement are the phase-two verticals; hydrogen and electricity are thin markets for now.
SOM (year 3): 900 declarants onboarded, about 8 percent of SAM's declarant count: 18M euros in platform ARR, 5.6M in campaigns, 3.6M in declaration packs, 3M in desk subscriptions, for about 30M euros in year-three revenue. Aggressive, but the acquisition funnel is the free scope triage answering the market's most-asked question, and every triaged declarant with a September 2027 deadline is a customer whose defaults we can still beat.
Startup Costs
| Category | Cost | Notes |
|---|---|---|
| MVP: declarant cockpit, supplier campaign engine, declaration prep (7 months) | $380K | 3 engineers. CN-code mapping, tonnage annualization, campaign workflows, methodology version tracking, registry data formats. |
| CBAM methodology lead + verifier partnerships | $140K | A practitioner who has filed transitional reports, not a generalist consultant. Verifier referral agreements are the distribution; negotiate them before launch. |
| Trade and customs counsel | $70K | Authorized declarant obligations, liability boundaries, opinion letters on the scope triage. The platform gives workflow, not legal advice; counsel draws the line. |
| Multilingual supplier outreach operations | $90K | Campaign staff covering Turkish, Chinese, Hindi, and Arabic supplier bases. The data is in those inboxes or it does not exist. |
| E&O and cyber insurance | $40K/yr | Table stakes when you touch client declarations and certificate purchasing. |
| Customer acquisition (forwarder partnerships, trade associations) | $90K | The scope triage is the funnel; forwarder and association partnerships are the distribution. |
| Founder salaries (year 1) | $260K | 2 founders, modest. The CEO should be able to sell to trade directors; the CTO builds the campaign engine. |
| Operating buffer | $60K | Hosting, registry interface costs, support line. |
| Total | $1.17M |
Risks and Challenges
The WTO could wound it. Russia has filed dispute DS639, and India, China, Brazil, South Africa, and Indonesia have lodged objections, with India threatening retaliation. A WTO ruling against CBAM's design would not kill the regulation overnight, the EU would redesign rather than repeal, but it could freeze or reshape obligations mid-build. Hedge: the supplier-data asset survives any redesign, because every version of a carbon border measure needs installation-level emissions data.
Brussels could simplify it again. The de minimis itself is proof the scope moves: 182,000 importers were in scope until they were not. Further simplification, a higher threshold, or a delayed downstream extension would shrink the customer base. The July 2026 Commission proposal to stretch the free-allocation phase-out from 2034 to 2038, flagged in the OECD's implementation review, would push the ratchet out and soften near-term urgency. Mitigation: price the platform on the data problem, which exists at any factor above zero, not on the factor itself.
Suppliers can simply refuse. Non-EU producers face no penalty for ignoring a data request, and some will treat emissions data as commercially sensitive or as leverage in price negotiations. The platform's answer is process, multilingual requests, buyer-pressure templates, documented good-faith effort, but a mill that will not answer is a mill that will not answer, and its buyer's declaration will carry defaults. Churn risk concentrates in declarants whose supply base is least cooperative, which is knowable from the campaign data and should shape ideal-customer profiling early.
The verifier bottleneck is real and outside your control. Accredited verifiers are few, their capacity is finite, and every declarant needs them before September 2027. If verification queues blow out, declarations slip through no fault of the platform's, and customers will blame the nearest vendor. The referral agreements and pack-quality reputation are the defense, but they do not create verifier capacity.
Liability cannot be outsourced, which caps what the product may promise. The authorized declarant remains legally responsible for the declaration no matter who prepared it. The platform must be engineered, messaged, and contracted as workflow and evidence, never as a liability shield. One marketing slide implying otherwise ends the company at the first audit.
Limitations
The 11.7-billion-euro cross-check is an original calculation, not a measured fact. It annualizes a single week's customs pace, assumes a uniform 1.8 tonnes of CO2e per tonne of steel, and holds the certificate price at 75.36 euros. First-week flows may be front-loaded; real intensities vary by production route; the price moves quarterly now and weekly from 2027. The convergence with the independent 12-billion estimate is encouraging, not confirmatory.
The 18,000-declarant figure is the Commission's estimate after the de minimis, and the downstream extension to 180-plus product codes is still under negotiation, with Parliament's environment committee pushing past 400. The addressable market could be 15,000 or 30,000 depending on where the final list lands. ARPU is modeled from adjacent compliance software, not observed in this market, because this market is eleven months old.
This analysis relies on trade-press reporting of Commission figures, first-week registry data as reported by industry press, and draft rules on certificate purchasing that are still under consultation. The immutable-purchase rule, the 5-cent fee, and the one-repurchase limit could change before February 2027. Nothing here is legal advice; the scope triage is a workflow that must be fenced by counsel's opinion letters.
Finally, the WTO dispute and the 2038 phase-out proposal are live political variables. A full invalidation is unlikely, the EU treats CBAM as a flagship, but the trajectory of the factor schedule, which is the business's growth engine, is subject to legislative revision.
Strongest Counterargument
This is a feature inside CarbonChain, carbmee, or osapiens, not a company. The 18,000 declarants are concentrated and already have customs brokers, enterprise carbon suites will add declarant workflows the moment the market looks real, the Big Four will sell it as advisory, and certificate purchasing is a treasury function that belongs in the ERP. Anyone starting today is building a point solution in a market where the incumbents have the customer relationships, the verifier partnerships, and ten times the capital. The honest end state is an acqui-hire by whoever owns the trade-compliance suite, and the window to matter is maybe 24 months.
There is real force in this, and the honest version of the business concedes the timeline pressure. The rebuttal is not that incumbents will ignore the market but that they are structurally mispositioned for it. Brokers file entries; they do not run supplier data campaigns at Turkish mills and will not take emissions liability at any price. Enterprise carbon suites sell to sustainability teams; declarant liability sits with trade and customs teams, who buy different software through different budgets and have never opened the sustainability platform. Advisory firms sell PDFs by the hour against a recurring quarterly data problem. None of them will build the multilingual supplier-chasing machine, because it is unglamorous, operational, and low-margin per unit, which is exactly why it is available. The default-value spread, the immutable certificate purchases, and the retroactive methodology corrections make this a data-quality business that compounds with the factor ratchet, not a reporting feature. And the 2028 downstream extension delivers a second wave of first-time declarants who have no incumbent relationship at all. The acquisition-by-incumbent end state is not a failure mode. It is a plausible exit at a multiple of the 30M-euro year-three revenue, provided the company owns the declarant workflow rather than renting a module inside someone else's suite.
What You Can Do
If you import CBAM goods above 50 tonnes a year: Annualize your 2026 tonnage by CN code this month and confirm your authorized declarant status. Then start the supplier data campaign now, not in Q1: installation-level figures, methodology version pinned, precursor emissions included. Every consignment received without evidence is a consignment you will file on defaults, and the markup on defaults rises to 20 percent next year. Get the carbon-price-paid documentation from your suppliers' jurisdictions too; the deduction is real money and it requires evidence.
If you supply EU importers: Your customers' declarants will come asking for installation-level emissions data, and the ones with a platform will ask in your language with worked examples. Answering with actuals instead of forcing them onto defaults is a competitive advantage measured in euros per tonne times the CBAM factor. Suppliers who provide verified data win share from suppliers who do not, once buyers can see the spread.
If you're a customs broker or forwarder: Your customers hold the liability and you hold the relationship. White-label a scope triage and let the free scan do the CBAM conversation your team cannot have at per-entry economics, because the forwarder who productizes the declarant workflow keeps the customer when the 2028 extension arrives, and the rest watch a software company intermediate the relationship.
If you're building this: Sign the verifier referral agreements before writing the campaign engine, because without verification capacity you have a dashboard, not a product. Hire the methodology lead who filed transitional reports, not a generalist. Pin every methodology version in the audit trail from day one, because July 2026 proved the Commission rewrites inputs retroactively. And build the "you are out of scope" path into the triage with the same care as the "you are in" path: the exempted importers are your referrers, and the 2028 wave is your second act.
The Bottom Line
Eleven-point-seven billion euros a year at full phase-in, implied by the first week's customs tonnage and confirmed within a few percent by independent sector estimates. Eighteen thousand declarants left holding the liability after Brussels exempted the other 182,000. Certificates near 80 euros a tonne, defaults priced punitively by statute with a markup that ratchets to 30 percent, purchases that cannot be amended once submitted, and a first declaration due September 30, 2027 for data that is being created right now, in inboxes in Turkish, Chinese, and Hindi, by suppliers with no legal duty to reply. The enterprise suites sell carbon accounting to sustainability teams. The liability sits with trade teams. A platform that triages scope for free, runs the supplier data campaigns nobody else will run, compiles verifier-ready packs, disciplines the certificate buying, and radars the 2028 expansion is not betting that climate regulation keeps expanding. It is betting that 18,000 declarants would rather pay 20,000 euros a year than file on defaults at 48.5 percent. The screen costs less than one default-value declaration. The spread is the business.
Related Startup Ideas
The Supreme Court Voided $166 Billion in Tariffs. 330,000 Importers Are Owed Refunds, and 6.1 Million Entries Already Failed. · 1.36 Billion Parcels, $47 Each: Duty Compliance for the Importers De Minimis Left Behind · Washington Collected $195 Billion in Tariffs Last Year. The Supreme Court Ordered $166 Billion Back. A 1934 Law Cuts Whatever Comes Next.