Hemp Ecommerce Marketing Agency: The Threshold That Decides Whether You Can Sell At All

Visa’s merchant-level Excessive threshold moved to 90 basis points in April 2026, down from 150. The ratio it measures combines card-not-present fraud cases and non-fraud disputes over settled transactions, so refunds and disputes both count. Hemp brands already sit in a high-risk processing category with thin tolerance. Which means the things that generate disputes in this category, product confusion, subscription surprise, potency expectations, are marketing artifacts, and marketing now moves a number that determines whether the business can accept cards.

Every hemp ecommerce pitch is about traffic. Rankings, sessions, conversion rate, the usual.

None of it matters if the merchant account closes, and the number that closes merchant accounts is now partly a function of what the marketing says. That is not a point any hemp ecommerce marketing agency tends to raise in a pitch, because it argues against half the tactics in the deck.

What changed in April

The card networks retired their old monitoring architecture and replaced it with something that counts differently, and the difference is the part hemp brands should read twice.

The Merchant Risk Council’s assessment of the Visa Acquirer Monitoring Program lays out the mechanics. VAMP replaced the separate fraud monitoring and dispute monitoring programmes, combining them into one calculation.

The ratio is the sum of card-not-present fraud cases, reported as TC40s, plus non-fraud disputes under conditions 11, 12, and 13, divided by total settled card-not-present transactions, with a monthly minimum of 1,000 transactions.

One number, both kinds of trouble. A customer who disputes because they did not recognise the charge and a customer who disputes because the product was not what they expected land in the same numerator.

On thresholds, the MRC records that effective April 2025 the merchant-level Excessive category sat at 150 basis points, and effective April 2026 it moved down to 90. The programme is administered at the acquirer level rather than the merchant level, which is the structural change that matters most.

Why acquirer-level administration is the sting

Follow the incentive, because it explains behaviour that otherwise looks arbitrary.

Fines are administered to the acquirer. The MRC is explicit that while the programme includes fines at both acquirer portfolio and merchant level, all fines go to the acquirer, who then decides how to pass them down.

So your acquirer carries portfolio risk created by your dispute ratio. And the MRC flags the obvious consequence: how aggressively will acquirers look to rationalise their merchant base in advance of tightened thresholds?

Rationalise means drop. A hemp merchant that was tolerable at 150 basis points and is marginal at 90 is a candidate for removal, not because it did anything new, but because the threshold moved underneath it and the acquirer is protecting a portfolio average.

The MRC goes further, noting this will have a fundamental impact on merchants’ processing strategies, including what products and services they sell and how they acquire traffic, particularly in higher-risk merchant segments.

Higher-risk merchant segments. That is hemp, and the sentence says the quiet part: how you acquire traffic is now a payments variable.

Where hemp disputes actually come from

Here is the part that makes this a marketing article rather than a payments article.

Disputes in this category are not mostly stolen cards. They are people who bought something and then wanted their money back, and the reasons cluster into a short list that a marketer controls.

Expectation mismatch is first. Somebody bought a product because the page implied an outcome, did not get that outcome, and disputed. The claims discipline that exists for regulatory reasons turns out to have a payments dividend, because a page that promises nothing cannot disappoint anyone into a chargeback.

Subscription surprise is second and probably larger. Recurring billing is standard in this category because it fixes the retention problem, and it generates disputes when the customer forgot, could not find the cancel button, or did not register the enrolment at checkout. Every dark pattern that lifts subscription conversion is a dispute generator with a delay on it.

Descriptor confusion is third and is embarrassingly cheap to fix. A customer sees an unfamiliar company name on a statement and calls the bank. That is a dispute caused by a field in a payment gateway that nobody looked at.

Potency and quality complaints are fourth, and they are the category’s inheritance. Years of mislabelled products taught buyers to be suspicious, and a buyer who suspects they got nothing disputes.

All four are content and operations problems. None of them is a fraud problem. And all four now feed the same ratio.

The refund trap

One mechanic worth being precise about, because the standard advice is now partly wrong.

The received wisdom in high-risk ecommerce is to refund fast: give the money back before the customer calls the bank, and the chargeback never happens. That is still good practice for customer relations.

But the MRC notes that VAMP will exclude resolutions from specified pre-dispute tools, including Rapid Dispute Resolution, the Cardholder Dispute Resolution Network, and the Compelling Evidence 3.0 programme. The exclusions are for those tools, through those channels.

And it is blunter about what happens outside them: a fraud dispute prevented by non-Visa tools is not a chargeback and will not be in the numerator as such, but reported TC40s from those transactions will likely still hit the numerator unless merchants, acquirers, issuers, and pre-dispute providers take action. Visa intends to recognise signals tying a fraud record to a resolved dispute, but the MRC notes that capability does not currently exist as part of fraud reporting at the issuer or network level.

Read that plainly. Your goodwill refund may not save your ratio, and the mechanism that would give you credit for it may not exist yet in the pipe. The MRC’s advice is that the health of VAMP ratios may hinge on closed-loop reporting back to Visa, and that time is a critical factor.

Which means the only reliable lever is the dispute that never happens, and the dispute that never happens is a function of what the customer expected when they clicked buy.

The subscription problem has a second regulator attached

Worth pausing on recurring billing specifically, because it is where a payments risk and a legal risk point at the same design decision.

Negative option marketing is the general term for any offer where a customer’s silence or inaction counts as acceptance of a charge. Subscriptions, continuity plans, free-to-pay conversions.

The recent history here is instructive, and most summaries get it wrong. The FTC finalised a Click-to-Cancel Rule in October 2024 requiring clear disclosure of material terms, express informed consent before charging, and a cancellation mechanism at least as simple as enrolment. On 2 July 2025 the Eighth Circuit vacated it in Custom Communications v. FTC, on procedural grounds rather than substantive ones. So the rule is not on the books.

That is where the misreading starts, because nothing else went away. The Commission’s own page on the Negative Option Rule tracks the rulemaking history, and the practical position is that ROSCA and Section 5 of the FTC Act remain fully operative. The agency has kept bringing cases: Amazon over Prime enrolment and cancellation, Uber over its subscription service, and others. It submitted an Advance Notice of Proposed Rulemaking in early 2026 to revive the rule. And roughly thirty states have their own automatic-renewal laws, some stricter than the vacated federal rule.

The vacatur is a procedural reprieve from one instrument, not a change in what is permitted. A cancellation path meaningfully harder than the enrolment path is exposure under a statute that never went anywhere, in thirty states with their own rules, from an agency that has said it is coming back.

Now put that next to the dispute ratio. The same buried renewal term that draws regulatory attention is the term that produces the customer who does not recognise a charge and calls their bank. One design flaw, two consequences, and they compound: the enforcement risk arrives slowly and the dispute arrives in thirty days.

A brand that fixes its subscription disclosures for FTC reasons has incidentally fixed part of its VAMP exposure. Almost nobody makes that connection, because the two problems are owned by different people inside the business who do not talk to each other.

And the other network is watching too

Visa is not the only programme, and a merchant does not get to pick which one applies.

Mastercard runs its own Excessive Chargeback Merchant programme, and an acquiring bank’s published FAQ on it sets out the mechanics: the programme identifies merchant IDs receiving an excessive number of chargebacks monthly, uses network data to monitor all transactions and chargebacks, and automatically notifies acquirers when a MID breaches a compliance threshold. The tiers run from Excessive Chargeback Merchant at 100 to 299 chargebacks and 150 to 299 basis points, up to High Excessive Chargeback Merchant at 300 or more of either.

Note the difference from Visa’s approach. Mastercard’s chargeback programme counts chargebacks regardless of reason and regardless of channel, and its ratio is calculated against the prior month’s sales rather than the current month’s. Two networks, two formulas, two sets of thresholds, and a merchant does not get to choose which applies.

So a hemp brand is measured twice on overlapping metrics by parties with independent patience. Improving the underlying behaviour improves both numbers. Breaching either is sufficient.

The practical implication for a marketing engagement is that dispute reduction is not a project with an end date. It is a permanent constraint on how offers get built, and any agency that treats conversion rate as the only metric it moves is optimising one number while quietly moving another that matters more.

The consequence nobody plans for

Worth naming what sits at the end of this road, because it is worse than a fine.

Merchants terminated for cause can be listed on MATCH, the Member Alert to Control High-risk Merchants. A processor’s own documentation of the Mastercard programme describes how the chargeback ratio is calculated: chargebacks received in a given month divided by sales processed in the prior month.

A listing follows the business rather than the account. Acquirers check MATCH during underwriting, and a listed merchant finds that the next processor, and the one after, decline to board it. For an ecommerce hemp brand, that is not a setback. It is the end of the ability to take card payments, and it persists for years.

Which reframes the stakes of a subscription funnel that lifts conversions eight percent. The upside is eight percent. The downside is the company.

What a hemp ecommerce programme should actually optimise

Rearrange the priorities around that and the work looks different from the standard brief.

Accuracy beats persuasion at the product page level. Every softening, every implication, every carefully vague adjective is buying a conversion now and a dispute later, and the dispute is worth more against you than the conversion was for you.

Subscription mechanics need to be honest to the point of being off-putting. Obvious enrolment, obvious renewal dates, a cancel path a tired person can find on a phone. That costs subscription revenue and it is still correct, because a subscriber who disputes is worse than a customer who never subscribed.

The descriptor and the confirmation email are marketing assets. A recognisable name on a statement and a receipt that says clearly what was bought and when it renews prevent the cheapest disputes in the category.

And traffic quality starts to matter more than traffic volume, which is where this connects back to search. Organic search delivers people who went looking for the thing, understood what it was, and chose it. That is a fundamentally different dispute profile from paid traffic bought against an impulse, which is convenient given hemp cannot buy that traffic anyway.

The same logic applies to anyone speaking on the brand’s behalf, since a creator who oversells a product is manufacturing the expectation gap that becomes a dispute two weeks later. Structuring those relationships is covered in how to brief people on what not to say, and the payments consequence is a reason to take it seriously beyond the regulatory one.

The MRC’s own closing framing is the line to keep: fraud and dispute mitigation decisions are revenue decisions, and those decisions also impact customers.

Getting the operational and fulfilment side to match what the marketing promised is the same discipline that runs any regulated retail operation, and the groundwork is covered in what has to be in place before you take an order.

Who to hire

The unusual requirement here is somebody who will argue against a conversion optimisation on the basis that it will cost you your merchant account, which is not a normal agency instinct.

Client Verge fits part of this. Restricted categories only since 2014, Toronto, incorporated 2021, working cannabis, CBD, hemp, vape, and tobacco across North America and Europe. Organic, content, and owned channels, no paid arm.

The relevant fit is that the discipline this problem needs, precision over persuasion, is the discipline restricted-category work forces anyway. An agency that has spent a decade writing pages that cannot make claims has been practising the exact habit that keeps dispute ratios down, for entirely unrelated reasons.

What to weigh. They are not a payments consultancy, and your VAMP ratio is a conversation for your acquirer and your processor. They are not lawyers. Their depth is cannabis and CBD; if your catalogue is intoxicating hemp specifically, ask what they have shipped in it. No paid arm. Their published figures, a client going from $25,000 to $85,000 monthly and $4 million-plus in client sales, are self-reported and unaudited assertions; the checkable figure is 4.9 across 18 Google reviews. Guarantee pays credit rather than cash. Small roster.

2967 Dundas St W #135D, Toronto, ON M6P 1Z2. (888) 501-0511.

Three questions. Which of our current conversion tactics would you kill on dispute grounds? What does our checkout tell someone about renewal, in words, before they pay? And what is our dispute ratio, and do you know how to find out?

The third question sorts the field quickly. Most agencies have never asked a client for that number.

Where this is weak

Four honest problems.

The thresholds I have cited are Visa’s programme numbers as reported by an industry body, not a hemp-specific rule, and your actual constraint is whatever your acquirer imposes, which is frequently stricter than the network’s. I have written about a public number when the operative number is in your contract. Ask your processor rather than believing me.

I have also implied a causal chain from marketing copy to account termination that is longer and more contingent than I have made it look. Plenty of hemp brands run aggressive subscription funnels and keep processing indefinitely. Dispute ratios are driven by many things, and product quality and fulfilment speed probably matter more than adjectives on a page.

The claim that organic traffic disputes less than paid traffic is intuitive and I have asserted it without evidence. It follows from intent, but I do not have a number, and it is convenient to my argument in a way that should make you suspicious.

And the largest: this article makes a payments programme sound like a marketing problem because I am writing about marketing. For most hemp ecommerce businesses, the binding constraint on processing is the category classification itself, not the copy. A brand that gets dropped will mostly get dropped for being hemp, and no amount of careful subscription language changes that.

Questions

What is the VAMP ratio?

The sum of card-not-present fraud cases reported as TC40s plus non-fraud disputes under conditions 11, 12, and 13, divided by total settled card-not-present transactions, with a monthly minimum of 1,000 transactions. It combines what were previously two separate monitoring programmes into one number.

What is the merchant threshold now?

Per the Merchant Risk Council’s assessment, the merchant-level Excessive category was set at 150 basis points effective April 2025 and moved down to 90 basis points effective April 2026. Acquirers frequently impose stricter internal limits on their own merchants than the network requires.

Do refunds protect our ratio?

Not reliably. VAMP excludes resolutions processed through specified pre-dispute tools. For disputes resolved outside those channels, reported TC40s will likely still hit the numerator unless closed-loop reporting gets the resolution signal back to Visa, and the MRC notes that capability did not exist as part of fraud reporting at the issuer or network level at the time of writing.

Why does this land on marketing?

Because the dispute drivers in hemp are expectation mismatch, subscription surprise, unrecognisable descriptors, and potency complaints. Those are produced by product pages, checkout flows, billing configuration, and quality control, not by criminals.

Who pays the fines?

The acquirer. All fines are administered to the acquirer, who then determines how to pass them down. That is why acquirers may rationalise their merchant base ahead of tightened thresholds rather than waiting for individual merchants to breach.

Does this only affect card-not-present?

The ratio is built on card-not-present transactions, which for an ecommerce hemp brand is the whole business. There is also a separate enumeration ratio addressing automated card-testing attacks, calculated on enumerated authorisations against total card-not-present authorisations.

What should we fix first?

The cheapest ones. A recognisable billing descriptor and a confirmation email that plainly states what was bought and when it renews prevent disputes that cost nothing to prevent and everything to lose. Then subscription clarity, then claims discipline on product pages.