Your dispute rate is climbing, someone has quoted you $29 an alert, and every comparison page you open turns out to be published by the reseller selling the alerts. This ranks the real providers on what each one actually stops, what it only reports, and what it does to the ratio your acquirer is watching. Where a provider publishes pricing, it is here. Where a provider does not publish, that is said plainly.
There are 4 products here, not 1
Chargeback prevention is marketed as a category. It is really 4 separate jobs with 4 separate failure modes, and buying the wrong one is the standard mistake. A merchant buys alerts to fix a ratio problem that alerts only partly fix, or buys representment to stop disputes that have not been filed yet.
That split decides everything else. Alerts and deflection stop a dispute record from ever being created. Representment recovers cash after the chargeback has already been counted against you. A guarantee moves the loss off your books without moving your ratio at all.
If your processor has already paused payouts, none of this is your first move, because the first 14 days after a freeze run on a different clock.
- Prevention alerts. The issuer tells you a dispute has been raised and pauses it. You refund, and the chargeback never posts.
- Deflection. Your order data reaches the issuer and the cardholder's banking app before anyone disputes. Verifi calls this Order Insight, Ethoca calls it Consumer Clarity.
- Representment. The chargeback already posted, and someone builds the evidence file to fight it.
- A guarantee. A fraud platform approves the order and absorbs the loss if that order charges back.
Ethoca and Verifi own the pipes. Everyone else resells them.
There are only 2 real alert networks. Ethoca is owned by Mastercard, Verifi is owned by Visa, and every consumer facing brand in this market is buying from one or both and marking it up. That is the shape of the market, and it explains why 6 vendors quote 6 different prices for identical alerts.
Ethoca Alerts fires when a cardholder disputes with their bank. Ethoca's own FAQ says the merchant should refund or stop settlement with an objective of 24 hours, then record the outcome in the portal or through the API. It states that actioning alerts is at the sole discretion of the merchant, and that not actioning them results in a chargeback being processed by the card issuing bank.
Coverage is the part to interrogate, because it moves with the card network and your processor together. Chargeblast's published comparison puts Ethoca at roughly 95% coverage on Mastercard and roughly 30% on Visa, which is why nobody serious runs it alone. An alert reseller's own help docs put Visa coverage over 95% on a traditional processor but 80% to 90% on Stripe or Shopify Payments. Ask for a vendor's numbers on your processor, not in general.
Verifi runs 2 products worth knowing. CDRN gives you 72 hours to credit the cardholder on Visa and non-Visa disputes before a chargeback is filed, and Verifi describes the issuer network behind it as reaching Mastercard, Discover and American Express as well as Visa. Its seller page states directly that disputes resolved with CDRN or RDR are not eligible for future disputes, and that disputes resolved before a chargeback initiates do not count against Visa's dispute ratio.
American Express and Discover are still the hole in all of it, and the reason is administrative rather than technical. Both need an SE number to enroll a merchant in alerts, and that same documentation says Stripe and Shopify Payments do not hand the number over when asked, which puts expected Amex and Discover coverage at 0% on those platforms and 15% to 25% on a traditional processor.
RDR is not an alert. It is a standing instruction to refund.
Rapid Dispute Resolution is the product operators misbuy most often. It is not a message you respond to. Verifi calls it a customizable decision engine that instantly credits the cardholder before a chargeback is initiated, and says a case can be auto decisioned in as little as 1 second. It runs on rules you set for amount, reason code and issuer, with no case review. That is an accurate description of what you are agreeing to.
It is a good trade if your average order is small and you were going to refund anyway. It is a bad trade if you sell a $300 bundle and had evidence worth submitting, because RDR does not ask you. It also runs through your acquirer, which has to read the field Visa adds to the TC15 message to see a pre dispute was resolved this way, so whether you can have it is not entirely your decision.
A timing detail that matters if you run both. Chargeblast's documentation says the Ethoca feed front runs the RDR feed by about 7 days, because Ethoca sits at the consumer bank and RDR sits at the acquiring bank. Running them together without dedupe rules means paying on both feeds to resolve the same dispute.
What an alert removes from your ratio, and what survives it
This is the part the comparison pages skip, and it decides whether the spend was worth anything. A disputed transaction can generate 2 records: the dispute itself, filed as a TC15, and, where the issuer codes it as fraud, a separate TC40 fraud report. Refunding on an alert removes the first. It does not remove the second.
Visa's acquirer monitoring program counts both. The ratio is the count of TC40 fraud reports plus TC15 disputes over the count of settled transactions. Disputes resolved through RDR and CDRN come out of the dispute side of that numerator. The fraud report does not, because Visa confirmed in 2025 that TC40s resolved through RDR would still count, reversing an earlier position that they would be excluded. A merchant buying alerts to fix a fraud driven ratio is paying per alert for a number that keeps climbing. Read that next to the reason your chargeback ratio is reporting last month's ad spend, because the 2 lags compound and the spike lands after you have already scaled.
The mechanism that does remove a fraud record is Compelling Evidence 3.0, submitted through Order Insight, and it carries conditions most stores cannot meet. You need 2 prior undisputed transactions from the same cardholder, each between 120 and 365 days before the disputed one, and at least 2 matching data elements across all 3 transactions, where the matching set has to include either the IP address or the device ID. If you have never logged device IDs, you cannot qualify, and no vendor can retrofit that history for you.
For scale, the thresholds effective April 1, 2026 put a merchant in the excessive band at 1.5% with at least 1,500 combined fraud and dispute records in a month, and enforcement is reported at roughly $8 per record above the threshold. That 1.5% applies in the US, Canada, the EU and Asia Pacific, while CEMEA stays at 2.2%. Check your acquirer's current notice rather than any blog, because these numbers have already moved more than once.
Per alert pricing, and the cost nobody puts in the table
Alert pricing is published by resellers, not by the networks. Neither Verifi nor Ethoca posts a rate card, so any figure attributed to them is somebody's markup.
Now the number missing from every one of those tables. The alert fee is not the cost of the alert. The cost is the fee plus the entire order you just refunded. On a $90 order, a $29 alert saves you a dispute fee that typically runs $15 to $100, and costs you the whole $90 sale plus goods that already shipped. You did not save money. You bought a ratio point, which is a fair thing to buy when the ratio is what threatens the account.
Which makes the buying test simple. Alerts are worth it when your ratio is the thing about to shut you down. They are a poor trade at a stable 0.3% dispute rate, where you are handing back margin on orders that were never going to become chargebacks.
- Chargeflow publishes $29 per deflected chargeback, with a stated guarantee that you only pay for alerts that prevented a chargeback, and better pricing above 50 alerts a month.
- Chargeback.io publishes $29 for Ethoca and $15 each for RDR and CDRN, with no monthly fee, no setup fee and no minimum.
- Chargeblast bills per alert with no contract, but publishes no rate on its own site, so the number arrives at signup. The per alert figures that circulate for it come from competitors' comparison pages, not from a published price.
Representment, once the chargeback has already landed
Different job, different vendors, and the pricing models split cleanly enough to compare.
Then the test that separates the good ones. Ask who reviews the evidence before it is submitted, and ask to see a redacted submission for your worst reason code. A vendor who will not show you one is selling a template.
And do the arithmetic on percentage pricing before you sign. 25% of a recovery is cheap on a $60 order and expensive on a $600 one, because that percentage is not capped. On high ticket a flat per case fee usually beats a share of the win, which is why the vendors quoting flat rates rarely lead with it.
- Chargeflow. 25% of each recovered chargeback, billed only on recovery, with alerts at $29 and transaction screening at $0.20 a scan after the first 1,000 free, against a $0.40 list rate. Stack all 3 and you have 3 bills on 3 different units, not 1 success fee.
- Justt. Success based, and its FAQ says you only pay when you win a chargeback. It does not publish a rate. It pulls evidence by API, CSV or manual upload and builds a case specific file rather than sending a template.
- Chargebacks911. Publishes no pricing. Quote based, and the quote moves with your volume and what is bundled into it, so get in writing whether you are billed on the cases you lose.
- Midigator, acquired by Equifax in 2022 and sitting alongside Kount. Also unpublished, positioned at enterprise volume. The monthly minimums you will read about come from competitors' comparison pages rather than from Equifax, so treat them as unconfirmed until you hold a quote.
- Signifyd and Riskified. Not representment. A guarantee: a percentage of approved order value, nothing charged on declines, and they absorb the fraud loss. Neither publishes a rate, and both route pricing through a sales conversation.
The order to buy in
Run these in sequence. Skipping straight to alerts is how operators end up paying per alert for volume that the first 2 steps would have removed for nothing.
- Fix the billing descriptor. It is free, and it kills the disputes that start with a customer not recognizing the line on their statement.
- Turn on deflection. Order Insight on Visa and Consumer Clarity on Mastercard push your order detail into the banking app before a dispute exists. Both run through your acquirer or your processor's enrollment, so that is a phone call, not a purchase.
- Start logging device ID and IP address on every order today. Without them you can never qualify for the evidence rule that removes fraud records, and you cannot backfill history you never captured.
- Then buy alerts, sized to the ratio problem rather than to your dispute count.
- Then buy representment, on the reason codes where you actually hold evidence.
The account none of this protects
The honest limit of everything above: it protects your merchant account. It does nothing for the ad account, and for most operators reading this, the ad account is what stops the business when it goes down.
They are linked through billing. Disputes climb, the processor tightens, cards get declined or replaced, and payment failures on the advertising side get read as risk by a completely separate system. Ad accounts get disabled for unusual activity on billing signals far more often than on anything in the creative, and the timing looks identical to a chargeback spike because it shares the cause.
That is structural rather than a tooling problem, so be blunt about the trade. On an agency account issued off a portfolio spending around $50 million a month, you inherit that portfolio's payment history instead of building your own, and you give up direct control of the account to get it. Nobody can promise an account will not be banned, and you should walk away from anyone who says otherwise. What changes is the Friday afternoon it goes down: a replacement is a support request rather than a rebuild. The account tiers and their spend ceilings are worth reading before your ratio forces the decision, and the questions operators ask before they move cover the parts that actually matter.