ChargebackX Session Recap
Robbie MacDiarmid, VP, Consulting at CMSPI
Key Takeaways
- Every transaction requires consensus. A single โnoโ from any of the 6โ8 parties in the payments chain can kill a sale, regardless of the merchantโs decision.
- Overly cautious risk settings cost merchants more revenue through lost approvals than actual fraud losses.
- Your inputs shape issuer trust. Pre-authorization screening builds long-term โcredit healthโ with issuers and improves future approval rates.
- Measuring โapproval rateโ inconsistently across tools hides lost sales; the only true measure is your gross approval rate.
- AI will blur the line between customers and bots. As agentic commerce grows, merchants must evolve risk models to avoid blocking legitimate AI-driven purchases.
Introduction
Itโs one of the most common frustrations for any online merchant: your business is growing, your marketing is working, and customers want to buy your product, but you’re still losing sales. Legitimate customers are being declined, and it’s not always clear why. The assumption is that a payment is a simple exchange, but the reality is far more complex.
A single transaction is a journey through a supply chain where a cascade of decisions are made in milliseconds. At the recent inaugural ChargebackX conference, Robbie MacDiarmid, VP, Global Product Lead at CMSPI, presented a powerful framework for understanding this complexity.ย
He posits that every payment decision is a balance between three competing forces: Fraud, Cost, and Approvals. Merchants often focus intensely on the fraud lever, but MacDiarmidโs insights reveal how this negatively impacts the approvals lever, ultimately driving up the effective cost through lost sales.
Here are four surprising truths from his session that reveal how an unbalanced payments strategy is costing you.
#1 โ A Sale Requires Total Agreement, But a Decline Only Takes One “No”
For a transaction to be successful, every single partner in the payment chain must agree to it. This concept of “consensus” is fundamental. On average, a single transaction passes through the hands of six to eight different parties, from your payment gateway and fraud-screening tool to the card network and the customer’s issuing bank. Each one must give a “yes.”
โEvery approved transaction in payments is the result of consensus. Every partner in the chain has to say yes in order for that transaction to be approved. So it’s full agreement and alignment that this transaction is legitimate. The flip side of that is it only takes one no for that sale to vanish.โ
Read that again. It takes a โnoโ from just one party, out of the six to eight parties involved in the chain, to cancel the transaction.
This challenges the belief that the merchant has the final say. In reality, a merchant’s internal risk decision is just the first of many. As MacDiarmid notes, “The old adage you’re only as strong as your weakest link very much applies here. Although by weakest, we mean just whoever is the most risk averse in this case.”ย
The final, effective policy is often set by the partner most focused on fraud, who can single-handedly veto a sale and override your desired approval rate.
#2 โ The Hidden Cost of Caution is Higher Than the Cost of Fraud
Naturally, every merchant wants to eliminate fraud, but an overly cautious approach can be more damaging to the bottom line than the fraud it’s meant to prevent.ย
False declines (legitimate customers who are incorrectly identified as risky and rejected) are estimated to cost businesses over $300 billion globally. That figure is higher than the cost of actual fraud losses! Worse, when a good customer is rejected, they might retry the purchase, but more likely than not, theyโll go to a competitor, or simply never make that purchase again, damaging your brand and future revenue.
According to a survey by the Merchant Risk Council, false declines account for a staggering 2% to 10% of all attempted transactions on a merchant’s site. This is the direct result of over-optimizing for the fraud lever at the expense of the approvals lever. While no one would argue against the importance of fighting fraud, an overly aggressive strategy turns away good customers and ultimately loses more revenue than the fraud itself would have cost.
#3 โ Your Upstream Decisions Create Downstream Ripples
The choices a merchant makes before sending a transaction into the broader network significantly impact its final success rate. Issuing banks and card networks maintain a kind of “credit health” for each merchant ID (MID). Sending them cleaner, well-screened traffic builds trust and makes them less risk-averse toward your future transactions.
This concept is most tangible in the practice of “MID warming,” where new merchant IDs are slowly ramped up with low-volume, clean traffic to signal to issuers that they aren’t risky. The same principle applies to ongoing traffic. The data makes this clear: merchants who screen for fraud after authorization have, on average, a 5% lower authorization rate than those who use pre-authorization screening.
Itโs worth double clicking on the insight here. Implementing robust pre-authorization fraud checks isn’t just about managing fraud at your own gate; it’s about conditioning the entire supply chain to improve the long-term health of your approvals.
#4: You’re Probably Measuring “Approval Rate” Incorrectly
Donโt worry; itโs not your fault. The term “approval rate” sounds simple, but different partners in the supply chain define it differently, leading to significant internal confusion. To understand the disconnect, let’s follow a short narrative of 100 customer purchase attempts.
Step 1: The Issuerย
The journey begins. The customer’s bank, the first major gate, declines 9 attempts for various reasons. From the issuer’s perspective, your approval rate is a solid 91%. But 9 sales are already gone.
Step 2: The Fraud Toolย
Your post-authorization fraud tool now sees the 91 approved transactions. It flags and declines 3 more. The tool proudly reports a 96.7% approval rate, but that’s 96.7% of a much smaller pie.
Step 3: The Realityย
When the dust settles, only 88 of the original 100 customers have succeeded. Your True Gross Approval Rate is 88%. This is the only number that reflects your bottom line.
This discrepancy matters immensely. Your fraud team might celebrate a 96.7% rate, focused on its slice of the fraud pie, while your finance team struggles to understand why 12% of potential sales are vanishing. Aligning on a “True Gross Approval Rate” is essential for accurately balancing fraud prevention with overall approvals.
What Happens When You Can’t Tell a Customer From a Bot?
The payment ecosystem is a web of interconnected decisions. A merchant’s fraud strategy cannot operate in a silo; it must be holistic and account for the behavior of every partner in the chain. This has never been more critical as we enter the era of “Agentic Commerce,” where AI agents will automate purchases on behalf of consumers.ย
The immediate challenge is that these AI agents will be nearly indistinguishable from the malicious bots that fraud systems have been designed for decades to block. But the deeper threat MacDiarmid highlights is the potential loss of critical data. Fraud systems rely on rich user and transaction signals, but with an AI agent making a purchase, you might lose access to all of it. The system will be starved of the very information it needs to make an informed decision.
This leaves every merchant with a crucial question to ponder: In a world where AI agents are expected to drive a $1.7 trillion market by 2030, how will your business adapt its risk strategy when your best customer starts to look like your biggest threat?
Interested in speaking with chargeback experts about your risk strategy? Schedule a demo today.ย
FAQs
What does โconsensusโ mean in the payment chain?
It means every party involved, gateway, fraud tool, card network, and issuer, must agree to approve a transaction. A single โnoโ anywhere stops the sale, which makes alignment across partners crucial.
Why are false declines so costly?
Because they turn away legitimate customers who often donโt retry. Studies show false declines now cost more globally than actual fraud losses.
How can merchants improve their approval rates?
Start with upstream decisions using pre-authorization fraud screening, optimizing transaction data, and maintaining healthy merchant IDs, all of which build issuer confidence and increase long-term approval success.
What is the โTrue Gross Approval Rateโ?
Itโs the percentage of all customer purchase attempts that are successfully completed after every layer of screening. It reflects the real impact of fraud prevention policies on revenue.
How will AI and agentic commerce change risk management?
AI agents will soon make autonomous purchases, but theyโll resemble bots that fraud systems try to block. Merchants must adapt fraud models to recognize trusted AI agents while keeping real threats out.
