Mark Wagner believes the best chargeback recovery systems are automated and data-driven. He founded Disputifier, an Austin, Texas-based chargeback software company, on that premise in 2021.
He told me, “We’ve developed an intuitive system over the years. It combines data from the transaction with our testing and identifies an appropriate response.”
He and I recently discussed the state of ecommerce chargebacks and how merchants can recover false claims. The audio of our entire conversation is embedded below. The transcript is edited for length and clarity.
Eric Bandholz: Tell us what you do.
Mark Wagner: I run a software company called Disputifier. We’re an automated chargeback recovery agency. We see over 60% of chargebacks being fraud. These are not impossible to win. It’s more about separating the valid credit cards. Say a crook bought someone’s credit card info on the dark web. That’s a very different situation than a customer trying to get free stuff.
We help with duplicate chargebacks [where a cardholder wins a chargeback, then loses it, then refiles it], which are hard to prevent but easy to win. Duplicates are our highest win rate — around 90%. We attach screenshots of the checkout page and the purchase process for duplicate responses. We submit all the evidence to the card issuer after testing. We have a ton of data identifying the exact way to format a response, which can have a huge impact.
We present the evidence via PDFs. So, instead of using the Shopify Payment’s response, we built our own from scratch. We can highlight specific areas and make it almost like a lawsuit with different sections. We try to format it differently from Shopify.
Bandholz: Do real people at the issuing banks read the documents?
Wagner: Yes, the banks will print your chargeback response and throw it on someone’s desk. That person will manually flip through it and decide whether to side with the merchant when he or she has already agreed with the cardholder. So the formatting and images matter. We keep text to a minimum — two to three sentences. Folks are visual. It’s all in the format, the graphics, the images, and how it’s presented.
We’re software-based, meaning we programmatically ingest data from Shopify and other sources and then add those into our automated response. We manually review our responses to ensure they’re up to par and if we have any custom evidence, but typically over 90% of responses are unchanged from what our system generates.
Bandholz: Can’t you just use Shopify’s fraud analysis?
Wagner: Shopify’s fraud analysis is too basic and not always helpful. It might have 10 data points without explaining the reason for flagging a chargeback as low or high risk. For instance, Shopify might mark a chargeback as low risk even if the order was placed outside of North America and shipped to California. It doesn’t make sense. Conversely, many are flagged as high risk with no serious indicators. If you’re refunding those, then you’re losing money. We’ve run tests. Roughly 7% of Shopify’s medium-risk orders (and 35% of high-risk) turn into a chargeback. So the vast majority are legit buyers.
Bandholz: How much effort should merchants put into fighting chargebacks?
Wagner: It depends on your size, business model, and average order value. It becomes a necessary but labor-intensive process if we’re talking about higher average order values — hundreds to thousands of dollars. If your AOV is lower, you should not spend time on it.
When I ran ecommerce brands, we had an employee who would try to determine if an order was fraudulent. She’d call everyone in the office and say, “Guys, look at this.” End of the day, we still had a ton of chargebacks. It’s an imperfect process that is better not done by humans.
Bandholz: What’s Disputifier’s approach?
Wagner: We’ve developed an intuitive system over the years. It combines data from the transaction with our testing and identifies an appropriate response. It merges the two. It’s a customized response for every order but matches the template. That format has worked for us. It then goes through a manual review and gets submitted on a merchant’s behalf.
We make money by taking a percentage of orders we win.
When Shopify brands come to us, they’re winning around 25%. Our win rate is a bit over 50%, depending on the processor. Alternate payment methods seem to have a fair dispute process, whereas credit card issuers can be unpredictable.
Merchants should always require customers to agree to terms and conditions, including the refund policy, during the checkout. Customers cannot complete their order unless they click the box to agree. Sellers can then reference it if a customer falsely claims a refund. It significantly helps the win rate.
Again, this is for high AOV. I wouldn’t do it on low AOV. Plus, for very high orders — $5,000 or more — merchants should make an actual contract with the customer. This will help with a win, too. Never take a chance with a big purchase.
Merchants should test and determine what that winning response looks like. It’s tough for brands to figure out the entire chargeback process on their own. It’s murky. Every bank has slightly different rules.
Bandholz: Where can folks get your software?