Going to MRC next week to check out fraud tools? You need a checklist.
The fraud prevention industry is exploding with new solutions and fresh players. You need a methodical way to asses how suitable their solutions are for your business. Here are 6 essential questions for your checklist.
Business Integration: What is the full scope of integrating your solution into my business processes?
Evaluation: How do I evaluate your service? Can I A/B test it against my existing solution?
Getting better: How does your solution improve over time?
Chargeback guarantee? Great, but what happens when you decline good transactions?
Statistical models: What information do I get back on low scores/declines?
Friendly fraud: How do you help me with the various flavors of “friendly” fraud?
Why are these questions critical?
Of course you’re going to inquire about the technical aspects of integration. But bigger issues pop up when you examine how a new component fits into your business processes, and how it “talks” with your existing systems (CRM, accounting, CS). For example:
Does the solution handle all of your orders? (All channels? Devices? Payment methods? International customers? Phone orders?)
What’s the chargeback process? How does chargeback information feed back into the tool? Does it help you win representments?
When a customer calls CS about a declined order, what’s the process? What tools and information are provided to CS agents?
The true cost of choosing an inadequate solution is unbearable. You need to measure the shiny new tool’s performance on your transactions before you commit. You want to do it with as little investment as possible, so that you don’t get sucked in without a way to turn back.
Get details from the vendor on cheap and quick ways to test their product.
Fraudsters adapt and get better all the time. Any prevention solution that doesn’t continuously improve will ultimately become ineffective. Get nosy with the vendor: What features did they add in the last few releases? What new/improved variables have they made available for rules? When and how are statistical models refreshed? How many top-notch fraud analysts are looking at their system’s false positives and false negatives and fixing the leaks?
The industry has recently sprouted several players who are ready to take responsibility for your fraud. They make the decisions and they pay for chargebacks. It’s a great value proposition but there are some issues to discuss with these vendors: If they take the risk, what is their incentive to approve more of your transactions? What if your transactions are atypical or riskier than average? Is there a way to prevent insult to your returning or VIP customers? What do you do when a long-time customer calls you about a decline?
They’re all the rage now. Big Data geeks promise they’ll solve all your problems. In reality, they raise unique questions: When a long time customer gets declined with low score, will your CS agent know what to tell her? How often are models refreshed? Does your business have enough data for training an accurate model? What if your data is too diverse for a single model? Some of these issues are highly technical, so bring along your own geeks to make sure you’re getting good answers.
Friendly fraud is on the rise: online shoppers are abusing merchants without batting an eyelash. Their transactions are different from those of fraudsters, so it’s imperative that you ask the vendors if they are dealing with this challenge. It is no longer a marginal issue, and in some industries - a serious impediment to business.
Make the best checklist for your business
Many other questions did not make the cut. Notably, I left out some fundamental ones, that go back to your fraud prevention strategy: Outsource fraud or manage internally? Rules, statistical models or both? Buy or build? What about seller-side fraud? Hiring and training…?
It is a good idea to consider solutions in the context of a fraud prevention strategy that is tailored to your business. Otherwise, you risk asking the wrong questions or forgetting the really important ones. There are fantastic solutions out there, but ultimately it’s up to you to choose the right ones for your business.
If you haven’t yet done so, go and chart your strategy. Here are some tips. A robust strategy will inform the questions that you ask, and many other decisions down the line.