Some things you don’t want to start looking for once you really, really need them: Plumbers, life jackets, criminal defense lawyers, flashlights, restrooms — and fraud protection.
Think about it: If you’re an ecommerce merchant reeling from a fraud attack, you’re likely in no shape to carefully consider your fraud threats and costs and how you can best avoid both. The one thing you know is that you want it to stop — now.
But even if you’ve got all the time in the world (you don’t), how do you do it? How do you go about determining what kind of fraud protection is right for you — or even if you need fraud protection at all (you do)?
“There have been different challenges that merchants have faced in ecommerce, depending on the timing of where they are in the evolution of digital commerce and what steps they’ve taken to mitigate fraud risk” says Jeff Herrera, vice president of marketing and channel development at ecommerce consultancy Guidance. “You kind of sort through it that way.”
Herrera, who worked for Visa for eight years, including in payments, says the lens through which merchants should consider fraud management isn’t a very complex one. But it does require a lot of introspection, he says. A few key questions:
- What kind of fraud do you typically experience?
- What investments have you made to avoid fraud?
- Is your online fraud increasing, perhaps due to relatively recent in-store credit card chip requirements?
- Have you become too conservative in your fraud-fighting efforts?
Think of fraud protection as a math equation
From there, the search for a fraud protection partner becomes something of a math problem.
“There is the cost to manage fraud,” Herrera says, “and also, what is the actual cost of that fraud that currently exists with a merchant?”
And while the equation is simple, the variables can be more complicated than some merchants give them credit for. Sure, chargebacks and fraud costs in the form of shipping, taxes and penalties contribute to the cost of fraud. So do operational costs, the money merchants spend for fraud teams that scrutinize orders and research the true identity of buyers. And apropos to Herrera’s question about being too conservative in fighting fraud, false positives are yet another contributor to fraud costs that need to be considered. A false positive refers to legitimate orders that are mistakenly declined for fear of fraud.
As important as all these considerations are, the truth is they are often down the priority list for retailers juggling challenges involving marketing, merchandising, fulfillment, customer support and brick-and-mortar priorities.
“It gets buried in the CFO’s office,” Herrera says of fraud prevention. “It’s not a marketing play. It’s not a sales play. This is an overhead expense and cost that comes from the finance team, who is responsible for managing chargeback settlement disputes and keeping fraud in check. It’s not sexy.”
But how about cutting costs and increasing sales? That’s sexy. And getting fraud protection right can achieve both.
“It does have significant impact on sales and on customer experience,” Herrera says. “If it gets screwed up, and that’s happened quite a bit with some of these overzealous merchants that take a very conservative view and decide to decline. They’re turning away good clients and good customers and that ultimately hurts the brand, because these customers get angry and upset. They tell their friends and family that this is the worst business ever.”
And so, once you decide you do need some help with fraud, how do you go about figuring out who should help you?
There are all kinds of fraud management companies out there, but they are hardly the same. Some rely on static rules that require manual tuning to keep up with the rapidly changing tactics and tools that fraudsters use. Some provide a score on a sliding scale and leave it to the merchant to decide whether or not to ship an order. And there is a new model, the guaranteed fraud protection model, which relies on big data and machine learning to shift fraud liability from merchants to the fraud prevention provider.
Signifyd is of this breed, so maybe it’s no surprise that we come down in favor of the innovative model. But it’s not just Signifyd. Guaranteed fraud protection is on the cusp of mainstream adoption. Industry analysts have pointed to machine learning as the path to fraud prevention’s future. All of the well-known ecommerce platforms offer options for guaranteed protection.
Guaranteed fraud protection provides predictable costs
The big data and machine learning aspects of guaranteed fraud protection mean merchants can receive accurate decisions on orders at scale and during peak sales periods. The financial guarantee means that merchants — from small businesses to enterprise operations — have a predictable cost of fraud and no need to worry about the damage fraud losses can do.
Which leaves merchants to choose among the few guaranteed fraud protection solutions out there. One way to frame the decision is to consider what makes successfully offering a guaranteed model so hard and then considering which potential partner is likely to have best met the challenges.
By its nature, deploying big data and machine learning to solve a big problem requires a broad data set built over time.
For Ecommerce Fraud Management, that means accessing a database of retailers of different sizes, in different verticals, with different strategies and in different geographies. It means understanding consumers’ different buying habits and data footprints in different countries and regions of the world. It means understanding the proper mix of human and machine to make the right decisions on orders. In an era when fraudsters are constantly changing their tools and tactics, the importance of domain experts in designing models and identifying fraud is paramount.
Finally, getting guaranteed fraud protection right means having the ability to analyze false positives across thousands of merchants to ensure that models aren’t overly conservative when it comes to approving orders. Not only do overly conservative models mean that legitimate, valuable orders are never shipped, they result in legitimate would-be customers on the other end of those orders who’ve been made to feel like criminals.
So yeah, creating a clear vision of your fraud protection plan does take a lot of introspection. But as someone important in your life has no doubt told you more than once: When it comes to getting started, there is no time like the present.
Photo by iStock
Contact Mike Cassidy at firstname.lastname@example.org; follow him on Twitter at @mikecassidy.