The discipline of fraud prevention has changed dramatically over the past five years and continues to evolve rapidly. Consequently, former truths about fraud prevention are increasingly becoming outdated myths. In this series of posts, we’re tackling the four most widely circulated myths. Check out the other three parts on fraud insurance, reason codes, and chargeback guarantees.
In this post, the myth we’ll tackle is that you must use manual reviews to maintain oversight and control. The implication is that fully automated decisions remove visibility and control. There are (at least) four reasons why fully automated solutions are superior to any solution that requires some portion of decisions to be manually reviewed.
Automation doesn’t replace the Fraud team; it amplifies their impact
Forter is a fully automated solution, and the single most frequent misconception we encounter is that automation will replace the Fraud team. In our experience, nothing could be further from the truth.
Too much of the work that Fraud teams complete today is reactive — reviewing individual transactions and making approve/decline judgments. Automation frees the Fraud team from that reactive role. AI and machine learning provide a more complete view of fraud, meaning your Fraud team can take a more proactive role, focusing on highlighting business trends, enabling new payment methods, identifying vectors for digital commerce growth and more.
But don’t just take our word – here’s what our customers had to say:
“We transformed our Fraud analysts into business analysts, and that team has accelerated our digital commerce growth. Rather than review transactions, we’ve been able to rapidly add more payment methods, draw more insights from data and expand into new markets.”
— Jim Gallagher, VP Customer CARE and Fraud, Nordstrom
“As a former Fraud Manager, my team and I were stuck in Groundhog Day, reviewing the same types of transactions again and again. Now, instead of combating loss, we’re accelerating growth—we’re an engine for the business.”
— Jenna Posner, Chief Digital Officer, Snipes
AI and machine learning are far more accurate (and secure) than manual reviews
Manual reviewers are conditioned to look for known patterns, but sophisticated (and successful) fraudsters are constantly changing tactics. That’s why the efficacy of manual reviews has continued to decline over time.
Manual reviews also introduce other potential issues. Humans bring bias to decision-making, and our analysis has shown that their decisions are inconsistent — two transactions with near-identical attributes can produce one approval and one decline. Not to mention that exposing customer data to employees is a risky privacy practice; the more people that touch customer data, the less secure it is.
That’s why the next generation of fraud prevention is built on AI and machine learning. AI and machine learning together can identify and stop known fraudsters, seeing patterns in data that humans cannot, surfacing — and ultimately stopping — previously unknown forms of fraud. At Forter, our Trust Platform uses patented machine learning to consider thousands of attributes for anyone’s online identity and pattern match against our dataset — all in fractions of a second. That depth delivers precision that is impossible for a manual reviewer to match.
Manual reviews don’t scale
Almost every merchant we work with has business fluctuations. That can be a holiday rush in the online travel and hospitality industries, Black Friday-to-Cyber Monday in digital goods, or flash sales and drops of limited-edition merchandise for apparel retailers. That’s why Forter’s Abuse Prevention solution can be so integral in assessing trustworthiness and abuse.
When you have a dependency on manual reviews, what do you do when you have a 40% increase in transaction volume for a short period of time? Internal teams are not staffed to handle these bursts, nor are solution providers. You can add contractors who lack the context to make decisions, let through more borderline transactions without review, or pile up a backlog for your full-time team to grind down. Neither are good options — that’s a high cost to absorb for perceived incremental visibility.
Of course, fully automated fraud prevention solutions completely remove this concern — they can process hundreds or even thousands of decisions per second and scale seamlessly to match business needs. For example, between Black Friday and Cyber Monday in 2021, Forter processed $4.7 billion in transaction value, approving 25 million interactions across 236 countries. That was 50% more value and volume than we processed in 2020, and execution was flawless.
Manual reviews impede value-add services (e.g. BOPIS, mobile checkout)
The global pandemic changed consumer expectations for digital commerce experiences. The prevalence of Buy Online Pickup In-Store (BOPIS) increased by more than 200%, and leading merchants introduced contactless ‘mobile checkout’ experiences in retail locations. The vast majority of these value-add services are dependent on instant assessments of trustworthiness.
What happens when a consumer completes a purchase online and shows up at a merchant to pick up their items, only to learn that their order has not yet been approved?
This happens all the time with other fraud vendors. Even those that tout machine learning still manually review up to 5% of total transactions to maintain reasonable approval and chargeback rates. Ask a vendor how many full-time or contract reviewers they employ to avoid surprises. If your business is using or intends to use BOPIS, BORIS, mobile checkout, or similar value-add services to stand out in the market, you should insist on a fully automated solution.
At Forter, we’re champions of the Fraud team — that’s precisely why our mission is to rescue them from reactively reviewing transactions all day. We enable the Fraud team to apply their talent and insight proactively so that they can elevate their careers.
Sure, we provide unmatched visibility into all transactions — through our UI, you can see high-level trends and individual transaction details. But transformation is achieved when you have visibility into new vectors for growth.