Manual Review: The 5 Step Process of Detecting Fraud

Fraud is a prevalent issue that many businesses and merchants face today. According to the 2023 Annual Fraud Report from the Federal Trade Commission (FTC), losses increased by $1.2 billion, totaling a whopping $10 billion. It's worth noting that over 101,000 people fell victim to credit card fraud due to stolen information.

While there are systems in place that automate fraud detection, some can be flawed and require cases to be manually reviewed. Discover the importance of the manual review process, how it's done, and why businesses might still need it despite technological advancements.

‍

What is a manual review?

A manual review for fraud detection involves human analysts examining transactions, accounts, or activities that automated fraud systems have flagged as potentially fraudulent throughout the KYB and KYC processes. There are many types of transaction fraud, and a manual review aims to assess the flagged accounts and instances.

The reviewer can determine whether fraud has occurred or the activity is legitimate by applying human judgment, experience, and contextual understanding.

‍

Manual reviews vs. automated systems

It's no secret that automated detection tools use AI and machine learning to speed up fraud detection. While this can help, it's important to consider that each has its benefits when reviewing transactions. In fact, other businesses employ a hybrid approach to detecting fraud.

‍

The benefits of doing manual reviews

Manual review processes are time-intensive procedures conducted by a manual review team. It involves meticulously checking historical purchase records and finding discrepancies in payment behavior, which could mean only a small number of transactions or cases are reviewed. Manual checks can also be used to verify documents. Some benefits of opting for this approach include:

• Exercising human judgment
  Manual reviewers assess complex and nuanced situations by applying contextual understanding and years of experience to determine whether a transaction might be fraudulent.
• Manually checks false positives
  Manual checks help distinguish the difference between genuine and fraudulent activities. This can greatly help in reviewing who might be legitimate customers conducting these payments.
• Protects high-value transactions
  Conducting manual inspections means there is extra scrutiny for significant transactions. This can minimize the risk of substantial financial losses for high-value transactions.
• Provides a more personal customer interaction
  With a person working behind the scenes, this offers a personalized touch to verify suspicious activities. This can enhance customer service and satisfaction.

‍

The benefits of using automated tools

Machine learning and artificial intelligence aren't perfect, but they work with an ever-expanding knowledge base. This means that once the software is programmed with the right procedure, it can also continue to absorb new learnings the more cases it reviews. Let's look at some benefits of using a tool to automate the process.

• Enhanced efficiency
  Using the right tools to comb through large volumes of transactions means they can spot illicit activity or fraudulent orders for merchants and businesses quickly and continuously without fatigue.
• Cost-effective
  Since transaction monitoring software requires technology built into the system to detect fraud, it reduces the need for extensive human resources, which lowers operational costs.
• Easy to scale
  It easily scales to handle increasing transaction volumes, making it a suitable option for businesses of all sizes.
• Reduced false positives
  The right tool provides a uniform application of rules and criteria, ensuring consistent decision-making.
• Real-time detection
  Monitoring tools can run 24/7 and identify potentially fraudulent activities in real-time. This makes immediate responses possible.

‍

The process of a manual review

Manual fraud reviews are crucial for handling complex cases that require nuanced understanding, and contextual judgment that automated systems cannot provide. A review expert must go through several steps to determine whether or not the transaction makes sense.

1. Flagging suspicious activity

Organizations with fully manual processes will receive reports from fraud victims. This can include information from customers or online merchants.

When an organization uses a transaction monitoring tool, suspicious activity is automatically reviewed with algorithms and rules to flag transactions or activities that deviate from normal patterns or exhibit characteristics commonly associated with fraud.

‍

2. Data gathering

Flagged items are then placed in a queue for analysts to manually review. Once ready, analysts gather additional information.

Some key details that are considered include:

• Transaction history
• User behavior patterns
• Geographic location
• Any other relevant context that gives insight into the situation

‍

3. Data analysis

Analysts can then evaluate the flagged activity by considering various factors indicating whether the transaction is fraudulent. Some considerations include:

• Transaction patterns
  The reviewer looks for patterns or behaviors that are normally clues of fraud, such as multiple transactions in a short period of time, purchases from high-risk locations, or inconsistencies in the customer's provided information.
• Comparison against known customer data
  The flagged transaction should also be compared against compromised cards, blacklisted IP addresses, or previous fraud cases.
• Examining additional external data
  Analyzing public records like PEP and sanctions lists is crucial for AML screening processes. It ensures that customers are not high-risk individuals or possibly involved in illicit activities.

‍

4. Decision making

After reviewing the data and considering the context of the situation at hand, the review expert comes to a final decision on whether the activity is, in fact, fraudulent. Important

• Confirming detection of illicit activities
  The activity or incident has factual information that points to fraud. Appropriate actions are set into motion. This could be blocking the transaction, freezing the account in question, and reporting the incident to authorities.
• Dismissing detection of illicit activities
  When an account or activity is found to be legitimate and there isn't enough evidence to support that it is a fraud, no further action is needed.

‍

5. Documentation and reporting

Proper documentation and reporting are necessary for record-keeping, compliance, and AML audits. This means all findings, final decisions, and actions should be visible in the final report. This step is often used to refine algorithms.

‍

When should businesses do manual reviews for fraud prevention?

There are many reasons for an organization to use manual reviews in fraud detection. Despite advancements in automated fraud detection systems, there are instances where businesses should conduct manual reviews to prevent fraud. This is because they are valuable in spotting illicit activities based on these factors.

1. Uncovering complexity and nuance

Automated software can flag fraudulent activities based on rules and patterns. However, they might miss capturing the complexity and nuances of human insight. Human reviewers can assess fraud risk and suspicious activities with context and effectively make more informed decisions on some more complex cases.

‍

2. Cross-checking false positives

In the event of false positives, manually reviewed information can more effectively distinguish whether or not fraud and false alarms have occurred. This can minimize any inconvenience for legitimate customers.

‍

3. Adapting to evolving fraud techniques

While a good KYB tool should be able to provide you with the right features that monitor key metrics for fraudulent activities and detect illicit activities, relying solely on this can mean missing out on evolving fraud techniques. A quick manual review to supplement the process can help you identify new patterns and methods that aren't known to technology yet.

‍

4. Improved customer service

Conducting a manual review for crucial cases can greatly enhance customer service with a human touch. This can prove useful when customers need to be contacted for additional verification or if further explanation is needed as to why a transaction was flagged.

‍

5. Low cost for smaller businesses

Most merchants and smaller businesses can benefit from working with manual checks, considering they may not have too many cases to to review. This means they don't need to invest much into a sophisticated system.

‍

Strengthening your fraud detection process

While automated tools should be able to handle the bulk of fraud detection due to their efficiency and scalability, manual reviews are just as critical to a comprehensive fraud prevention strategy. They provide oversight, expertise, and adaptability to address automation's limitations. Additionally, they are a great alternative to fraud prevention and detection for smaller businesses.

If you're looking for a KYB verification service to complement your manual review process, Vespia offers a full suite of features that scales with your business. Schedule a demo to learn more. 

‍