The finance world was left reeling after federal regulators revealed that Wells Fargo employees had created millions of unauthorised credit card and bank accounts (without their customers knowledge or permission) since 2011.
- 5 million deposit accounts were opened without authorisation
- 565, 443 credit card applications were submitted without customer’s consent or knowledge
- 14,000 of those fraudulent credit cards earned over $400,000 in annual fees and interest charges
- 5,300 employees are said to have been fired because of ‘shady behaviour’
A toxic culture of unrealistic sales goals and relentless pressure – alongside Wells Fargo’s infamous strategy of cross-selling eight products per customer – are to blame for the widespread misconduct. But even if human error and lack of judgment were the underlying culprits, could technology have still prevented the scandal?
The right technology may not have prevented widespread ethical failure, but it would have stopped employees from being able to open millions of fraudulent accounts. Here’s how:
1. Fraud analytics
Analytics software can be used to detect fraud across multiple channels and transactions, combining real-time data with historical analysis to identify unusual behaviours and flag anomalies such as:
- Fake email addresses like email@example.com or phone numbers such as 123456789
- Accounts that have been inactive for substantial periods of time
Dormant accounts, consistently low balances and unusual contact information are all potential signs of an account that has been opened without the customer’s consent. But this information doesn’t necessarily point to an internal failure; it could suggest that a malicious external party is committing fraud too. That’s why fraud analytics alone is not enough to mitigate conduct risk. Other tools, like data monitoring, are needed too.
2. Data monitoring
Data monitoring is a process in which critical business data – such as customer contact information – is checked against a set of rules and controls established to create consistency in formatting. It ensures that an organisation’s data maintains a high standard of quality, and highlights anomalies that may indicate fraud.
Using data monitoring software would have helped Wells Fargo implement data quality controls and:
- Identify bank accounts and credit card applications that don’t match the attributes of a customer’s existing profile
- Track individual employee access to different data types across applications
- Map the data entry and collection process to track consent and permissions
With this kind of technology, Wells Fargo could have identified fraudulent accounts – and their creators – before the issue became systemic.
3. Artificial intelligence, big data and machine learning
Fraud analytics and data monitoring can help automatically identify anomalies in customer profiles, but they’re not enough to mitigate risk on their own. In some cases, a customer’s email address might really be firstname.lastname@example.org. They may even have multiple email addresses, or have new attributes that wouldn’t match their typical profile.
This is where Artificial Intelligence (AI) technology comes in. Machine learning software drawing on big data libraries can understand and learn things about a customer, such as:
- How many different email and postal addresses they’re likely to use;
- How often they’re likely to access certain account types;
- Where and when they’re likely to open an account; and
- How many accounts they’re likely to have at any one time
When something doesn’t fit within the parameters of expected behaviour, an AI program can alert management and trigger a response – such as an account freeze, or an investigation into the account registration process. Had Wells Fargo been using AI to audit and monitor customer data, the bank could have addressed systemic fraud earlier on.
4. Call recording and unified communications
Many Wells Fargo customers weren’t aware that accounts had been created without their consent or knowledge, but others were. CNN reports that some customers had reported and closed fake accounts more than a year ago, and subsequently changed banks.
It’s likely that many Wells Fargo customers would have contacted the bank to enquire about the mysterious accounts, file complaints and request closures. Call recording service, such as Cognia, and a unified communications platform could have flagged these complaints and identified consistent themes in customer engagement.
More specifically, call recording would have helped Wells Fargo:
- Record and analyse customer interactions on helplines, social media, SMS, online and more;
- Search and investigate interactions using analytics and automated surveillance features
- Alert and respond to high volume queries and complaints
Using other communications technology, such as natural language processing, would have been helpful too. ‘Cognitive technologies’ can be used to perform semantic analysis of customer interactions, spotting common themes in conversations – such as recurrent complaints of unauthorised or fake accounts.
Technology puts customers first
The scandal has cost Wells Fargo $190 million in fines and refunds, in addition to mounting legal costs, reputational damage and lost business.
But this is almost spare change to a bank with a valuation that exceeds $250 billion. It’s a relatively small price to pay for the damage inflicted on the financial positions and credit scores of customers.
Technology would have helped Wells Fargo prevent systemic fraud, and it would have protected customers and their interests too. Customers are better protected when the right technologies and processes are put in place, such as:
- Automated fraud monitoring and detection;
- Fraud tests that are replicated and performed consistently across full data sets, not just individual random transactions; and
- Early warning systems that alert to suspicious behaviour before it evolves into systemic fraud.
Implementing analytics software, call recording and AI technologies would have enabled Wells Fargo to do all this and more, protecting their customers from widespread ethical failure and conduct risk.