5 ways to speed up fraud detection in invoice finance
Invoice finance?fraud?is when a client receives undue cash from an asset-based lender through false invoices, diverted receipts, hidden disputes, re-aged invoices and collusion. This is different from ‘invoice fraud’ where criminals pretend to be suppliers and send fake invoices.
On the one hand, your clients are doing more business and have more debt to borrow against – great! On the other hand, higher invoice volumes and the apparent legitimacy of increased trading make it easier for fraudulent activity to slip through the cracks – not so great.
Even though lenders will eventually detect invoice finance fraud, the sooner they can identify it, the easier it is to reduce their exposure to it. The longer it takes, the less likely they are to recover loses.
Why fraud discovery takes so long
Most lender operations still rely on people to navigate a complex web of spreadsheets and disparate systems to review and validate invoices, check credit notes and reconcile ledgers.
The more time it takes your people to assess risk, the more time it leaves for desperate business owners to secure funding without getting spotted.
When operations rely on time-consuming manual processes, it’s hard to track a single client’s amalgamated risk exposure. The result is that far too often fraud is only discovered months after the invoice has been approved.
5 better ways to spot invoice finance fraud
To detect fraud more rapidly, the entire operation needs to move faster. That is, people need simplified workflows that make it easier to work across lending, banking and accounting systems.
Automations or ‘bots’ can be deployed at various, key stages of the process to handle high-volume, routine, repeatable tasks that take a lot of time. This allows lenders to process more documents, faster and more accurately without adding staff overheads.
Automation accelerates fraud detection in five ways.
1. Simpler review processes (invoice factoring)
If your people spot fraud by manually reviewing every bit of data and painstakingly trawl through invoices, credit notes, receipts and borrower information to look for red flags and inconsistencies, there will always be errors.
This process can be greatly simplified with automated exception management.
Exception management simply means spotting trends that deviate from the norm. In this model bots run automated background checks to review the bulk of the data. The automation can be programmed to look for unusual patterns across:
This means false invoices (for which no goods or services were delivered), re-aged invoices or otherwise tampered with documents can be spotter faster. Exception management saves time and reduces the rate of errors.
Importantly, it’s only when these bots detect something suspicious or anomalous that they bring an invoice or a debtor to an employee’s attention. So, human intervention is saved only for when it is really needed.
2. Increased visibility (invoice factoring)
Diverted receipts, where the client collects the payment but does not pass it back to the lender is still one of the most common frauds in invoice factoring. It’s also hard to spot.
Often, it’s only identified when the outstanding invoice gets old or the client’s day sales outstanding (debt turn) extends way beyond terms.
Automation can help detect these kinds of issues a lot faster than most invoice finance businesses do today.
Bots can be used to check client’s banking activity and raise alarms when funds are being banked and not transferred to the lender within agreed timeframes. They can even be programmed to spot faulty credit and returns notes by looking out for identical sums remaining in client accounts.
领英推荐
3. Improved verification (invoice factoring and discounting)
Whether you’re checking the legitimacy of potential new borrowers, invoices or creditors, verification is a crucial but time-consuming part of invoice factoring.
The simple act of contacting end customers to check if goods and services have been received is a cumbersome multi-step process. Your people have to review invoices and creditor records for the relevant information, send it to clients and chase them up if they don’t reply promptly.
When you have hundreds of invoices a day to verify, this process takes a significant chunk of time and makes it easier for fraud to slip through in those unanswered mails and missed calls.
Bots can be set up to automatically run?the whole verification process.
They can find documents, extract the data and put it in an email. They can send reminders, chase for information and process documents they receive from the end customer.
Any exceptions and particularly suspicious activity detected by the bots can be automatically escalated to be handled by a person. The whole verification process is smoother for your people who only deal with the red flags and suspicious activity.
Bots can also expedite and improve new borrower verification.
Instead of your people manually researching company directors, bots can automatically and quickly scan companies house data and link with other data sources like Experian to look for suspicious patterns in director names and the percentage of credit notes issued by a given debtor.
4. Faster ledger reconciliation (invoice discounting)
Typically it takes a client two weeks to submit their sales ledger to a lender. It then takes the lender another ten days to reconcile with their records. The result is lenders detecting mismatch and potential fraud across ledgers at the end of the month when it is harder to recoup what was lost.
With the intelligent use of automation, lenders can set up new workflows where the client’s ledger is pulled automatically at the very start of the month and entries are reviewed by bots before being handed to employees for exceptions and anomalies.
Automation applied to ledger reconciliation can be programmed to:
Not only does this accelerate the detection of fraud, it speeds up the entire business process and shaves hours off manual activity every week.
5. Applying more intelligent analytics
Only a small number of lenders are can do this today, but it is possible for to use advanced technologies like machine learning and analytics to spot fraud even faster.
Once manual processes across disparate systems are automated you have more clear data to work with. By applying intelligent analytics to this data, you can spot more patterns that signal suspicious activity across transactions, files and even communications patterns.
Once these technologies are set up, machine learning can be used to traverse much larger data including external and context data for more sophisticated pattern recognition and faster fraud detection.
A familiar pattern or a sign that things need to change?
The annual uptick in invoice finance fraud may be predictable in asset-based finance today. But lenders don’t have to concede that every year they’ll lose some unknown percentage of revenue to fraud.
By using automation to handle certain steps and processes, lenders can more efficiently serve customers and protect themselves from risk.
--
This article is written by @LuisHuerta, VP – Intelligent Automation, Firstsource and was first published at the Firstsource blog