SaaS automation applied to identity fraud in trade credit

Part 3 of 3-part series.

The pace of technology development has a profound impact on most industries and our everyday lives. The world of trade credit management is no exception. From fraud assessment checks to business process automation, building scale through the cloud, to leading the charge on data-led decision making, the power of digital technologies should not be underestimated as your best-practice enabler.

Let’s look at trade credit fraud incited by SMEs.

The challenge is knowing which are disguised as fraudsters amongst them. This becomes even more of a challenge when credit applications are made online or through a channel that is more removed than face to face.

“Think about it. Any person can register with ASIC, get an ABN and buy a legitimate looking website and logo for just a few hundred dollars. In many instances, that’s all they need to be eligible for trade credit. It could be anyone applying.”

The Australian Institute of Criminology (AIC) report of 2020 shows that a 33% increase in identity theft has been reported in one year. Complete 100-point identity packages of stolen or fabricated identities are for sale on all darknet markets for as little as a few hundred dollars (IDCARE 2020). AIC valued the financial impact of identity theft in Australia between $1.4bn and $2bn annually (Research report 15, 2018).

“There’s no doubt that identity fraud is hurting business. As fraudsters become more sophisticated, business is turning to AI/ML and automation for the answers”.

Enter AI powered SaaS as your fraud investigator, credit checker and scam mitigator.

Detecting and preventing fraud with artificial intelligence (AI) makes sense. It’s immensely scalable and increases in accuracy over time when used in conjunction with machine learning (ML) capability.

For context you can differentiate both AI and ML. AI creates intelligent machines that can simulate human capability and behaviour especially beneficial for those repeatable tasks. In contrast, ML is an application or subset of AI that allows machines to learn from data without being programmed explicitly.

AI can help prevent and detect suspicious activity especially where it comes to identifying the person you are doing business with. It will come as no surprise that this technology is not going anywhere anytime soon, infact will only evolve as its capability grows.

Let’s take an example.

Meet Mark.

Mark is a small business owner, and he applies for trade credit.

On paper, Mark checks out. However, as a Credit Manager, you don’t necessarily have access to all the information you need about Mark. You may do a 100-point identity check with an ASIC check and possibly a credit bureau check.

On the other hand, Robotic Process Automation will ingest Mark’s application – that could be online or using a QR code in-store. If needed, it will use optical character recognition (OCR) to read the application, recognise the text and complete Mark’s data set.

Then Machine Learning algorithms run checks across large datasets automatically. From ASIC and Bank transactions to Government Licensing. It will automate facial recognition using biometrics and, in under 10 minutes, will predict whether Mark is a suitable applicant for trade credit. Not only is the process far quicker, but the magnitude of feedback signals makes it a far more robust process.

For other user stories download the 16 page eBook here, on this topic.

Machine learning models monitor, evaluate and detect red-flag anomalies predicting and prescribing the ‘next best action’.

This means practically that when a borrower is deemed high risk, the algorithm you have determined may not automatically decline the application. Rather, it will convey the information to the Credit Manager to make a more informed decision.

Where credit amounts are small (under a set threshold), auto-decisioning can be used. Here AI/ML will predict the likelihood of fraud and, if it’s found to be low risk, can make the approval on your behalf.

Artificial intelligence as the Credit Manager’s co-pilot:

Automation and AI/ML will help you with these critical functions:

Predictive Analytics – Analytics on large integrated datasets which can detect suspicious behaviour correlated with past instances of fraud.

Anomaly Detection – Detecting deviations from normal activity compared to historical data over a period that can be scaled differently for separate investigations.

Recommendations – Recommending a “next best action” immediately following detection.

Inaccurate data increases the risk of identity theft.
Hello tokenization.

The quality of customer identity data within your business is critical in reducing fraud and efficiently onboarding customers. Quality issues in identity data are plentiful. It could stem from insufficient or incorrect information collected during onboarding or because customer information (like addresses and phone numbers) changes over time. Then, try getting all your application channels – from sales reps to online applications – to be standardised, accurate and rationalised into one master data system. This fragmentation makes quality a more complex issue to solve.

Unfortunately, identity fraud increases with poor data, technology, and accessibility. This is where OCR technology, AI technology and even different types of facial recognition technology are available today for you to reduce this risk significantly. These technologies are a more scalable and more cost-efficient solution for problems created by poor quality in data. Goes without saying my money is on digital identity via NFT’s as referenced earlier.

Reducing the burden for your credit team and your customers.

An undesirable consequence of mitigating identity fraud is a bad customer experience. This is manifested in either lengthy procedures while you manually work through your due diligence; or in setting the risk threshold so high that it results in low pass rates.

Furthermore, with risk mitigation as the aim, many strategies tighten authentication and make it more burdensome for customers. The outcome of which is a significant drop in revenues from approved customers.

“Adopting automation solves this issue. You get the stringent checks, but the burden is on the machine, not your Credit Department nor the customer.“

It’s better to be the fence at the top of the hill, making views (and dreams) possible rather than the ambulance at the bottom of it. Typically, businesses that adopt automation for trade credit see a 30% rise in safe revenue within the first month. That’s the business case for automation, right there. Evolution or revolution, the benefits of advanced technologies are immense for finance leaders.

Author
To learn more about how technology can help your credit department mitigate risk and unlock revenue more quickly for the sales team, talk to Miriana Lowrie, CEO at 1Centre. Miriana and her team have produced an eBook titled “How to spot identity fraud across all your sales channels” which you can download today here. Before becoming the CEO of 1Centre, Miriana had 20 years of experience in big banking and understands how poor process and use of technology can frustrate employees, customers and doing great business.

Part 3 of 3-part series. Part 1 is here >

Read more articles related to this topic below:

SaaS automation applied to identity fraud in trade credit

Let’s look at trade credit fraud incited by SMEs.

Enter AI powered SaaS as your fraud investigator, credit checker and scam mitigator.

Meet Mark.

Machine learning models monitor, evaluate and detect red-flag anomalies predicting and prescribing the ‘next best action’.

Reducing the burden for your credit team and your customers.

Recent Blog Posts

Are you ready to process trade credit applications faster?

Helping business, get on with business

Let’s have a conversation

Support Enquiries

Location

Let’s have a conversation.