How AI, Machine Learning and Low-Code/No-Code Approaches

AI, Machine Learning and Low-Code, No-Code approaches are ushering in the next generation of future-proof Buy Now, Pay Later (BNPL) initiatives.

As the BNPL space expands rapidly, organizations need to infuse their go-to-market strategies with advanced technology to make these programs sustainable – to manage risk and respond quickly to market needs, and to be agile to shift as needed to adapt and keep pace with the evolving regulatory environment.

Technology decisions made now will have a direct and tangible impact on the future adaptability, growth and longevity of your BNPL offering.

Here are eight key technology requirements to consider:

1. Ability to quickly leverage alternative data beyond traditional credit checks

In the high-risk, fast-moving BNPL sector, risk decisioning that’s accurate and based on real-time information is essential. Basic, soft pull credit checks often don’t report the most recent activity; this can make decisions riskier and less accurate.

Looking to data outside of the traditional credit score, such as alternative data such as behavioral scores, telco information, transactional data and open banking, can offer BNPL providers real-time insights into affordability and risk. To improve decisioning accuracy, seek to leverage data from a wide variety of sources.

New approaches eliminate hard coding to streamline data integrations, empowering users to quickly integrate and test new data. And the market is shifting toward the best-practice approach of using prebuilt connections to data vendor APIs that reduce integration times from months to minutes. This emboldens BPNL initiatives with newfound agility to access and use of data where needed across decisioning processes, onboarding processes, and/or for performance analysis.

2. Rapid onboarding for merchants and customers

Improving the ease and velocity of the BNPL onboarding experience for both merchants and customers is vital. After all, the onboarding experience is the first customer impression and a critical first interaction. According to recent research, unless a financial institution can open a new account or complete a new loan application in less than five minutes, the potential for the consumer to abandon the account opening increases to as much as 60 percent or more. Alternatively, faster account openings reduce abandonment rates down to 25 percent or less.

Automation in digital onboarding can significantly minimize customer effort. Ideally, automation augments customer data with the additional information needed to perform robust compliance checks, identity verification and risk decisioning all in real-time.

3. Agile compliance processes to address evolving regulations

A solid technology foundation can help BNPL providers accommodate shifting compliance regulations, in whatever industry sectors or geographic regions they operate in.

Building agile processes in areas such as Know Your Customer (KYC) and affordability requirements can ensure your BNPL offerings remain fully compliant. Solutions that leverage no-code, drag and drop user interfaces can empower risk teams to update processes, add in new data sources and make changes on-the-fly. By adopting these capabilities, providers can reduce their reliance on outside technology vendors while freeing up development resources to focus on other areas.

4. Integrated fraud detection

Fraudsters have been quick to exploit BNPL consumer-friendly onboarding and purchase experiences. Fully integrated fraud processes, such as robust Anti-Money Laundering and KYC tools, digital footprint tracking, transaction monitoring, simple integration or advanced fraud tools can thwart those looking to exploit system weaknesses. This is important, as catching fraud early in the process prevents bad debt being passed down the credit lifecycle.

5. Continuous improvement via analytics

Constant innovation requires constant iteration of analytics models. To this end, it’s essential to have the ability to monitor performance data as it’s happening and use that real-time information to identify trends. In turn, it must be easy to take those insights and make rapid changes to onboarding processes, models, credit line limits and more, forging a continuous improvement loop that drives innovation.

BNPL providers can leverage key capabilities critical to support rapid learning and iteration. Real-time visual performance dashboards offer a data analytics visualization “cockpit” to identify insights that empower innovation. The ability to use performance and decisioning data to train and retrain models in real time, rather than waiting months to insert updated models back into production environments, also plays a key role in accelerating product innovation.

6. Support for rapid time-to-market and BNBL business model diversification

Because your BNPL business may need to power consumer BNPL as well as business-to-business BNPL, it’s important for technology to support your BNBL business model today as well as your future strategy plans and diversification into new sectors. Technology elements that enable BNPL providers to pivot and enter new markets quickly include simplified data integration, low-code/no-code approaches, rapid model deployment and even prebuilt reusable decisioning templates.

7. Full customer lifecycle support

BNPL providers must grow and nurture customer relationships throughout the customer lifecycle. Look for technology that is extensible to support all aspects of the customer lifecycle, from onboarding to fraud management and ongoing credit line management to collections.

Having an enterprise risk decisioning ecosystem to manage the entire customer lifecycle results in smarter decisioning and superior consumer experiences. When all customer and decisioning data is consumable by that ecosystem, it eliminates data silos that prevent the business from fully identifying risk and empowers rapid iteration and innovation as well as greater operational efficiency and cost savings.

8. Use of AI/Machine Learning to support rapid risk modelling

It can take weeks or months for risk models to go live, with many never making it through the deployment process. As well, whether based on model drift parameters or a predetermined schedule, retraining models can be a time-consuming process. Machine learning – or ML Ops capabilities – can help BNPL providers retrain models in real time, for significant improvements in decisioning performance. Faced with data science talent shortages, many BNPL providers are finding significant value in prebuilt or custom-built models to accelerate the time-to-market and make strategic shifts in risk strategy.

Building Your Buy Now Pay Later offering for speed, agility and sustainability

Today AI, machine learning and low-code/no-code technology approaches offer BNPL providers tremendous advantages in architecting their BNPL offerings for speed, agility and sustainability. Careful forethought and purpose-built technology can address these eight key considerations for competitive advantage.

Source :