Skip to main content

Resource: Blog

Optimizing Your Data Strategy

Written by Allison Karavos on . Posted in Blog.

BLOG

Mo’ Data, Mo’ Problems:
Choosing the Right Data

Why the right data, not more data, is key to optimizing your data strategy

Big data and the term ‘data strategy’ gets thrown around a lot – but what is a data strategy when it comes to financial services, and how can you optimize it for more accurate, smarter risk decisioning? The answer isn’t more data, it’s the right data. Read on and discover how to choose the right data for your business use case and why optimizing your data strategy is key to your decisioning success.

Make Your Data Work Smarter, Not Harder

In our increasingly digital society, it seems like everyone is focused on more. More data, more choice, more speed, more competition, more options (how many different entertainment streaming services are there now??). More, more, more. So, our rebel yell is that when it comes to your data strategy, it’s not about more data, it’s about the right data, at the right time. According to IDC, “this year alone, over one hundred thousand exabytes of data will be generated, crossing the 100k threshold for the first time.” Yet 74% of decision-makers we surveyed said they struggle with their organization’s credit risk strategy because data is not easily accessible. The data is there, but it’s an incredible amount of wasted effort if you don’t know which data sources to use when.

74% of decision-makers struggle with their organization’s credit risk strategy because data is not easily accessible.

2022 GLOBAL FINTECH AGENDA, POWERED BY PULSE

When developing a data strategy for your financial services offerings, you need to look for ways to minimize costs and maximize innovation. And that means being able to select only the data you need, exactly when you need it, in order to make more accurate decisions across credit, identity, and fraud. According to McKinsey, “industry leaders tap multiple internal and external data sources to improve the predictive power of credit signals… both the internal and external data sources used in a credit-decisioning model will affect the decision quality.”

What can the right data do for your decisioning strategy?

As McKinsey put it, “Data marketplaces enable the exchange, sharing, and supplementation of data, ultimately empowering companies to build truly unique and proprietary data products and gain insights from them.” When it comes to risk decisioning specifically, that translates into several key benefits – and competitive advantages:

  • Improved customer experience: Ensure a frictionless digital experience for low-risk customers and enable data-driven actions on potential risk in real-time
  • Improved accuracy in your decisioning: The right data at the right steps in your decisioning processes across the customer lifecycle means more efficient, accurate risk decisions
  • Minimized data costs: Reduce the time/effort/resources necessary to source, build and maintain data integrations if all the data you need is right at your fingertips
  • Scalability: With the right data sources on both a local and global level, you can get new products to market in new regions faster by duplicating and iterating your data strategy
Types of data that are critical to optimizing your decisioning strategy across the lifecycle:
  • Identity Data: Verify identities and documents for better onboarding compliance, prevent identity fraud, and be sure that you are protected with ongoing due diligence data.

      Includes: KYC/KYB, PEPs/sanctions, document verification, synthetic ID fraud
  • Fraud Data: Identify potential first-party and application fraud in real-time to proactively detect/prevent fraud and reduce losses; reduce false positives by leveraging signals from mobile, email, behavior, device, IP, social and other fraud data sources.

      Includes: Email and mobile data, global fraud intelligence, social validation, device data, IP, and geolocation
  • Credit Data: Minimize credit exposure and loss by leveraging credit bureau, open banking, and alternative data sources. Ensure optimized credit onboarding and add value throughout the entire customer lifecycle with dynamic customer risk profiling, mitigate collections and optimize customer lifetime value.

      Includes: Credit bureau data, business data, open banking and alternative data including social media, rental payments, travel info, utilities and more
Data supply chain challenges and how to overcome them
Choosing the right data can seem daunting, but it’s critical to have an optimized data supply chain, with the right data in the right place, in order to deliver the most effective products to your customers. And depending on the type of financial product you are offering there are regional regulations to consider, third-party vendors required, technology requirements and more. These are some of the most typical challenges known to slow down deployment of even the most well-thought-out data strategies:
  • Identifying relevant local data sources
  • Negotiating multiple contracts
  • Complying with varying regulations
  • Ensuring data privacy for different regional requirements
  • Normalizing data formats
  • Building and maintaining integrations
  • Supporting global strategies
But you can overcome these challenges by ensuring you have the right data for each and every product offering you have. How? Work with a partner that provides an all-in-one data solution. Building your own data supply chain, for whatever your use case, is possible of course, but it’s time-consuming and resource intensive. If you want to work with a partner look for a data solution that offers:
  • One data contract that provides access to multiple data sources
  • A single API to replace numerous integrations
  • A wide variety of data types and sources, including alternative data
  • Expert data source curation customized to your needs, that can be easily modified as your needs evolve
  • Simplified, no-code data supply chains that non-technical users can understand and control
  • Global data access, as well as local sources, to ensure success of both regional tactics and the ability to iterate and expand to new markets
  • Seamless integration into your decisioning technology to ensure accurate, smarter decisions
If you’re a Buy Now, Pay Later provider, or are thinking of diving into the fray, check out our blog highlighting specific ways to optimize your data supply chain for BNPL.

Read the Blog

Simplifying your data supply chains (sourcing, building, integrating, and maintaining data sources and connections) and optimizing your data strategy is critical to continued success – and your competitive advantage. Don’t let yourself be overwhelmed by the immense variety of data available out there – remember, the right data is much more important than more data. Accessing the right data at the right time means enhanced risk models, strengthened onboarding processes, more accurate decisioning across the lifecycle, and optimized customer experiences.

For further reading, check out these articles that may be of interest:

What the Data-Driven Bank of the Future Looks Like

– The Financial Brand

Designing Next-Generation Credit-Decisioning Models

– McKinsey

The Data-Driven Enterprise of 2025

– McKinsey

Ready to unleash the power of data?

Meet Provenir Data

LATEST BLOGS

Provenir's Data Integration Tools

Provenir’s Data Inte...

Accessing the right data at the right time is
Read More
The Superman Effect: The Human Side of Banking UX Design

The Superman Effect:...

In today's digital-first banking environment, customer expectations are higher
Read More
Driving a Better Consumer Experience in Auto Financing

Driving a Better Con...

Customer satisfaction is crucial in auto financing. This article
Read More
Credit Risk Software: Build vs. Buy Options (Complete Guide)

Credit Risk Software...

The decision to build or buy credit risk software
Read More
The Ultimate Guide to Decision Engines

The Ultimate Guide t...

BLOG The Ultimate Guideto Decision Engines What is a
Read More
Ten Fintechs Using Alternative Data for Financial Inclusion

Ten Fintechs Using A...

This article explores the use of alternative data to
Read More
Finally, The Secret to Credit Risk Modeling with Python

Finally, The Secret ...

Credit risk modeling has evolved significantly, with Python playing
Read More
7 Reasons to Use Salesforce for Credit/Loan Origination and Risk Decisioning

7 Reasons to Use Sal...

Imagine scrolling through a long list of song choices
Read More

Continue reading

Guest Blog: Three Best Practices for Implementing Digital ID Verification

Written by Allison Karavos on . Posted in Blog.

GUEST BLOG

Three Best Practices
for Implementing Digital ID Verification

  • Christina Luttrell, Chief Executive Officer at GBG Americas

A recent report showed that 86% of businesses view identity verification as a strategic differentiator, allowing them to capitalize on digital adoption while delivering a seamless customer experience. Consumers who don’t trust the digital identity verification process are more likely to use guest checkout (54%) and less likely to keep a payment card on file (43%), thereby creating a drag on profits while compromising the end-user experience.

The following best practices can help fintechs locate, verify and approve new customers without friction or fraud while streamlining the customer journey.

Onboarding in The Digital Landscape

Being successful in a digital environment means being able to onboard and verify users in a purely digital way. This means doing all the required elements, such as KYC, AML, checking against sanctions lists, etc., in a digital-only environment, which can be challenging.

This means needing to design a UX that is inclusive of digital identity verification at its core, with access to multiple verification layers that can be deployed in each required scenario. Fintechs make money by people utilizing their service. Providing a digital experience that opens the door to more good customers—while also meeting regulatory requirements—is a goal for all fintech providers.

A robust ID verification solution gives fintechs the confidence to onboard more legitimate customers faster, with nominal friction, while staying compliant.

Data Diversity & Consortium Networks

Central to the requirement for effective digital identity verification is data diversity. Incorporating other identity verification data sources is essential, as the more indicators are used, the more robust the system is compared to a traditional system reliant on credit checks, which can be breached.

The other consideration is data transparency – data must be sourced and explained, as a critical requirement for ongoing regulatory compliance, and justify decisions to customers.

This is where the idea of consortium networks, where data is shared between a large network of interconnected parties, becomes highly important, as they enable new account openings at different institutions to benefit from fraud data and learnings elsewhere in the ecosystem, securing the whole market more effectively.

Ongoing Verification

Onboarding is an important element of fraud prevention, but ongoing verification is necessary, which is the authentication part of the equation. Opening a fraudulent account is a risk, but account takeover of an existing account is also a significant risk, as payment account fraudsters have access to make payments and view transaction history and payment details.

The requirement is for fintechs to design strategies that ensure that verification is carried out continuously. This could be when an unusual transaction is made, or when a new payment method is set up, or in any number of given scenarios. 

Given what’s at stake, if fintechs fail to implement robust systems based on more than just point solutions for ID document scanning, they will struggle to deal with evolving fraudster tactics. For this reason, the industry could see the continued fusing of physical and digital attributes for verification, such as taking name, address, date of birth, etc. Only by taking a multi-layered, customizable approach will banks achieve the best anti-fraud and customer experience outcomes.

Visit IDology.com to discover innovative solutions that streamline customer acquisition, deter fraud, and drive revenue.

Ten Companies Using Alternative Data for the Greater Good

Read the Blog

LATEST BLOGS

Continue reading

Data Now, Fewer Losses Later: Optimize Your BNPL Data Strategy

Written by Allison Karavos on . Posted in Blog.

BLOG

Data Now, Fewer Losses Later:
Optimize Your BNPL Data Strategy

Buy Now, Pay Later (BNPL) products have exploded across the globe, offering a new spin to point-of-sale (POS) financing for both consumers and businesses. From 2019 to 2021, BNPL loan originations increased by 970% from the top five lenders alone, and the industry continues to expand to include new verticals such as auto repair, grocery purchases, airline ticketing, and more. Consumers are beginning to rely on BNPL for everyday costs in order to help manage their cash flow. But none of this would be possible without data – more specifically, a strong data supply chain. 

If you’re a BNPL provider, the data supply chain is the powerhouse for your solution. When you have the right data, you can better determine risk, protecting your business against fraud and loan default. 

BNPL data strategies look beyond traditional data like credit scores and use alternative data to make credit more accessible and faster to approve without increasing your risk. While this allows you to expand your customer base in a secure way it also adds complexity to your data needs. So, how do you build a BNPL data supply chain strategy that gets the right data to the right place exactly when you need it?

Building Your BNPL Data Supply Chain

We know every potential BNPL customer must go through a process, but what does that process look like? Each step is built with different data checks that tell your decisioning engine whether to move that customer forward. An optimized data supply chain pulls only the necessary data needed for a customer at each checkpoint – data that comes from your data integrations and data partners. 

An optimized data supply chain has these hallmarks:

  • Multiple steps with distinct requirements
  • Multiple checkpoints at which consumers either pass or get denied
  • Steps that increase in complexity and cost of data
  • No unnecessary data is exposed and paid for before you need it

Launching with an MVP:

Are you a startup launching your first BNPL solution? A finserv expanding your product line? Maybe you’re an online shop looking to reach more customers. Whatever the case, when building a new data supply chain for your BNPL offering or optimizing an existing one, you should begin with your minimum viable product (MVP) – the basics you know you need to launch your product. 

An MVP has the least amount of checks in the process, pulling in the least amount of data. You might want to begin with an MVP if you want to:

  • Go to market quickly
  • Minimize the cost of development
  • Analyze basic performance to optimize more complex iterations in the future

To launch with an MVP approach you’re going to need data to support three key areas: 

  • Regulatory compliance checks like KYC/AML
  • Identity verification 
  • Credit risk

 An MVP for consumer lending could look like this: 

Step 1: KYC

The first step of the process is validating the most basic data to confirm the customer’s age, address, and identification. If you can’t verify a person’s ID, you certainly can’t lend to them. 

Step 2: Fraud Prevention

The second step digs deeper into a person’s identity to ensure they are who they say they are and help prevent fraud. There is a wide variety of data you can pull for a fraud check, including email address verification, if a SIM card has been swapped, and other behavioral and alternative data. If not all of this information matches, it could be a sign of attempted fraud, and the person would be rejected. 

Step 3: Credit Risk

The final step is to check creditworthiness. A bureau check is done through a soft credit check that grants you access to a consumer’s credit score without impacting it. With an MVP, BNPL providers would likely reject anyone with a score below a certain threshold or someone without enough credit history to have a score at all. If a person has made it through the process, the data is assessed holistically by a decisioning engine to determine whether and at what terms to grant the loan.

Beyond the MVP: Optimizing Your Data Strategy

Beyond the foundation needed for an MVP launch, you can optimize your supply chain based on your company’s risk appetite and goals. Before updating your data supply chain it will help to: 

  • Analyze success against your goals
  • Identify weak points in your data strategy

While you may want to initially launch your BNPL solution using an MVP, as you grow and want to add complexity, you can incorporate new data points and data partners. Think about the kind of customer you want to capture, as well as business goals and preventative measures you may want to take, and ask yourself:

What percentage of fraudulent applications is our current process letting through? Is this in line with our business goals? If not, look to:

  • Add additional fraud checks on existing steps
  • Add standalone fraud prevention steps to the process
  • Amend data sources to optimize as you go

Are we offering the most competitive terms to our customers? How can we improve conversions? For competitive edge and increased personalization, use data such as:

  • Behavioral trends
  • Geolocation
  • Activity and usage 

How effectively are we reducing defaults? Are we filtering out non-viable customers at the right point in the process? Make sure your flow features:

  • Prescreening
  • Scoring
  • Additional data checkpoints on existing steps

For BNPL providers that want robust data supply chains across credit, identity, and fraud while maximizing efficiency, an optimized flow could look like this:

BNPL Data Supply Chain 2

Prescreening 

Prescreening breaks down the identification verification steps even further, making sure the minimum requirements are met. It’s a faster, more efficient way to filter out unqualified applicants without using unnecessary time and resources. 

What does prescreening look like in an optimized supply chain? Say you have a person under 18 – they’re not legally allowed to take out a loan, so their application would be rejected. In an MVP, someone that can’t even use the product would still have their identity verified, but it’s a waste to run those checks, since they’re not a viable customer. Optimization ensures you expose only the data you need at each step.

Scoring

Scoring pulls supplementary data that helps paint a clearer picture of a consumer’s risk. This includes mobile device data, additional fraud checks, or any other kind of alternative data you want to feed into your decisioning tech. 

Why include scoring in your process? Again, it comes down to building your process for optimal efficiency and minimal cost. At this point, you would know if the customer was viable, who they are, and what their financials look like – this is all straightforward data to pull. Scoring adds behavioral information that is more time-consuming and costly to analyze and should be incorporated only when everything else checks out. 

Ultimately, the more relevant data you have, the more accurate your decisions will be, the better you can predict future defaults, the easier it will be to identify upsell and cross-sell opportunities – whatever your business goals, the right data can help you get there. Optimizing your consumer BNPL supply data chain is dependent on finding the ideal number of checks and steps to accurately determine creditworthiness and risk, while keeping the process fast and efficient.

Ready to launch and expand your BNPL products? Look out for these data supply chain challenges

As BNPL products continue to grow around the world, new markets have emerged, and with them new challenges. To build a global supply chain, you have to know regional regulations, vendors, tech requirements, and more. Some of the challenges that can slow down deployment of your data strategy include:

  • Identifying relevant local data sources
  • Negotiating multiple contracts
  • Complying with varying regulations
  • Ensuring data privacy for different regional requirements
  • Normalizing data formats
  • Building and maintaining integrations
  • Supporting global strategies 

BNPL is a fast-moving industry, so it’s also important to ensure your supply chain can be easily iterated on to incorporate evolving legislation and market demand. 

Data Powers BNPL

Regardless of trend, customer type, or region, your BNPL solution is powered by data. Diverse data sources pulled at the right time in the right order is the calling card of an optimized data supply chain. And an optimized data supply chain feeds your decisioning engine the information necessary to give you a smarter decision every time. 

Building a data supply chain on your own, however, can be a huge undertaking and an even bigger headache. Instead, consider choosing a data partner that can build it for you, while connecting you to the integrations you need to grow your BNPL business. 

Ideal features include:

  • One data contract that gives you access to multiple data sources
  • A single API to replace numerous integrations
  • A wide variety of data types and sources, including alternative data
  • Expert data source curation customized to your needs
  • Simplified, no-code data supply chains that non-technical users can control
  • Global data access
  • Integrates into your decisioning technology to ensure seamless and smarter decisions

Do you want your data on-demand? Meet Provenir Data.

Get the Datasheet

LATEST BLOGS

Continue reading

Innovations in Risk Decisioning Fuel YapStone’s Rapid Global Expansion

Written by Allison Karavos on . Posted in Blog. No Comments on Innovations in Risk Decisioning Fuel YapStone’s Rapid Global Expansion

BLOG

Innovations in Risk Decisioning
Fuel YapStone’s Rapid Global Expansion

The global sharing economy continues to transform the online payments landscape, as we know it.

We spoke to YapStone, a payment platform taking the world by storm, and asked them how their new credit decisioning model has helped expand their company at a rapid rate, on a global scale. Through advanced analytics, model development has taken place quickly and effectively for YapStone. Here, they talk us through how credit scoring, open banking data, and credit risk models have facilitated their growth, allowing them to sit amongst industry leaders.

In a time when virtually anyone can sell goods or services on the internet, real-time merchant onboarding and risk and fraud monitoring capabilities have become imperative.

At YapStone, we know this very keenly because the bulk of our users are not simply selling products to strangers they will never meet – they are inviting consumers into their homes.

Twenty years ago, we couldn’t have imagined we’d be comfortable inviting complete strangers to stay in our home for extra income, but thanks to vacation rental marketplaces like HomeAway, Airbnb, and Kigo, “living like a local” has become the preferred way to travel. As a result of this undeniable trend, YapStone now processes about $18 billion (and growing) in electronic peer-to-peer transactions every year.

Further challenges to marketplaces are emerging with the rise of Alternative Payment Methods (APMs). Consumers in different countries or regions have their APMs of choice, using them to pay securely with their local currency. As the payment partner, we have to ensure that these methods suit our customer’s lifestyle, and that they can securely pay using their preferred payment methods in their local currency, while sellers receive the funds seamlessly in their local currency.

YapStone has been able to capitalize on the growth of apartment and vacation rentals, where the average ticket size is large and the risk is high, by developing proprietary technology focused on reducing the risk of fraud and loss to our marketplace partners. YapStone’s trust and safety solutions are highly flexible and designed to service all marketplace types, allowing us to diversify our portfolio and grow our business beyond apartment and vacation rental marketplaces.

YapStone is unique in that we offer a full service, end-to-end payments acceptance, customer service, and risk management solution, including instant and advanced payments, to our marketplace partners.  Therefore, it is a necessity for YapStone to verify the traveler and authorize their payment method, as well as verify the vacation property’s existence and ownership. Our goal is to deliver trust and safety to our marketplace partners, allowing them to spend more time focused on growing their business, while leaving risk management to the expert team at YapStone.

YapStone uses a layered, risk-based approach focused on the persona of any particular client interaction.  Within the persona, we are continuously monitoring the purchaser of the good or service, the payment instrument being presented, and the asset or property they are renting.  To accomplish this, YapStone utilizes proprietary data science and predictive analytics augmented with 3rd party data to achieve the most accurate risk scoring.

Given our scale, and the risk associated with high average tickets and the speed at which fraud can happen, it was critical for us to choose a tool that allows a risk analyst to react quickly to an escalating threat.

In 2017, we selected Provenir as a key strategic partner in the development of YapStone’s next-generation risk decisioning platform. The tool provides the ability to house our proprietary underwriting and fraud models, serve as the hub for our third-party risk vendor integrations delivering powerful adapters to augment our proprietary risk methodologies and data, and conduct A/B and regression testing for new or proposed model changes, all delivered through a simplified user interface that doesn’t require a PhD in computer science to use.

We maintain a highly competitive edge over other payment facilitators because of the way we mitigate risk for our marketplace partners and assume the liability for each transaction. Using our proprietary technology in partnership with Provenir, YapStone is able to provide marketplaces with high levels of automation for merchant onboarding and risk management aimed at improving speed to revenue and reducing losses for our clients.

We have a very exciting future ahead of us, made particularly bright by having partners like Provenir who help us deliver innovative solutions to our existing customers and new faces. As we expand into new territories, the demand for innovation in risk decisioning will be high. The team at YapStone looks forward to staying on the forefront of this new wave of marketplace payments.

Companies invest lots of time and money developing risk models to figure out which  customers are the best bets for loans and credit, including auto lending, mortgages, credit cards, BNPL and more.

Operationalizing these models,  developed in tools like Excel, SAS, Python, and R, within risk decisioning processes often turns out to be challenging. This is especially true with complex models built in R. Lots of risk decisioning ‘solutions’ demand that you manually translate the R model (or any other model that you are using) into code that it can understand. You need high-priced programming resources and lots of time to connect the R model to the risk decisioning process.

It’s much more efficient to use a risk decisioning solutions which are model-agnostic – in other words, a solution that doesn’t care how the model is constructed. Provenir’s AI-Powered Decisioning Platform is a great example of this model-agnostic approach. With this platform, models developed in a variety of tools can easily be imported, mapped, tested and validated using simple wizards.

Provenir automatically generates a list of the data fields; all users have to do is pick the data Provenir needs to send to the model, as well as the data the model should send back to Provenir to drive the decisioning. This entire process takes just a few minutes, which means you not only gain an effective way to maximize the value of your models, but can also instantly adapt risk decisioning processes whenever a model changes.  Automating your risk decisioning not only saves you time and money – it improves your decisioning accuracy and allows you to focus your resources on growing and scaling your business.

The Ultimate guide to Decision Engines

What is a decision engine and how does it help your business processes?

Take a look

LATEST BLOGS

Continue reading

Simplifying the Merchant Onboarding Process with Automation

Written by Allison Karavos on . Posted in Blog. No Comments on Simplifying the Merchant Onboarding Process with Automation

BLOG

Simplifying the Merchant Onboarding Process
with Automation

The Challenges of Manual Merchant Onboarding

Merchant onboarding is a critical process for acquiring businesses that involves acquiring, analyzing, and integrating large volumes of data. However, the manual and time-consuming nature of this process often results in delays and errors. For instance, if data or knowledge of the merchant is lacking, then identity can’t be validated. Compliance with Know Your Customer (KYC) and other governmental regulations has to be determined, as does creditworthiness. This takes days and can still involve a high degree of manual handling. To streamline this process and make it more efficient, automation is the way forward.

Compliance with KYC and Other Regulations

KYC stands for “Know Your Customer” and is a process that financial institutions and other regulated companies use to verify the identity of their clients. This process involves collecting and verifying various types of information about the client, such as their name, address, date of birth, and other identifying information. The objective of KYC is to prevent financial crimes such as money laundering, terrorist financing, and other fraudulent activities.

During the merchant onboarding process, compliance with KYC and other governmental regulations is required. Failure to comply with these regulations can result in fines and other legal consequences. By automating the merchant onboarding process, companies can streamline the KYC process, making it quicker and more efficient, while also ensuring compliance with regulatory requirements.

The Benefits of Automation

Simplified Data Integration

To simplify data integration, acquirers need to access and efficiently handle and analyze all data sources such as bank account information, commercial data, address verification, KYC checks, credit score, and more. To achieve this, a merchant onboarding solution with integration capabilities that can rapidly aggregate data from various sources is required. Non-standard data, such as that from social media, can supplement sources – if the acquirer has the means to get at it and pull out what’s relevant. To achieve this, the best solutions offer pre-built adaptors built on industry standards.

Operationalized Risk Models

Operationalized risk models play a critical role in the merchant onboarding process. They integrate with the other elements that make up the end-to-end merchant onboarding process, ensuring that risk decision-making is not a bottleneck in the process. Technology and model-agnostic solutions can integrate with SAS, Excel, and anything else besides. Business-defined rules lay down the terms and conditions for each merchant and identify exceptions that require further investigation. A visual interface lets business users quickly establish the relationship between the risk model and the automated onboarding process.

Effective management of data is essential to a rapid, efficient merchant onboarding process. Technology for automated risk analytics and decision-making integrated into the onboarding process taps into multiple data sources and systems for a streamlined end-to-end process. To learn more about simplified data integration and operationalized risk models for merchant onboarding, check out our guide on our website.

Learn how you can choose the best merchant onboarding automation solution.

Get the eBook

LATEST BLOGS

Continue reading

Machine learning – all a bit ‘Skynet’?

Written by Allison Karavos on . Posted in Blog. No Comments on Machine learning – all a bit ‘Skynet’?

BLOG

Machine learning –
all a bit ‘Skynet’?

Machine Learning – Revolutionizing Financial Risk Analysis and Decision-Making

While the concept of Machine Learning (ML) may conjure up images of a dystopian future where machines have taken over the world, much like Skynet in the Terminator movies, the reality is quite different. While Skynet may have been a malicious and all-powerful AI system, Machine Learning is simply a tool that can help us better understand and leverage data. So, while Skynet may have been the ultimate villain, Machine Learning is more like the trusty sidekick that helps us save the day.

Here’s how ML is rapidly becoming a game-changer in the field of financial risk analysis and decision-making:

The Power of Data

Machine learning enables businesses to gather and analyze data faster, thereby arriving at insights quicker. This is because the software program uses pattern recognition to build automatic analytical models, eliminating the need for human intervention.

Dynamic Fraud Detection

Machine learning algorithms can learn from a customer’s previous transactions and use them to identify patterns of behavior, allowing for dynamic fraud detection. This eliminates the inconvenience of manual validation processes while also increasing fraud detection rates, saving considerable costs.

Huge Cost Savings

According to analysis firm Oakhall, global financial services firms could save $12 billion annually through machine learning fraud management. This underscores the tremendous potential for risk analysis and decision-making with machine learning.

Harnessing Machine Learning for Predictive Analytics

To fully benefit from the predictive analytics power of machine learning, financial institutions need a fast, simple way to connect their machine learning application to their credit and lending decisioning processes.

Machine learning is revolutionizing the financial risk analysis and decision-making process. Its power lies in its ability to gather and analyze data faster, dynamically detect fraud, and save costs. By harnessing its predictive analytics capabilities, businesses can unlock its full potential for risk analysis.

Deploy Machine Learning in Your Financial Institution Rapidly

Get the Whitepaper

LATEST BLOGS

Continue reading

Loan Origination Software Plays Its Part in Banking’s Digital Transformation

Written by Allison Karavos on . Posted in Blog. No Comments on Loan Origination Software Plays Its Part in Banking’s Digital Transformation

BLOG

Loan Origination Software
Plays Its Part in Banking’s Digital Transformation

Loan origination — and, subsequently, loan origination software — is at an interesting intersection right now.

At a less institutional level, peer-to-peer lending is expected to grow at a CAGR of 53% between 2016 and 2020. But as lending technology matures, its impact could reduce profits in American banks by $11 billion per year, or roughly 7%.

This evaporation of margins is not a new problem for banks, though it is one that is increasingly disconcerting. In 2012, for example, the share of risk and compliance within general banking costs was 10%, a large portion of overall costs as they were. In 2017, risk and compliance are expected to consume 15% or greater. While costs are rising, it’s hard to actually mitigate risk with incremental risk management improvements. In large part, return on equity in banking often resides below the cost of capital, impacted by capital building projects and fines.

The result: to see increased growth into the next decade, banks are digitizing more processes. This has begun to happen, but a variety of studies — including many on millennial bankers — shows the digital transformation of financial services has not yet fully arrived. Since lending is a huge revenue source for banks across virtually all segments from small business to enterprise, making sure digital loan origination is properly executed is preeminent for many banks now.

As McKinsey has noted, the shift to increased digital transformation focus in financial services came about because of five distinct pressures (paraphrasing here):

  • Changing customer expectations: Consider the rise of mobile and on-demand experiences.
  • More regulations and risk-function effectiveness: Seen in increased regulations in most first-world economies, as well as more fines being dealt since the 2008 crisis.
  • Data management and advanced analytics are hallmarks of competitive banks now: Buzzword or not, we are living in a Big Data era.
  • Disruption: Risk management programs are essential for banks to compete with upstarts — if the upstart makes a big bet and misses, the established bank can favorably reposition.
  • Increasing pressure on costs/returns: And as noted above, risk management doesn’t necessarily deliver in this way on the balance sheet.

As such, rapid-fire, on-demand loan origination programs and software have begun cropping up in the financial services world. Why? When risk decisions are made in seconds, loan origination cycles shorten for the customer. Shorter cycles create positive customer experiences, brand loyalty, connection to the bank and its relationship managers, and continued business. Speed can be good.

But, banks attempting a new approach to loan origination need more than speed. While quick risk decisioning and credit scoring is crucial, a loan origination software program also needs:

  • The ability to process both structured and unstructured data: This would allow for the incorporation of both standard and alternative decisioning, i.e. credit documents vs. items from a loan applicant’s blog.
  • Compliance: Perhaps the most important at the bank level, compliance calculation and True in Lending Act (TILA) disclosures need to be compliant, and documents must be in compliance with the Electronic Fund Transfer Act.
  • “Look for initiatives within easy technological reach:” That’s advice from McKinsey above, and it makes good sense. Some loan origination software barely requires advanced coding anymore, so your IT side can work on more value-add internal projects.

There are other considerations such as ease to operationalize risk models that should be deployed.

Financial services, and especially loan origination have long suffered from a lack of transparency and simplicity. It oftentimes seemed that financial services firms were underserved by technology, or “square-peg/round-holing” the problem. That’s not the case anymore, and loan origination software and approaches are of huge value for established banks as a way to drive a growth culture forward. The crucial step is the right partner for your specific needs.

LATEST BLOGS

Continue reading

Loan Origination in the Golden Age of Instant Everything

Written by Allison Karavos on . Posted in Blog. No Comments on Loan Origination in the Golden Age of Instant Everything

BLOG

Loan Origination in the Golden Age
of Instant Everything

It may look like the golden age of “instant.” The casual iPhone user can get an Egg McMuffin delivered to their door while it’s still steam-hot in the bag. The average Twitter user can use their ampersand key to swat through a brand’s customer service obstacles like Arnold Schwarzenegger cutting through the jungle in Predator. It may look like we’re in the golden age of immediate-results technology, but we’ve only just reached the earliest, primordial stages in its existence, and consumers have already and instantly adjusted to the instant  age.

As the world of apps grows, consumers have, and will, expect every corner of our daily lives, especially the institutions that manage our financial loans or our medical information, to be as cloud-efficient and scalable as the app that schedules an undergrad to walk your dog while you’re in a meeting. Technology tends to evolve inwardly and sensitively, finding it’s way into our banks and homes–not broadly and impersonally. Notwithstanding, loan origination systems are a deeply personal technology.

As a microcosm-example, a 2016 piece in Forbes summarized that brands who engage in direct customer service via Twitter see a 19% increase in customer satisfaction. Loan origination is one of life’s sensitive areas–vulnerable like our medical information or mortgage payments–that needs to adapt to this instant-evolution.

Picture the expectations of instant response when we have an artificially intelligent platform accessible from a contact lens. How long do you think a customer will tolerate waiting for approval or a green “success” check or a loan when all of the technology around us has already reached the speed and accessibility only dreamed up in Ray Kurzweil novels?

Waiting, whether for your breakfast sandwich or your loan decision, is as a dead and dusty a concept as your CD-RWs. Instant technology that learns for the individual and can execute in real time is the present and future. We’re already expecting banking to happen in the blink of an eye. In the future–especially in light of the cascading failures in recent financial technology–it will need to happen even faster, more efficiently, and securely.

Deploy Machine Learning in Your Financial Institution Rapidly

Get the Whitepaper

LATEST BLOGS

Continue reading

The Algorithm Challenge – Using AI for Risk Decisioning

Written by Allison Karavos on . Posted in Blog.

BLOG

The Algorithm Challenge –
Using AI for Risk Decisioning

  • Giampaolo Levorato, Senior Data Scientist, Provenir

How to implement advanced AI algorithms for improvements across the modeling lifecycle

We’ve all heard the term Big Data, and the world of financial services is no exception. Big data refers to large, structured and unstructured sets of information growing at ever increasing rates. Data drives key decisions made by fintechs and financial services organizations – everything from helping determine identity and approving a car loan or a mortgage to optimizing pricing and deciding when to upsell a current customer. The surge in volume, variety and velocity of data has led financial institutions to use advanced machine learning algorithms to make smarter, faster decisions. But using AI is not without challenges. There can be several obstacles to successful deployment, including choosing the right algorithms, interpreting and explaining complex models, deploying the models, ensuring the infrastructure is sufficient, and managing bias.

AI Challenges

  1. Choosing the right algorithm: not all algorithms perform equally well on the same dataset. Depending on the nature of the data, organizations must be able to choose and configure the best algorithm to fit their data.
  2. Model complexity, interpretability and explainability: the intricacy of AI algorithms can make them “black boxes” in the sense that often even the developers don’t know why and how the algorithms make the decisions they do.
  3. Model deployment: deploying a model into production requires coordination between data scientists, software developers and business users, posing a challenge with regards to the different programming languages and approaches that need to be unified into one solution.
  4. Infrastructure requirements: many organizations lack the infrastructure required for data modeling and reusability. Being able to quickly develop and test different tools, across different, large datasets, is essential to producing more accurate, manageable results.
  5. Exclusion bias: many consumers globally remain ‘credit invisible’ or thin-filed, meaning that little-to-no credit scores are available for them.

Overcoming the AI hurdles

What’s the best way to tackle these challenges? Financial services organizations should transition from traditional Generalized Linear Models (GLM) to explainable AI algorithms to improve the speed and accuracy of their decisions. According to a recent survey conducted by Pulse and Provenir, 69% of companies plan to invest in AI-enabled credit decisioning in 2022.  AI algorithms can also help to more easily identify fraud and creates opportunities for improvement of the customer experience across the entire lifecycle.

Benefits of AI

  • Algorithm Optimization: choose the most appropriate algorithms from a wide variety of options, including Gradient Boosting Decision Trees, Random Forests and Deep Neural Networks, depending on the nature of the dataset.
  • Interpretability and Explainability: through a careful adoption of SHAP and LIME explanation methods it is possible to explain how and why your model has made a prediction.
  • Ease of Deployment: use of a unified platform enables seamless deployment, allowing businesses to take fast, effective action.
  • Scalability: reduce the development time from months to days by automatically training, testing, monitoring and managing your model.
  • Diverse Data: by leveraging traditional and alternative data, improve your model accuracy, while managing bias and promoting financial inclusion.

Moving to AI algorithms has numerous benefits – including higher accuracy, improved compliance and superior scalability – all of which have tremendous impact on your overall business stability and growth. Using AI algorithms means more predictive, more accurate models, resulting in increased profits, reduced losses and more up-to-date risk assessments. After conducting internal research, Provenir has observed that AI algorithms can improve a model’s accuracy by up to 7%, while automated model development and deployment can reduce time and effort by up to 90%. This automation ensures faster speed-to-market with more accurate models and the ability to quickly respond to consumer needs and market trends, for true scalability. And the effects of this go beyond an individual business when you consider the further-reaching implications on the economy as a whole – The Wall Street Journal forecasted a 14% increase in the global GDP by 2030 thanks to the advancements of AI.

More legislation is now in play that requires full explainability of models. Fully interpretable and explainable models meet these requirements by clearly demonstrating how and why models make the decisions they do. In addition to compliance, model governance can be incredibly difficult with traditionally siloed environments. Separate environments for data collection, model development, deployment and monitoring require an immense amount of time and resources to integrate.  With a cohesive, all-in-one environment you eliminate that integration time and effort, enabling live, real-time results and helping reduce human error from manual processes.

The Value of a Unified Platform

Further to the siloed environments of data collection, model development, deployment and monitoring, models are also often built separately from decision engines and unnecessarily moving data between them increases time, effort and the probability of errors. With a unified platform that incorporates data, AI and decisioning, models are built and implemented in the same platform, ensuring seamless data and model integration, eliminating recoding delays and ensuring maximum performance of your models. In Provenir’s experience, models implemented in a unified platform can save up to 30% of a modeling project’s overall time and effort.

But what makes AI so powerful and capable? It’s all about the data. The more data your AI models have, the better your advanced algorithms will perform. A data-agnostic platform that can integrate and enrich your existing data sets with any other type of data set (i.e. various forms of alternative data) is critical. This seamless integration to a wide variety of data sources helps to encourage financial inclusion, manage bias and improves the predictive power of your models. And it’s not a one-and-done deal – true value comes from the continuous improvement that happens when you bring data, AI and decisioning together. Model monitoring and a constant feedback loop helps you fine-tune your decisions for continual optimization.

Being able to increase your predictive power and make more accurate decisions has impacts across the entire customer lifecycle. Real-time dashboards and reports help you stay up-to-date on changes with your customers, your portfolio and all of your models – allowing you to automatically generate updated predictive models, with everything available for live monitoring. This helps to enable better relationships with your customers, increases your agility in responding to market needs, and better predicts (and prevents!) fraud and loss.

According to The Economist, 86% of financial services executives are planning to increase their investment in AI – but most AI projects never make it out of the concept/planning stage. Despite how daunting moving from linear models to advanced AI models can seem, it is possible to implement AI and see results in under 60 days.

Check out our cheat guide to leveling up your risk decisioning with AI.

Get the Cheat Sheet

LATEST BLOGS

Continue reading

Replacing Your Legacy Credit/Loan Application Processing Software

Written by Allison Karavos on . Posted in Blog. No Comments on Replacing Your Legacy Credit/Loan Application Processing Software

BLOG

Replacing Your Legacy Credit/Loan Application Processing Software

Your business has moved on, did your processing solution keep up?

Before The Gap was an international clothing brand, it was a small record store in San Francisco’s Lakeside district.

Similarly, tech giants LG got their start selling cosmetics, toothpaste and other personal hygiene products.

That’s right.

LG was originally the Lak-Hui Chemical Industrial Company.

There are dozens of stories like these. Businesses grow and evolve over time. And while you might never pivot as dramatically as The Gap or LG, your products and services have and will continue to evolve to meet new market demands.

It goes without saying, but if you shift your offering, expand into new markets, or even grow, your current software may no longer meet your needs. And while it can be tempting to try and adapt existing technology to meet current business requirements, it’s often like trying to fit a square peg into a round hole. When assessing the long-term feasibility of your existing loan application processing solution ask the following questions:

  1. What’s the cost of maintaining the current system?
  2. How much will it cost to make significant changes to meet new business needs?
  3. How long does it take to make changes?
  4. Is it making you less competitive?
  5. Do you rely on the vendor to make key updates?

Over time, keeping your software operating smoothly will cost much more than investing in new technology.

Don’t believe it?

Consider this. Outdated technology cost businesses $1.8 trillion in wasted productivity in 2016.

Is your software making you more, or less, competitive?

Can your current solution:

  • Be adapted to new business processes?
  • Support a growing number of users?
  • Automate repetitive tasks?
  • Handle operations on a bigger scale?
  • Power a first-class consumer experience?
  • Enable business users to make changes quickly?
  • Make integration to data sources and other tech solutions easy?

    Your credit application processing solution should power not impede business growth and help make you more competitive. If you’re constantly fighting the system to make changes, waiting on integrations due to complex coding, or sacrificing the consumer experience because the system can’t support instant approvals, then it’s time to make a change. Why? Because, if you can’t make changes quickly your business is exposed to increased risk and missed opportunities.

    Consumers demand instant decisions and the best user experience. For today’s tech savvy customer making them wait more than a few seconds for a loan decision is like expecting them to go back to the days of dialup internet. While it used to be fine to wait a minute for the internet connection to kick then another minute for a website to respond, it’s now considered slow if a website isn’t visible in just a couple of seconds. If you continue to use the ‘dialup’ of loan application solutions, expect your customers to have found an alternative option before the modem even starts to warble!

    Telltale signs your credit/loan application processing system is past its sell-by date include:

    • You rely on your dev team to make simple changes
    • Making sure it works properly is becoming increasingly expensive
    • Waiting on changes is slowing down business growth
    • It can’t scale to meet your business needs
    • It’s preventing you from making real-time decisions
    • Tie-dyed t-shirts, leisure suits, and mullets were acceptable fashion choices when you first started using the software

      What should you look for in a replacement?

      The benefits of replacing a legacy system far outweigh the temporary inconvenience of implementing a new loan application processing system, but how do you know which replacement solution to select?

      Here are five key things to look for in a replacement:

      1. A low-code solution – Low-code solutions allow you to configure, maintain and even create new processes without having to rely on your dev team. Instead, you can drag and drop different components to make changes quickly and easily. The right low-code solution can reduce or eliminate the delays caused when business teams have to rely on over-burdened dev teams or the solution vendor to make updates.
      2. Simplified integration capabilities – Integration, whether it’s to internal or external sources, is a challenge for many businesses but it shouldn’t be. Your credit application processing solution should make integration easy, so when new integrations are needed, which they will be, the reliance on dev involvement will be minimal and business users can take the lead.
      3. Advanced automation options – Process automation is a vital component to powering business growth and ensuring a first-class customer experience. Your new solution should make it easy to automate processes and also enable you to reuse automation components across multiple business processes.
      4. Scalability – You would never invest in a one bed property if you knew you’d need something bigger in a few weeks. So, why treat a processing solution any differently? If you’re investing the time and money in changing solutions you should choose one that you can keep for many years, which means picking one that is able to scale as your business grows!
      5. Flexibility – It’s impossible to predict what changes your business will need to adapt to in the future, so your credit application processing solution needs to be flexible enough to allow your business to remain agile. For example, Provenir’s simple drag and drop interface, allows you to build new tools easily when you need them, allowing the business to respond to changing markets and take advantage of new opportunities as they arise.

      Wrapping Up

      Saying goodbye is never easy. But when you find a credit application processing system that configures to your needs with minimal coding, integrates at lightning speed, and that actually makes your life simpler, you won’t regret making the change.

      After all, the (first) end of Michael Jordan’s basketball career brought a triumphant return, a record-breaking winning season, three championships (and how dare we forget about Space Jam). And, the end of Genesis, as the world knew it, brought Phil Collins’ solo career. What would the world be without that rendition of True Colors?

      Endings are just the opportunity for something new. It’s time to take the leap!

      The Ultimate guide to Decision Engines

      What is a decision engine and how does it help your business processes?

      Take a look

      LATEST BLOGS

      Continue reading