Joseph Ndungi Follow

Alternative Data – The New Frontier in Financial Intelligence

Introduction and Definitions

Alternative data emerges from unconventional collection methods and nonstandard sources. Investors typically need to examine factors outside company reports and broker statements to gain market advantages. Alternative data encompasses elements such as credit card transactions, social media commentary, product reviews, and satellite imagery.

Traditionally, banks have relied on conventional data: credit scores, employment records, tax returns, and formal income statements to assess financial health. But in an increasingly digital and mobile-first world, these are no longer enough.

Enter alternative data—non-traditional datasets that offer deeper, often real-time insight into consumer behavior. These include mobile payments, app usage, ride-hailing history, delivery transactions, utility bills, and even entertainment consumption.

2. Why It Matters

a. Financial Inclusion: Seeing the Unseen

Imagine a world where millions of people are financially invisible. This isn’t science fiction; it’s the reality for a significant portion of the global population, especially in developing nations. These individuals are often “credit invisible” – they don’t have traditional bank accounts, credit cards, or loan histories. To a conventional bank, they simply don’t exist in a way that allows for risk assessment or product offering.

However, these same individuals are far from inactive in the digital world. They are increasingly using a variety of platforms that generate a wealth of data:

Mobile Money (e.g., M-Pesa, MTN Mobile Money): In many parts of Africa and Asia, mobile money is the primary way people send, receive, and store money. Consistent use and healthy transaction volumes paint a picture of income and financial activity.
Peer-to-Peer (P2P) Payment Platforms (e.g., Venmo, PayPal): These platforms are popular globally for everything from splitting bills to paying for goods and services, revealing spending habits and social financial interactions.
Ride-Hailing (e.g., Uber, Bolt, Little): Frequent use of ride-hailing services can indicate regular commuting patterns, which often correlate with employment or business activities.
Food Delivery Services (e.g., Uber Eats, Glovo): Ordering patterns can provide insights into lifestyle choices and discretionary spending power.
Streaming Platforms (e.g., Netflix, YouTube, Spotify, Apple Music): Subscription payments and usage patterns demonstrate a level of consistent disposable income and engagement with digital services.

These digital footprints, though unconventional from a traditional banking perspective, are rich with valuable signals. By analyzing these alternative data streams, banks can finally “see” these previously invisible individuals. They can begin to understand their:

Cash Flow and Income Stability: Regular mobile money top-ups and consistent bill payments (even for services like streaming) can indicate a reliable source of income, even if it’s not a traditional salary.
Financial Responsibility: A history of on-time utility payments or consistent contributions to a mobile savings wallet showcases responsible financial behavior.
Lifestyle and Spending Habits: Analyzing spending on ride-hailing, food delivery, and entertainment can help banks understand an individual’s financial priorities and discretionary income.

By tapping into this alternative data, banks can break down the barriers to financial inclusion. They can confidently onboard customers who were previously excluded and offer them a range of tailored financial products, from basic savings accounts to micro-loans and insurance, ultimately empowering individuals and fostering economic growth.

b. Credit Scoring Reinvented: A Fairer Assessment

The traditional credit scoring models that have been the gatekeepers of lending for decades are facing increasing scrutiny. These models can often be:

Biased: They can inadvertently discriminate against certain demographics or individuals with non-traditional career paths due to their reliance on historical data that may reflect past societal biases.
Outdated: In a rapidly changing economic landscape, traditional models can be slow to adapt and may not accurately reflect an individual’s current financial situation or future potential.
Unavailable to Certain Segments: As discussed, those without a formal credit history are simply left out, regardless of their actual creditworthiness.

Alternative data offers a powerful solution to these challenges, enabling banks to build more dynamic and inclusive “behavior-based” risk models. Instead of relying solely on past borrowing history, these models consider a broader range of real-time behaviors, asking questions like:

How regularly do you top up your mobile wallet, and what are your average balances? This provides insight into income consistency and savings habits.
Do you have a history of defaulting on payments for utilities or mobile phone bills? This can be a strong indicator of financial discipline.
What is your monthly expenditure on essentials versus discretionary items, and how does this compare to your income inflows? This helps assess an individual’s budgeting skills and financial stability.

By incorporating these and other alternative data points, banks can create credit scores that are not only more accurate but also more equitable. They can identify creditworthy individuals who would have been overlooked by traditional methods, leading to more inclusive lending practices and a more robust financial system. This shift represents a fundamental reinvention of credit scoring, moving away from a one-size-fits-all approach to a more nuanced and personalized assessment of financial risk and potential.

3. Dominant Sources of Alternative Data

So where does this precious alternative data come from? Too often, it is produced by the digital services we use every day, possibly unknowingly. Let’s examine some of the most significant sources and how they can offer up:

a. Mobile Money and Digital Wallets (e.g., M-Pesa, Airtel Money, T-Kash)

In most parts of the world, especially here in Africa, mobile money is not a choice; it’s a way of life. These digital wallets register a huge amount of financial transactions:

What it shows: The value flowing in and out (transaction volume), how often they’re making or receiving payments (payment frequency), and how they’re utilizing it – sending money to family and friends (P2P), paying a bill (bill pay), or buying airtime (top-up).
Why banks care: It’s an excellent means of assessing a person’s real income, even when not working in the classical sense with a payslip. Repeated bill payments and frequent usage demonstrate financial discipline and a hidden income stream.

b. Payment Platforms (e.g., PayPal, Wise)

Platforms are the lifeblood for lots of online entrepreneurs, freelancers, and gig workers. They enable payments across borders and between businesses and individuals

What it means: The type of business one is involved with (merchant categories), whether they’re making or receiving money from other peers (P2P activity), and the frequency and extent of these interactions.
Why the banks care: It helps identify small business owners or gig economy workers who can have significant revenues that traditional banks would otherwise be unable to capture. It’s a peek into modern-day entrepreneurship.

c. Ride-Hailing Apps (e.g., Uber, Bolt, Little)

The way you travel can tell you a lot about your lifestyle and expenditure management:

What it shows: How often you use these services and how much you spend (frequency and cost), and when you travel. For instance, daily commuting during peak times tends to point towards stable jobs.
Why banks care: It helps them infer employment status and consistency. Someone ordering a ride-hailing service to the same business district each weekday likely has a full-time job.

d. Food & Delivery Services (e.g., Glovo, Jumia Food, Uber Eats)

Our eating habits upon ordering food can give us tiny clues about our money situation:

What it shows: The total money spent each month (monthly spend) and when they’re ordering – are they weekly weekday lunches (suggesting work-related ordering) or occasional weekend splurges (indicating discretionary spending)?
Why banks care: This is a way of measuring disposable income and lifestyle choices. It helps in understanding how much cash a person has left over after they’ve paid the essentials.

e. Utility Payments (e.g., KPLC, Nairobi Water)

On-time payment of your bills is an old-fashioned measure of fiscal responsibility, and computerized records make it easier:

What it shows: How prompt your payments are, and the type and quantity of services that you are regularly paying for (electricity, water, internet, etc.).
Why banks care: It’s a positive indicator of reliability and willingness to pay financial obligations. Paying your utility bills consistently, you’re likely a better lending risk.

f. Streaming Services (e.g., Netflix, Spotify, YouTube Premium)

Even your leisure activities can provide useful data points, particularly in terms of payment:

What it indicates: The punctuality of subscription payments. While tastes and behaviour patterns can be interesting, in finance, the payment record counts.
Why banks care: Consistent subscription payments, even small ones, signal some degree of financial stability. This knowledge can then be used for lifestyle segmentation, which helps banks to understand their customers better and, for instance, provide appropriate products or services (cross-selling opportunities), e.g., bundle deals or reward schemes.

4. Real-World Applications by Financial Institutions

This is not theory; financial institutions and financial technology (fintech) companies are already applying alternative data in meaningful ways, changing how they interact with customers and evaluate risk. Here’s a look at some real-world examples:

a. Loan Approvals for the Unbanked: Opening Doors

This is perhaps one of the most exciting uses, especially in nations like Kenya, Nigeria, India, and Southeast Asia. Millions of people who were previously “invisible” to banks can now get credit thanks to innovative lenders:

How it works: Companies like Tala and Branch, both very active here in Kenya, look beyond bank statements. They analyze data (with the user’s consent, of course) from smartphones – like SMS messages (to see M-Pesa transactions, for example), call patterns, and even the apps you use. Similarly, lenders like Carbon pore over mobile transaction data to assess if one can pay back a loan.
The impact: This allows them to immediately decide one’s creditworthiness and provide instant loans, typically beginning with small and going higher as trust is built. It’s a game-changer for individuals who need a little boost or small entrepreneurs who need funds for growth.

b. Credit Line Adjustments: A Dynamic Approach

Banks are starting to use alternative data to make more informed decisions about how much credit to offer existing customers. It allows them to react more quickly to changes in a person’s financial situation:

How it works: Reading between the lines. Imagine if someone suddenly stops taking their daily Uber rides to the office (ride-hailing frequency drops). That could signal a job change or job loss. Or maybe their mobile wallet balance patterns show a steady decreasing trend. Perhaps they’ve suddenly stopped paying for several subscription services.
The impact: These signals can prompt a bank to actively alter a credit limit – perhaps reduce it to prevent someone from accumulating debt they can’t handle, or even increase it if positive habits (like growing mobile savings) are identified. It’s all about making credit more applicable to the situations of everyday life.

c. Fraud Detection: Finding the Unusual

Alternative data adds another layer of security by allowing banks to identify and prevent fraud much sooner:

How it works: Banks build a picture of your typical digital behaviour. If your phone data usually shows you making mobile money payments in Nairobi, but suddenly there’s a transaction attempted from London or Johannesburg, that’s a red flag. Inconsistent patterns – like sudden, unusual app downloads combined with large, out-of-character payments – can trigger alerts.
The effect: This allows banks to suspend suspicious transactions in real time and investigate potential fraud, protecting both their customers’ accounts and their own networks.

d. Marketing and Cross-Selling: Smarter, More Relevant Offers

Using customers’ online lives, banks can introduce products and services that are truly relevant and appealing:

How it works: If a bank sees a customer consistently paying for Netflix and Spotify, they might be a prime target for a lifestyle-oriented credit card that offers entertainment rewards or cashback on subscriptions. They might be offered a digital-first banking experience or products suited to a tech-savvy, millennial demographic.
The impact: This moves marketing from generic, typically irrelevant, promotions to personalized recommendations. This leads to customers getting promotions that they will probably be interested in and banks being in a position to build deeper relationships by showing that they understand their clients’ needs and preferences.

5. Benefits of Alternative Data Utilization

Embracing alternative data isn’t a trend; it’s a strategic decision with tremendous advantages for financial institutions and customers alike. Here’s why it’s generating so much buzz:

✔ Broader Coverage: Bank the Unbanked

The biggest benefit is perhaps the ability to finally open up the finance doors to everyone. Historically, millions, especially in developing markets, have been excluded, simply because they lacked the ‘correct’ form of financial history. New information shatters this by providing an alternative means of considering financial potential. By looking at such factors as the use of mobile money or utility bills, banks are now able to see and serve consumers who were previously invisible, stimulating financial inclusion on a never-before-seen scale and unlocking vast new markets.

✔ Faster Decision Making: Finance at the Speed of Life

In today’s world, waiting weeks for a loan approval or credit consideration is painfully slow. Alternative data, which is largely digital and real-time, allows for amazingly fast analysis. Extremely advanced algorithms can examine these disparate data points virtually in a nanosecond. What this means is immediate loan approvals, quicker risk assessment, and more timely customer service. For customers, it means receiving the cash or services they need, at the exact time they need them, not weeks later.

✔ Personalized Services: Products Tailored for You

We have all been presented with generic products by banks that are not suitable for our life. With alternative data, banks can move past a one-size-fits-all model of banking. By having a clearer understanding of lifestyle, spending habits, and financial behaviors (from sources like delivery apps or streaming platforms), banks are better able to segment customers. This leads to financial services and products that are genuinely personalized to the needs, preferences, and life phases of a person, making banking more useful and meaningful.

✔ Risk Reduction: A Smarter, Safer Approach

Risk is the core of finance. Traditional models make a snapshot based on past behavior, but new data gives a streaming video. It allows banks to build dynamic, adaptive risk models. These models will be able to adapt almost in real time as a customer’s online profile changes. If someone’s financial behavior shows signs of stress (e.g., missing small payments or unscheduled spending), the bank can move quickly. Good behavior can be rewarded, too. All this gives us more accurate risk prediction, fewer defaults, and a smarter, more resilient financial system overall.

6. Challenges and Considerations: Navigating the New Frontier

Just because alternative data promises to be absolutely fantastic doesn’t imply that exploring this new frontier is a walk in the park. Financial institutions have to tread carefully, balancing innovation with responsibility. Here are the key things to keep in mind:

This is probably the biggest hurdle. Think about this: do most people really know that how often they get takeout or use which apps can decide whether they get a loan or not?

The Problem: A lot of consumers click “accept” on terms and conditions without understanding the entire extent of ways their online traces can be analyzed. This raises grave ethical issues.
The Requirement: Europe’s GDPR, California’s CCPA, and most relevantly to us here, Kenya’s Data Protection Act (2019), are filling the gap. These legislations demand unequivocal, transparent consent from the user. Banks must be transparent about what data they collect, why they collect it, and how they will use it. Building and retaining trust with the customer is the solution; if they lack it, access to that data will disappear.

b. Data Quality and Bias: Shunning New Pitfalls

Alternative data isn’t magic; it has its own flaws.

The Problem: Data is inconsistent – data from one app may not align with another, and trends can deviate wildly by region or demographic. More concerningly, there is a genuine risk of introducing new biases to the system. For example, could one be unnecessarily stigmatised as ‘high risk’ due to low mobile internet usage (e.g. due to the use of Wi-Fi or a poorly covered location), or a spending profile that reflects lower socioeconomic status?
The Requirement: Banks require strong systems to sanitize and validate data. Most importantly, they must continually check their algorithms to guarantee that they aren’t inadvertently discriminating against individuals. Fairness has to be baked into these new models from the very beginning.

c. Over-Surveillance: The “Big Brother” Issue

There’s a fine line to walk between extracting meaningful insights and making customers realize that they’re under constant surveillance.

Summary: The more and more data points banks collect – from where you are to what you buy and when you’re online – customers start to feel overwatched and constrained. This builds a backlash against digital services, with consumers not wanting to use them or give their data in the first place.
The Requirement: Banks need to be attuned to this “creepiness factor.” They need to care about using data that has actual value and risk, not merely bring in everything. Transparency yet again – the customer will be more apt to take data usage if they can see the value and feel it won’t be used in an intrusive way.

d. Standardization: Searching for a Common Language

Unlike traditional credit scores, which have (if imperfect) methods, alternative data is more like the wild west.

The Problem: There aren’t industrywide standards for collecting, processing, analyzing, and scoring alternative data. What might be a good signal to one lender is ignored or used differently by another. It becomes challenging to compare models, keep things consistent, and even for regulators to regulate the market properly.
The Need: As the industry matures, there is growing demand for cross-industry collaboration to develop best practices and common standards. This will foster trust, ensure fairness, and allow the potential of alternative data to be more consistently and responsibly exploited throughout the industry.

Conclusion

We’re standing at a really interesting time when it comes to money and banking. Traditional methods of examining financial records such as old bank statements are losing their significance. Instead, a new picture is emerging, painted by how we use our phones and the apps on them every single day. This is “alternative data,” and it’s showing us a whole new way forward.

Think about what this means, especially right here in Kenya. For so many people that traditional banks couldn’t really help before, this is a game-changer. It means they finally get a chance to be ‘seen’ financially, maybe get a loan for their business, or just manage their money better. It also means faster decisions – no more waiting weeks – and services that feel like they were actually made for you, not for some average person.

And this is just the start. The advancement of intelligent computer programs (AI) will enable improved comprehension of this data. The process of secure data exchange between applications and financial institutions will become more straightforward. The ‘super apps’ that enable chatting, paying, and ordering services will provide an even more detailed portrait of user behavior.

But, as we step into this new world, we have to be incredibly careful. This new information gives us a lot of power, and with power comes a huge responsibility. We must make sure we’re always being fair, protecting people’s private information, and building trust. This technology aims to elevate individuals and unlock opportunities instead of developing methods to exclude them.

At the end of the day, this isn’t just about data and technology. It’s about people. If we use these new tools wisely and with respect, we can build a financial future that works better for everyone, giving more people a fair shot.

Happy Coding!

References

25 May 2025

#Credit

« Understanding Functors and Monads in F# with Statistical and Financial Models Internet Security - How to Protect Yourself from Malicious Attacks »