Hey Fintechers and Fintech newbies 👋🏽

We’re living in a transitional time in the financial services industry and the world as a whole. 

Costs are rising. Regulatory frameworks and policies are in flux. Traditional business models are being broken. New technologies have emerged and are being adopted faster than in previous eras. And, consumer expectations require more. More personalized, more seamless, more connected, more intuitive, etc. 

It’s a challenge to predict the next few months of innovation, let alone nail the classic “where do you see yourself in 5 years?” question for fintech. 

But, despite the pace, problems, and possibilities of today, there is one constant that will continue to underpin everything that we do:

Data. 

I’ve referenced data in pretty much all of my previous newsletters, but this is the first dedicated data deep dive (that will be the limit of the alliteration). 

Data has been, and will continue to be, an invaluable commodity. But many organizations, financial institutions and fintechs included, fail to effectively tap into and leverage the data they have to its full potential. 

That’s why I’m covering data in some depth in this edition and here’s what to expect:

• The Anatomy of Data in FS
  • The Fundamentals of Data
  • The Data, Information, Knowledge and Wisdom Pyramid
• The Value of Data in Financial Services
  • Smarter & Faster Credit Risk decisions
  • Fraud Prevention
  • Embedded Finance
  • Growing Deposits
  • Customer Engagement
• Challenges and Myths of Financial Services data
  • Connecting and Collecting Data
  • Normalisation
  • Privacy and Permissions
• How MX Helps Solve Data Challenges and Generate Value
  • Connect – Enable secure data sharing and verifications
  • Analyze – Transform data into actionable insights
  • Engage – Deliver personalized experiences
  • Act – Drive deposits, loyalty, and long-term revenue
• The Trends that Will Shape the Future of Data
• The Infinite Data Hack
• What Banks, Fintechs, Credit Unions, and Others Should Do Now

Zero to One: The Anatomy of Data in Financial Services 0️⃣1️⃣0️⃣

Let’s start with some fundamentals.

At its core, data is a symbolic representation of a fact, condition, event, or value. 

It can be:

• Quantitative (numbers like 98.6 or $500.00)
• Qualitative (words like “pending” or “overdue”)
• Binary (yes/no, true/false, 1/0)
• Temporal (time or date)
• Spatial (location)
• Multimedia (audio, images, video)

By itself, data doesn’t say much. But, with structure, analysis, and context, it becomes the foundation for knowledge, strategy, and action. 

In the same way that raw ingredients like flour (once milled) and sugar (once refined) cannot stand on their own as a meal, data has to be brought together and transformed into something. Flour, sugar, and other ingredients combined in the right way and cooked at the right temperature can become a birthday cake, brownies, or other tasty treat. (Is anyone hungry yet?) Once data is collected, enriched, analyzed, and put into action, then you’re cooking.  

Raw financial services data, just like raw recipe ingredients, come in different forms and require different things to become something worthwhile. For instance, how many of us have seen a transaction on our statements that we have trouble even deciphering? A random string of letters can cause more harm than good, creating unnecessary calls from consumers wondering what’s going on or worse, panicked consumers who are worried they are victims of fraud. 

Raw data is messy. It’s often unstructured, duplicated, incomplete, and inconsistent. It tells you what happened, but not what it means or what to do about it.

To become useful, this data must go through a multi-stage transformation:

• Captured (from bank systems, third-party apps, customer interactions)
• Enhanced (cleaned and categorized into understandable formats and consistent models, e.g., merchant names, spending categories, locations, etc.)
• Contextualized (with behavioral patterns, sentiment, historical norms)
• Activated (used to trigger insights, recommendations, or automated decisions)

Even that bank statement view, which some might think is a basic grouping of structured transaction data, goes through a transformation.

• Transaction data is pulled in. 
• Merchant names are standardized and mapped.
• Dates are normalized to ensure consistency across time zones and formats.
• Metadata, such as merchant logos, locations, or type of service, is enriched.
• Recurring payments are detected to flag subscriptions, bills, or salary deposits.
• Anomalies are surfaced, like duplicate charges or unusually high spending.

Data is valuable, but only when it’s given context (converting it into information), meaning (converting it into knowledge), and put into action to make intelligent, informed decisions (wisdom).

This is where the true value of data lies and it’s why so many organizations struggle to effectively use the data they have.

The best way to demonstrate this progression is using a model known as the DIKW pyramid, or my bespoke and catchier version, the DIKI pyramid. 

There are several other layers that go into this basic model to understand what needs to happen to transform raw transaction data into intelligence and ultimately, action. Here’s a simplified example using a classic transaction you might see on your bank statement:

📄 Raw Data: AMT:16.45|CCY:USD|MRCHT:Starbucks|DateTime:090420260945|Type:Debit

🔍 Enhanced Data: “$16.45 spent at Starbucks at 9:45am on 9th April.”

📊 Information: “You’ve spent $112.20 this month at coffee shops.”

📈 Knowledge: “That’s 40% above your average. Coffee spending is trending up.”

🧠 Intelligence: “If you want to save more for your upcoming holiday, reducing discretionary coffee spend could help.”

⚡ Action: Click here to automatically move unspent dollars at the end of the month into a savings account. 

Coffee spending is an overused and clichéd example, but it's a good one in this context because it inadvertently demonstrates the value of the effective transformation of data into knowledge and wisdom as it describes the core function of a sub-sector of fintech. 

Personal Finance Management. 

When you can bring together data from various sources, enhance it, give it the correct context, and give customers the ability to make better decisions, you have the workings of a Personal Financial Management (PFM) tool. 

Great PFM tools are more than just a record of where consumers have spent and plan to spend. They give personalized and informed actions to help customers reach their financial goals. But unfortunately, the average ones end up just aggregating transaction data and playing in that unactionable information layer.

The ones that use data to help consumers put financial management on autopilot and give them actionable insights and knowledge will win. The rest will fall by the wayside.

PFM rant over for now.

Another example of taking data through these stages is something a bit smarter:

📄 Raw Data: AMT:5400|CCY:USD|MRCHT:Citibank Ltd|DateTime:010420260900|Type:Credit

🔍 Enhanced Data: “You received $5,400 from Citibank Ltd”

📊 Information: “Your total annual income from Citibank Ltd is $64,800”

📈 Knowledge: “Your annual income is up 15% from last year, but your expenditure has not increased.”

🧠 Intelligence: “You should sweep some of this excess income into an inflation-beating savings account to reap the benefits over a year.” 

⚡ Action: “Click here to sweep 40% of the excess funds into an easy-access savings account and set up a regular transfer”

When you look at the difference between the original raw data and the transition to an intelligent outcome, it becomes clear how powerful data can be for consumers and financial institutions.

That single data point becomes a financial mentor which can lead to positive behavioral change, and more funds in the banks’ and customers’ accounts.

And, that’s just one use for that data point. There are many more.

 What starts as raw, disconnected numbers becomes the basis for insight, advice, and action. 

Not just a record of what happened, but an indicator toward what should happen next. 

This is the real alchemy of financial data: when connected, enhanced, contextualized, analyzed, and activated, data becomes a dynamic tool for change, whether that’s better budgeting, smarter investing, or faster lending decisions.

And, once data reaches this level of utility, it becomes more than just a byproduct of financial activity.

It becomes the most critical input.

The raw ingredients to make something truly special.

Why Data is the Hottest Commodity in Financial Services 📀

Data is a valuable commodity but, by itself, has no intrinsic value. 

It becomes valuable only when it’s organized, analyzed and contextualized, and it's the power of what you can do with data that gives it value. 

Here's a great real world example of that:

Yes, 4% of Americans are still asleep at 12pm.

Once you understand how to turn data into action, there are countless practical applications and benefits for both consumers and organizations across financial services including… 

Smarter & Faster Credit Risk decisions 🧠

Leveraging alternate data for credit decisions rather than relying purely on bureau data improves the credit decisioning process. Insights like identifying consumers who consistently paid rent, utilities, and mobile bills on time — which are often not part of credit reporting — can allow financial providers to lend more confidently to applicants with “thin file” credit reports. This leads to the right borrowers being approved, lower default rates, increased loan volume, and greater financial inclusion.

Fraud Prevention 🥸

With the right data, financial institutions can more effectively spot and prevent fraud. Anyone who says they are using AI to solve challenges with fraud in financial services are, in fact, using data to do the pattern recognition and create predictive algorithms. Building cohesive, AI-powered fraud models in this day and age relies on pulling in the right data points, which goes beyond pure transaction data to include geolocation, verified IP addresses, mobile device data, and more to detect and prevent fraud in near real-time.

Embedded Finance 🔌

Having worked on an embedded finance proposition myself, I have firsthand experience of the value of data in delivering embedded experiences. Small businesses who log into their accounting platform can see relevant finance options based on historic cash flow data to understand an incoming gap in cashflow and trigger action. Data -> Information -> Knowledge -> Intelligence in action again. 

It’s the same for banks embedding a fintech provider’s finance journey into their stack. Or a retailer like Shopify surfacing finance options for their retailers. Embedded finance (and embedding anything for that matter) relies on data to surface those options to the right people at the right time. 

Growing Deposits 🏦

Banks and credit unions are increasingly turning to behavioral data to grow and retain deposits. By analyzing spending, income patterns, and idle funds held at competitor institutions (via consumer-permissioned account aggregation tools), financial institutions can identify deposit attrition risks, spot cross-sell opportunities, and personalize savings incentives. Using the example I outlined before, understanding changes in account activity such as an increase in salary, spotting cashflow abundance, and using that to encourage savings is another great win-win situation using data. 

Customer Engagement That Feels Human 🫱🏽‍🫲🏼

This is where the real magic happens. Transaction data enrichment, real-time categorization, and contextual financial insights can form the basis for meaningful customer engagement, and lead to higher engagement, stronger customer loyalty, and business success.

Instead of bombarding users with generic alerts (“Your statement is ready”), financial institutions can now say:

• 🧠 “You spent 40% more on takeout this month. Would you like to set a budget?”
• 🧠 “Your subscription to Spotify increased. Do you want to review your recurring expenses?”

These small, data-driven touchpoints create a sense of understanding, support, and partnership, the kind that drives higher app engagement, product uptake, and long-term loyalty.

However, if utilizing data and transforming it into action was easy, everyone would be doing it. Well, there are key challenges in effectively leveraging data that banks, credit unions, fintechs, and many other organizations face… 

The Hard Truth About Financial Services Data: Myths and Challenges 🤨

Collecting and Connecting Data from Various Sources 🔌

Data is often siloed across systems, subsidiaries, products, and even teams. A customer may have a mortgage, savings account, and credit card with the same institution, but those datasets are often held in different systems by different departments. Now, factor in the myriad of financial accounts that customers have at other financial institutions. Most financial providers only see a small portion of a customer’s financial picture. 

But, through aggregated accounts and open banking APIs, financial providers are tapping into new permissioned data sets each day to gain a more complete picture of their customers. However, this also makes utilizing that data even more complex. 

The Normalization Challenge 💽

This is one of the biggest blockers to delivering great data experiences. Even when you manage to pull in data from multiple sources, it doesn’t come neatly packaged. It’s inconsistent, poorly labeled, and often miscategorized or missing vital context.

Take merchant categorization as an example. One bank might label Uber as "Travel", another as "Ride Sharing", and a third as "Other". Multiply that inconsistency across millions of transactions, and you get unreliable insights.

Even collecting basic transaction data from other banks via Open Banking APIs can be inconsistent. Different banks may return slightly different data formats, limited history, or missing metadata. You’ve “collected” the data, but it’s not yet usable.

Data normalization is the quiet workhorse of data enhancement and intelligence. Without clean, consistent, and categorized data, your dashboards are misleading, your insights are inaccurate, your AI models are underfed, and your recommendations are irrelevant.

Privacy and Permissions 🔐

A key part of the Open Banking mandate is to give customers ownership and control of their own data, enabling them to allow third parties (which includes other banks and fintechs) to securely access their permissioned financial data to get better products and services, experiences, and insights. In theory, this sounds simple. In practice, it introduces a new layer of complexity.

Today’s Open Banking ecosystem varies from country to country with the United States primarily market-led until the formal adoption of new regulatory requirements from the Consumer Financial Protection Bureau (CFPB) under Section 1033 of the Dodd-Frank Act go into effect. Under Section 1033 as it’s written today and in other parts of the world, customers must explicitly consent to share their data with third parties for specific purposes. That consent has to be managed securely, time-boxed, and revocable. This means institutions not only need the technical infrastructure to ingest and act on this open banking data, but also a consent management framework that respects consumer privacy, meets regulatory standards, and provides transparent control to the end user.

But, this isn’t just a compliance box to tick. It’s a trust issue. If permission flows are clunky, confusing, or opaque, customers will hesitate. And, if institutions misuse or fail to act on the data they’ve been granted access to, they’ll damage that trust.

“Consumers want their providers to know them — and they want providers to use their permissioned data to create better experiences. We have an opportunity and obligation to consumers to use their permissioned data responsibly to gain better insights about them and create targeted experiences that help them improve their financial outcomes.” says Jane Barratt, MX’s Chief Advocacy Officer.

Organizations will also often have to periodically refresh this consent, which means building a compelling enough case to get customers to share their data, prove it was worth it, and ensure that trust and understanding continues to be upheld. 

It’s a very tricky balance but one that is well worth it. 

MX Marks the Spot 🏴‍☠️

Tackling these data challenges and building the infrastructure to transform data into action is incredibly difficult. 

The challenges we’ve covered earlier — fragmentation, normalization, permissioning — are complex and time-consuming to solve alone.

That’s where MX, (other Financial Data Specialists are available), really adds value.

MX’s solution stack is organized into key value areas, each addressing a different step in the data-to-intelligence and intelligence-to-action journey. This isn’t just about pulling in data — it’s about making it usable, actionable, and valuable for both consumers and financial providers. 

Here is a brief overview of how MX solves data challenges.

Connect – Enable Secure Data Sharing and Verifications

This layer focuses on enabling secure, reliable access to external accounts and real-time verification to power onboarding, personalization, and money movement.  It helps consumers gain a 360-degree view of their financial lives, while giving financial providers the visibility needed to assess customer needs, verify accounts, and support onboarding or underwriting workflows.

Products/Services include:

• Account Aggregation – Enables consumers to link external accounts to build a unified financial picture.
• Account Verifications – Instantly verify account ownership, routing, and balances in real time.
• Data Access – Open Banking APIs for secure, compliant access to permissioned consumer financial data with full consent control.

Analyze - Transform Data into Actionable Insights

Once you bring together data, you have to make it understandable and meaningful. This layer enhances and organizes raw transaction data into clean, categorized formats, allowing both consumers and providers to understand spending habits, investment behavior, and financial health at a glance.

This step is critical to move from “we have data” to “we understand it.”

Products/Services include:

• Data Enhancement – Categorize, normalize, and enrich transaction data for clarity and usability.
• Investment Data Enhancement – Make portfolio data consistent, clean, and insightful across providers.
• Customer Analytics – Surface deeper insights with intelligent models, dashboards, and analytics tools to identify segments, trends, and behaviors to optimize targeting and messaging.

Engage – Deliver Personalized Experiences

Once data is analyzed, it can be used to engage users with more personalized, insightful experiences. This layer brings the intelligence back into the hands of customers via mobile and digital platforms, and helps providers engage with users more effectively.

It’s the difference between generic, one-size-fits-all experiences and personalized banking experiences

Products/Services include:

• Mobile Banking – A customizable digital banking experience built to increase engagement and retention.
• Financial Insights – Personalized, contextual nudges and recommendations to guide users on their financial journey(e.g., "You're spending more than usual on subscriptions").
• Personal Financial Management – Tools to help users set budgets, track goals, and improve financial wellness.

Act – Drive Deposits, Loyalty, and Long-Term Revenue

This is the final step: turning insights into measurable business outcomes.

Whether it’s helping a customer switch direct deposits, identifying accounts at risk of churn, or nudging them toward relevant offers, this layer is where data powers real growth. It’s where customer action aligns with business objectives.

Products/Services include:

• Customer Analytics – Predict behavior, spot risks (like attrition or dormant accounts), and personalize outreach.
• Direct Deposit Switch – Reduce friction and retain customer income flows by making it easy to move salary deposits into your institution.

With expanded data and clearer understanding, you can understand your customer’s goals and you can be there for them helping them in their financial lives. This is the power of MX. We are a data company. -Wes Hummel, MX Chief Product and Technology Officer

Collect, Connect, Process and Analyze data, then use that to Engage and Act, driving better outcomes for customers.

Using an expert partner or bringing in data strategists and experts is becoming even more important when you look at what the future holds where the value and volume of data is increasing exponentially…

The Future: Moore Data, Mo’ Problems? 📈

Now, the value of understanding and transforming data into actionable insights is clear. 

This is the perfect primer because this understanding, along with a detailed look into the future (keep reading), should give banks, credit unions, fintechs, and anyone in financial services overlooking the value of data, ironclad reasons to rethink their position.

Artificial Intelligence 🤖

No article, podcast, product release, etc. is complete without mentioning AI. In this case, it’s not for clicks. The GenAI tools the general public has been experimenting with, powered by Large Language Models, have all been trained on massive amounts of structured and unstructured data.

In Dec 2024, ChatGPT was the 5th most downloaded app globally, a signal not just of novelty, but of mainstream adoption. 

AI adoption is only going to accelerate. With it comes a growing need for high-quality, permissioned, domain-specific data, particularly in financial services, where trust, accuracy, and context are non-negotiable.

The real challenge will emerge as AI companies face greater scrutiny and restrictions around web scraping and open data access. This means that first-party financial data, collected with consent, cleaned, and enriched, will become a strategic moat. It’ll be the fuel that powers everything from intelligent assistants and predictive analytics to next-gen fraud prevention and hyper-personalized financial coaching.

Just as digital banking made mobile banking the norm, completely transforming how customers interact with banks, AI has transformed what it means to receive real-time, responsive access to intelligence. Banks will have to keep up with these changing expectations and deliver AI-powered products. And the only way to build AI-powered financial products that are actually useful is to start with solid data foundations.

Fraud 👮🏼‍♂️

Using data to better understand customer behavior and spot if an account has been hijacked and prevent it in the first place is the future. As financial products become more digital, more embedded, and more real-time, the surface area for fraud is expanding. Fraudsters are no longer just exploiting weaknesses in identity checks or payment rails. They’re using AI themselves to mimic behaviors, spoof identities, and hijack trust signals.

To stay ahead, financial institutions need to understand their customers more deeply than ever before. That’s where data comes in.

By using data to track behavior over time, institutions can build dynamic profiles of normal activity for each user, how they spend, where they transact, which devices they use, and what time of day they typically log in. So when something deviates, say a new device in a new country suddenly transferring funds at 3am, the system knows it’s off.

But more than that, the best fraud prevention is proactive, not reactive. It doesn’t just respond to anomalies. It predicts potential threats. That’s only possible when you feed models with clean, enriched, and diverse data: transaction history, geolocation, device metadata, behavioral biometrics, and more.

In the coming years, we’ll see a major shift from rule-based fraud detection to AI-powered fraud prevention engines that are constantly learning, adapting, and refining themselves based on streams of real-time data. And the most effective systems will be those that have access to deep, cross-channel, permissioned data—not just transactional, but contextual.

Because ultimately, you can’t prevent what you can’t see. And you can’t see clearly without trusted, structured, high-quality data.

Moore’s Law of Tech Acceleration 📈

The most underestimated trend, for me, is the perpetual and exponential growth of data.

Technological advancement is often measured in Moore's Law. The original principle formulated in 1965 was an observation from Gordon E. Moore, co-founder of Intel, that the number of transistors on a microchip would roughly double every two years at minimal cost. It’s a measure of physical, technological acceleration, but it’s been used as a proxy to measure the growth of all kinds of tech. .

Although the volume of data generated hasn’t followed Moore’s Law of doubling every two years, that could be about to change. It’s estimated that 90% of the world's data was generated in the last two years alone. 70% of that is user-generated data. And forecasts expect the volume of data to double from around 182 zettabytes (1 zettabyte is a trillion gigabytes) to 364 ZBs by 2028 fueled in part by the ‘Internet of Things’ (IoT), which refers to the growing network of connected devices and sensors generating and exchanging data online.

In financial services, this data growth shows no signs of slowing down.

More customers.

More channels.

More devices.

More sensors.

More data.

But more data doesn’t automatically mean more insight. In fact, without the right infrastructure, governance, and intelligence layers in place, more data often just means more mess.

That’s why getting a grip on data, understanding it, organizing it, connecting it, and making it useful is no longer optional. It’s a strategic imperative.

With this increased breadth and depth of data, companies that leverage telematics (in auto), wearables (in health), spending habits (in life insurance), or transaction data (in loan approvals) can redefine how products and services are built and made available to consumers (insurance premiums get smarter, credit decisioning gets more inclusive, and more). Instead of relying on proxies like age or zip code, they can assess real behavior and patterns to make better business decisions and deliver better products.

The list of better outcomes for customers and financial services providers is nearly endless. 

So yes, 'Moore' data does mean Mo’ problems.

But only for those who don’t start treating their data like the asset it really is.

The Infinite Data Hack ∞

So, what should banks, fintechs and others do to stay ahead?

The obvious one is to leverage an expert firm that specializes in collecting, connecting, engaging, and acting on financial data. 

Ideally, organizations should aim to design a self-sustaining loop. 

Pull in data → Enhance it → Contextualize it → Derive meaning → Trigger an action → Generate more data → Pull in that data → Enhance it and so on. 

That’s the real magic: using data not as a static report, but as a feedback loop that continuously improves outcomes. The better your organization is at understanding behavior, predicting needs, and acting intelligently, the better the outcomes and the richer the data you feed back into the system becomes.

That’s true data alchemy. 

Turning raw data into gold, and turning that gold, into more gold. 👑

So Where Should You Start? 🏃🏽‍♂️

My advice to anyone starting out on this journey is, ironically, to not look at the data first. 

Data fatigue is real. Looking at a bunch of entity relationship diagrams and running SQL queries across databases is not the best route to start the journey.

The best way to start, especially if you're new or have data fatigue, is to look at the objective you want to achieve and work backwards. 

Look at defining the goals and outcomes that matter to the wider business and the specific product lines.

Ask:

• Are we trying to increase credit card retention?
• Are we aiming to boost deposits from high-net-worth customers?
• Do we want to lower default rates on short-term loans?
• Are we aiming to improve financial wellbeing through better insights?
• Are we trying to speed up the end-to-end lending process?

Once the objective is clear, then you can work backwards.

• If you're trying to increase credit card retention, what knowledge do you need? 
• The largest causes and demographics of churned customers? 
• Then the specific breakdown of customers? 

Then what are the specific data points that enable you to build that breakdown, build out a picture of the behaviors that lead to churn and specify actions to address them? 

Starting with a clear objective, then working backwards to the specific data is a great step, even for data-literate companies.  

I’ll close with this poignant 1-liner about data that I use when explaining its value:

Data is the digital shadow of reality. Bits and signals capturing human behaviour, market movement, and the world's complexity. - Jas Shah

That’s why data is vital to positive outcomes for customers and companies alike.

Not because it holds innate value on its own, but because when refined, structured, and stitched together, it forms a mirror of the real world.

A digital shadow that helps organizations better understand human behavior, customer needs, and shifting market dynamics. It’s this understanding that enables better segmentation, stickier products, and more relevant decisions.

The organizations that succeed in creating a rich, real-time picture of these digital shadows will be the ones that deliver personalized experiences, faster credit decisions, smarter fraud prevention, higher payment conversion, and stronger deposit growth.

And if you’re making no headway using data to build these digital shadows of human behaviour and real-world actions, you’ll likely be left behind.

That's it from me. Hope you enjoyed this deep dive and guest write up 👋🏽

If you enjoyed this edition, hit the thumbs-up button below and reply to the email if you have more feedback or words of adulation (it’ll get back to me eventually). In the meantime, connect with me on LinkedIn and check out more of my fintech deep dives at my regular newsletter.

Jas.