Bank Director’s 2025 Technology Survey sends a loud signal: banks are still struggling to turn technology investments into business value. The survey reveals that one third of bank leaders admit their institution’s inability to use data effectively is a top challenge.

This isn’t just another operational headache. Banks are being squeezed from every side, fintechs moving faster, regulators demanding more, and customers expecting seamless digital experiences. If a bank can’t put its data to work, it’s not just falling behind, it’s putting its future at risk.

1. The numbers back it up:
2. The Illusion of a Single Customer View
3. Elevate Data Strategy to the Boardroom – With Teeth.
4. Build a Shared Semantic Layer – The Universal Translator.
5. Reduce Dependence on Spreadsheets with Governed Automation.
6. Stop Defending, Start Attacking: Monetize Your Data.
7. Apply the “Data Authority Test” for New Technology.
8. Path A: The IT-Centric Path. 
9. Path B: The Risk & Compliance-Led Path. 
10. Path C: The Business-Led Path. 
11. References:

The numbers back it up:

• 56% of banks still keep data siloed in the system where it was generated.
• 56% rely on their core provider to access data.
• 41% still manage business-line data in spreadsheets.
• 39% have deployed data lakes or warehouses, but adoption is uneven.
• Only 18% measure ROI on technology projects.

This is not simply a technology shortfall. These are the fingerprints of deep data misalignment.

The Illusion of a Single Customer View

From my experience advising financial institutions, the root cause is rarely a lack of tools, but a dangerous illusion of control.

I was once in a meeting with a proud CDO of a major African bank who showed us a dashboard claiming a “360-degree view of the customer.” Out of curiosity, I asked a simple question: “For a customer who has a mortgage, a current account, and a wealth management portfolio with you, how many different ‘customer records’ exist in your primary systems?”

The room went quiet. After a week, their team came back with the answer: seventeen.

Seventeen different versions of the same person, each with slightly different spellings, addresses, and risk profiles. This is the reality. The problem isn’t a lack of dashboards; it’s a fundamental failure to create a single, trusted source of truth before the dashboard is even built. That dashboard wasn’t providing insight; it was providing an average of 17 different lies.

The Real Problem: Three Layers of Data Debt

Banks aren’t just facing technical debt; they’re drowning in Data Debt, which is far more corrosive. This debt accumulates in three layers:

1. Technical Data Debt: The legacy cores, bolt-on platforms, and countless third-party systems. This is the easiest to see and fund, but it’s just the tip of the iceberg.
2. Semantic Data Debt: The fact that “assets” means one thing in wealth management and another in commercial lending. This is the layer that causes massive reconciliation costs, blocks real-time decision-making, and makes AI initiatives impossible to scale.
3. Governance Data Debt: The lack of clear ownership and accountability. When regulators ask, “Where did this number come from?” and the answer is a spreadsheet, you have a governance blind spot. This is the most dangerous layer, leading directly to regulatory fines and catastrophic decision-making.

Most banks are throwing technology (solutions for Layer 1) at a problem that is rooted in Layers 2 and 3. It’s like buying a faster car for a driver who doesn’t know the rules of the road or their final destination.

What Banks Must Do Differently: A Blueprint for Alignment

The solution is not more technology. It’s a modern data strategy grounded in governance, alignment, and a shift from defense to offense.

Elevate Data Strategy to the Boardroom – With Teeth.

Data is not an IT issue. It’s a business issue that impacts risk, compliance, revenue, and customer trust. Banks must treat data as a board-level agenda item, with a C-level executive (CDO or otherwise) accountable for outcomes and for liquidating the three layers of data debt.

Build a Shared Semantic Layer – The Universal Translator.

Stop the reconciliation nightmare. A shared semantic layer, tied to master data management (MDM), is the universal translator that ensures every system defines “customer,” “account,” and “product” the same way. This is the non-negotiable foundation for any successful AI or analytics program.

Reduce Dependence on Spreadsheets with Governed Automation.

Replace manual reconciliation with governed, automated data pipelines. This isn’t about banning spreadsheets; it’s about creating such good, accessible, and trusted data products that teams no longer need to retreat into their own manual, error-prone spreadsheets.

Stop Defending, Start Attacking: Monetize Your Data.

The conversation is always about defense: compliance, risk, and reconciliation. But the real prize is offense. A clean, unified data foundation allows you to launch products a fintech can only dream of.

Imagine:

• Using real-time cash flow data from a business current account to pre-approve and instantly offer a short-term loan at the precise moment it’s needed.
• Creating a truly unified view of a retail customer’s holdings to offer hyper-personalized savings or investment products that don’t compete with themselves.
• Partnering with a retailer by leveraging your combined, consented data to create targeted loyalty offers.

Fintechs win because they start with clean data. Legacy banks have vastly more data but can’t wield it. Solving the data problem isn’t just about survival; it’s about unlocking revenue streams that are currently trapped in your silos.

Apply the “Data Authority Test” for New Technology.

Before adopting any new tool, ask:

• Can it integrate cleanly with our semantic layer, or does it create a new dialect?
• Does it support compliance and regulatory requirements by design?
• Does it contribute to trusted, governed data, or is it just another silo in disguise?

The Fork in the Road

Bank Director’s survey is the latest warning sign. Bank leaders now face a fundamental choice:

Path A: The IT-Centric Path. 

You continue to fund “data lake projects” and “AI initiatives” without first solving the foundational data problem. You will spend the next five years with the same survey results: disappointed by ROI, frustrated by silos, and watching your market share slowly erode to nimbler competitors.

Path B: The Risk & Compliance-Led Path. 

You invest heavily in data governance, lineage, and controls purely to meet regulatory demands and avoid massive fines. This path makes you safer and more auditable, turning your compliance function from a cost center into a managed one.

However, it’s a defensive play. You will build a robust, clean data foundation, but primarily for reporting and risk mitigation.

The danger? You create a “compliance-only” data moat, failing to leverage this new asset for growth and allowing more innovative competitors to poach your customers with superior experiences. You survive, but you don’t necessarily thrive. And non-risk players in your business may turn away from data altogether.

Path C: The Business-Led Path. 

You declare data a core asset and make a leader accountable for liquidating your data debt. You invest in the unglamorous work of governance, master data management, and a semantic layer first. You tie every single tech investment to a specific business outcome, like “faster loan origination” or “reduced customer churn.”

The survey is clear. The market is clear. The technology exists. Path C is the only one that turns data into a competitive weapon. Path A is a slow decline, and Path B, while safer than A, is a missed opportunity. The only thing missing is the strategic will to treat data not as a byproduct of business, but as the business itself.

Which path will your bank choose?

References:

Bank Director’s 2025 Technology Survey