Unlocking Growth with Strategic Data Management

During my panel discussion at Acquire or Be Acquired 2026, we tackled one of the most pressing challenges facing lenders today: transforming abundant but fragmented data into a strategic advantage. In a landscape rapidly reshaped by AI, competitive strength won’t come from collecting more data—it will come from curating it, governing it, and enabling teams to act on it with confidence and precision. Here are some of the insights that came from this panel that I hope you will find insightful.

The Data That Actually Moves the Needle

As lenders, we’re surrounded by signals. But the data that drives decisions and competitive advantage falls into three categories: Risk, Performance, and Opportunity.

• Risk is mainly focused on borrower and financial data like traditional financials, tax returns, spreads, cash flow, and collateral.
• When we think about performance, we are thinking of behavioral data that gives us insight into the loan on the books and its payment histories, account activity, and early warning indicators.
• Opportunities exist within the portfolio data and insights in the relationship as well — think pipeline metrics, concentrations, risk ratings, exceptions, and workflow performance. What are my borrowers not using, and what could they expand on?

Unify these streams and you stop seeing snapshots. You see patterns. And patterns reveal risk, performance, and opportunity.

Why Banks Struggle: It’s Not Volume—It’s

Fragmentation

Banks don’t have a data scarcity problem; they have a data sprawl problem. Critical information lives in siloed LOS platforms, core systems, spreadsheets, credit files, and even emails. That fragmentation creates three barriers:

1. Inconsistent structures: Different formats, quality levels, and definitions.
2. Manual processes: Every rekey or manual manipulation adds time and error.
3. No single source of truth: Competing numbers erode trust and slow decisions.

The costs are real. More than one quarter of data and analytics professionals say their organizations lose over $5M annually due to poor data quality, a risk that compounds as AI scales, according to a 2024 Forrester study.

Regulators are also sharpening their focus. The FDIC’s Risk Review flags operational vulnerabilities (including data and cyber risks) alongside credit concerns such as CRE—another reminder that weak data governance can become a safety and soundness issue.

That’s why the concept of the Data Pond ™ from Baker Hill is so critical. It moves data from being a “collecting” exercise to a “curating” movement and using that curated data to make an impact.

Reconciliation: The Most Underrated Pain Point in

Lending

Multiple versions of a borrower’s financials or risk ratings can undermine every downstream decision. The solution is to embed reconciliation into the workflow—not bolt it on later. That looks like:

• Automated spreading and integrated ingestion to reduce discrepancies.
• Standardized definitions and shared taxonomies across credit, finance, and risk.
• Clear ownership of the golden record. It does not need to be your core system, which is a path many FIs go down, but it needs to be a trusted single source of truth on that data.

This aligns with supervisory expectations: Risk teams should apply established model risk management principles and robust data governance to AI-enabled processes.

Integration Is a Discipline (APIs Make It Scalable)

Integration isn’t just a technology project; it’s a discipline of data harmonization:

• Map fields across systems.
• Normalize formats and values.
• Track lineage so every metric is explainable.
• Build pipelines/APIs so data flows instead of stagnating.

Banks increasingly recognize this. In a McKinsey global survey, 88% of respondents said APIs have become more important over the past two years, and large banks now allocate ~14% of their IT budget to APIs.

From Data to Decisions: Make It Actionable

No bank or credit union is going to win a trophy that will matter for having the most data. But they will win in shareholder value for having the most actionable data. Actionable data has three traits:

1. Timely: Insights that arrive weeks late are history, not guidance.
2. Contextualized: Numbers plus trends, benchmarks, and “what-ifs.”
3. Trigger-ready: Alerts, recommendations, and automated next steps.

Getting there requires the right architecture. McKinsey estimates that banks spend 6–12% of their tech budget on data, yet the right architectural choices can halve implementation time and cut costs by ~20%, accelerating impact.

At Baker Hill, we design workflows where actionability is built in: streaming ingestion, standardized definitions, and decisioning logic that routes the next best step—whether that’s an automated covenant tickler, a risk rating review, or a proactive relationship outreach.

What High-Performing Institutions Do Differently

The best clients we see don’t treat data as an IT project; they treat it as a strategic asset. How can you do the same?

• Institutionalize governance: Shared definitions, data stewardship, and clear ownership.
• Automate relentlessly: Eliminate rekeys and manual reconciliations so analysts can analyze.
• Use shared analytics: Adopt dashboards and metrics across credit, operations, and the board.
• Evolve continuously: Think “digital evolution,” not one-time transformation.

Getting Data-Ready for the AI Age

The age ahead of AI-driven lending, real-time analytics, and continuous risk monitoring will reward institutions that prepare now. It’s not just theory. Fully embracing AI can improve a bank’s efficiency ratio by up to 15 percentage points, but only when the data foundation and workflows are ready.

Three imperatives:

1. Clean, structured data: AI is only as good as the data you feed it. Responsible use also demands explainability and lineage.
2. Modernized workflows: Reduce cycle times and human error by streamlining intake, spreading, reconciliations, and exception management end-to-end.
3. A culture of digitization: Technology doesn’t transform banks—people do. Align incentives, training, and adoption across the institution.

Momentum is building. A recent industry assessment found up to 91% of financial services firms are adopting or using AI across operations, from fraud detection to underwriting—raising the bar for data quality and governance.

Where We’re Focused at Baker Hill

Our mission is to help institutions move from collecting data to curating it within the Data Pond™ so teams can make faster, more confident decisions. In practice, that means:

• Unified data models that harmonize borrower, behavioral, and portfolio data.
• Embedded reconciliation in credit workflows to protect the golden record.
• Transparent lineage and definitions to restore trust in metrics.
• API-first integration so data flows to the right decisions at the right time.

Banks that get this right don’t just close loans faster; they identify risk earlier, surface opportunity sooner, and allocate capital with greater precision.

A Final Thought—and a Question

In lending, insight isn’t about having more data; it’s about having the right data, reconciled and in motion. The institutions that unify their borrower, behavioral, and portfolio signals and then wire them into disciplined workflows will set the pace in an AI-powered market.

What’s the single biggest data friction in your credit process today and what would it take to remove it for good? I’d love to compare notes and share what we’re seeing work across our client base.