Banks are being challenged to do more with less. what’s at stake?
The pressure on today’s banks to deliver business results is increasing. Interest rate headwinds are waning, price-to-book multiples remain compressed, and efficiency ratios remain consistently above 60%. At the same time, talent is scarce and competition from digital-native fintechs is intensifying. Global fintech revenues to grow by 21% in 2024, three times the growth rate of existing companies1In this environment, Banking leaders are being asked to do more with less – Increase fee income in a low-rate world, work harder on the balance sheet, and get more productivity from lean teams.
Since 2010, US regional banks have lost about 30 bps of productivity each year, eroding returns even as costs rose. A one percentage point gain in productivity could increase return on equity (“ROE”) by 100 to 150 bps – a significant change for boards and investors.
At Databricks, we work with banking leaders in revenue and cost centers around the world to reverse that trend. Of all banks investing in data and AI, only 5% of institutions have achieved measurable P&L impact. The reasons for failure are structural. Roughly 10% of initiatives fail due to a lack of integrated data and AI vision, 20% fail due to legacy technical debt and fragmented data assets, and 70% because democratization, governance, and enablement are not addressed.2
The top 5 percent of banks view productivity loss as an opportunity to better integrate data. They form the foundation of integrated, governed data for intelligence, automation, and ultimately measuring performance.
Databricks’ approach to unlocking step-change productivity
For banking leaders who are already modernizing their technology and integrating their data assets, Our gold-standard approach to improving productivity is two-pronged,
- take the fast track – Remove silos and old equipment. Examine every manual step in the sales process, starting with better data governance using the power of Databricks’ Unity Catalog. Unified governance enables you to democratize data to business users in natural language through AI/BI (Genie, dashboards) and Databricks One. IT teams are no longer stuck between bronze/silver data and business intelligence, helping to make faster and better decisions.
- change bank – By taking the fast track, banks earn the right to reimagine and transform their operating models, unlocking productivity gains. Using Databricks’ Agent Bricks, banks can build, evaluate, optimize AI agents at the cost-quality boundary to rapidly automate manually intensive processes, unlocking time for higher-value activities.
A Day in the Life of a Middle-Market Commercial Banking Leader
At Databricks, we work closely with commercial banking leaders in the US mid-market. Their job is to grow profitably, deploy capital efficiently and increase banker productivity. Yet the efficiency ratio remains above 60%, and relationship banker Productivity has been flat For more than a decade.
The number of employees is less. Since 2020, regional banks have cut front-line teams by almost 15%Even as credit demand, compliance requirements and customer expectations have increased. Each relationship manager now handles large portfolios with limited tools to prioritize customers, track activity, or focus on revenue-generating tasks such as new loan origination, deposit growth, and cross-selling treasury or payments products.
Pressure also comes from product teams. As banks launch new, often more complex offerings in lending, FX, or treasury management, RMs struggle to remain efficient. Training is slow, guidance is fragmented, and insights are trapped in esoteric CRM, spreadsheets, and product systems. The result is that opportunities are missed and “white-glove” customer service is weakened – on which much of the differentiated relationship banking depends on.
Leaders see these challenges as structural. Data is buried in loan systems, CRM instances and finance, with no unified layer to control or analyze it. Without a single, governed view of banker performance and client profitability, it is difficult to enforce accountability or direct efforts toward the highest-value opportunities.
Get on the fast track: manage and democratize your data
For the head of commercial banking, productivity is measured by pipeline volume and velocity, expense ratio, revenue growth and ultimately ROE. Databricks’ Unity List Enables this by giving leaders a single, governed data foundation – unifying information across loan systems, CRM, deposits and treasury to provide trusted insights at every level.
This foundation enables banking officers to access Databricks’ AI/BI Dashboard To track pipeline health and conversion rates in real-time, segment portfolios by banker, customer type or region, and identify cross-sell opportunities or margin leakage – all from a single source of truth. They can also link banker activity data (calls, meetings, client touches) to performance outcomes, helping to train managers for higher-value behaviors.
For frontline bankers, Databricks One These insights are made readily available through the AI/BI Genie, Databricks apps, and real-time AI/BI dashboards. Instead of relying on manual exports or month-end reports, bankers can simply ask, “Which customers require a deposit pricing exception?” or “Which renewals are at risk this quarter?” – and act on the answers immediately. This results in faster prioritization, faster renewals and higher wallet share.
Databricks helped leading bank Modernize your data and ML workflows, unifying client, product and channel data into a single governed platform. Within a year, the bank achieved a 90% increase in cross-sell effectiveness and reduced customer acquisition costs by more than 30% – demonstrating how integrated data and AI translate directly into productivity and profits.
Transform the Bank: Reimagine Workflow with Agentbricks
As leaders navigate the faster path, they earn the right to unlock the next wave of productivity through agentic workflows. In addition to making the same work faster, they can also change the way bankers create value.
With Databricks’ Agentbricks, banks can orchestrate AI agents that analyze internal and external data from FactSet, Moody’s, S&P, and Dun & Bradstreet to prioritize tasks and automate research, preparation, and outreach. For example:
- Automated Research: Prepare client documents by combining transaction history, risk signals and market news, cutting down on hours of manual preparation.
- Synthesize in action: Recommend products, services or outreach messaging with human-in-the-loop review for referrals and compliance.
- Automatic Execution: Close the loop from insight to action – send client emails, approve pricing exceptions, or automatically update systems of record.
Efforts to use AI to automate manual processes are well underway across the industry. A North American bank automated customer research on Databricks, reducing preparation time by 70% and expanding banker coverage without new headcount.
By simplifying technical complexity, Databricks’ Agent lets banks automate manual workflows such as customer research, credit memo creation or deposit pricing requests, giving bankers the ability to engage more customers faster and with more insights that will accelerate growth and reduce operating costs.
Databricks and our partners are here to help
Get in touch to see an “Agent Banker” demo built to maximize B2B seller productivity by recommending high-prospect engagements, automating customer research, and streamlining outreach. Using Agent Bricks, Databricks customers can rapidly build AI agents that remove administrative burdens so bankers can focus on what matters most: serving their customers. If you are a global banking or payments leader looking to reimagine your operating model and unlock productivity at scale, contact us to define and execute a roadmap that transforms data intelligence into measurable business performance. with consulting partners such as pocketHelping banks accelerate delivery and capture value faster from connected, data-driven solutions, you can grow efficiently at scale on the Databricks data and AI platform.
1 https://web-assets.bcg.com/04/4f/160d4f9d4d7e98ab417609b553d4/globalfintechreport-june2025.pdf
2 https://www.bcg.com/press/24october2024-ai-adoption-in-2024-74-of-companies-struggle-to-achieve-and-scale-value
