Why good lending decisions still fail in execution
A lending policy can be approved in a morning yet take months to reach every product, channel, and customer journey. This decision-to-execution gap can turn sound strategy into inconsistent outcomes, operational risk, and missed opportunities.
Summary
- Lending decisions create value only when implemented quickly and consistently.
- Disconnected systems and technology dependencies turn routine changes into complex projects.
- Adaptable lending platforms close the execution gap while strengthening governance and control.
Lenders operate amid constant regulatory change. Thomson Reuters reports that major banks receive nearly 200 regulatory updates a day.¹ The European Banking Authority has also acknowledged that the EU regulatory and supervisory framework has grown in both size and complexity, adding to the burden on financial institutions.²
Some regulatory updates require material changes to policies, products, processes, controls, or systems. But regulation is only one source of change: lenders must also respond to portfolio performance, funding costs, and commercial opportunities.
In my discussions with lenders, the quality of decision-making is rarely the constraint. Risk, compliance, product, and executive teams generally understand what needs to change. The difficulty lies in implementing that decision consistently across the entire lending operation.
I call this the decision-to-execution gap: the distance between approving a lending decision and seeing it applied consistently across the operation.
Closing that gap means making controlled change part of the lender’s normal operating model, with the appropriate oversight and safeguards built in.
Manage the decision across the lending lifecycle
For example, consider a change to affordability criteria. It may affect the information collected from an applicant; the affordability calculations and the rules used to approve, refer, or decline the application. It may also change approval authorities, exception handling, customer communications, and monitoring.
Inconsistencies emerge when the change is not applied across the lending lifecycle. Origination may adopt the new policy while loan management continues to use the previous one. One product may be updated while another waits for a later release. The unintended result is two operational versions of the same policy.
“A policy is not implemented because the document has changed. It is implemented when the application journey, decision engine, exception path and servicing process all apply it consistently.”
Lenders need to be able to follow a decision across the lifecycle, from application and credit assessment through approval, administration, and subsequent changes to the agreement.
These capabilities do not have to reside in a single system. What matters is that the relevant products, rules, workflows, data, and controls remain connected.
Give the right teams a controlled way to make changes
Product parameters, pricing rules, decisioning criteria and workflows are often embedded in code or divided between several applications. A modest adjustment can lead to multiple development requests, integration changes, tests, and release dates.
Suppose a lender decides that applications require additional review when existing exposure to a particular region or asset class exceeds an agreed threshold. The lender should be able to configure the concentration rule, use current portfolio data to identify affected applications, route them to the appropriate authority, and record and monitor the outcome. A transformation program is disproportionate to such a routine policy adjustment.
“If changing a single lending rule involves several systems, multiple teams, and a wait for the next major release, the constraint is no longer the credit decision. It is the design of the lending operation.”
Frequently changing elements, including eligibility criteria, pricing, decisioning rules, approval authorities, data requirements and workflows, should be managed as controlled configurations.
Lending, risk, compliance and operations teams can then manage the configurations they own within agreed permissions, tests and approval processes. Technology teams remain responsible for the wider environment: architecture, integrations, security, and resilience.
Keep the foundation stable and the operation adaptable
Some parts of a lending platform should change infrequently. Transaction processing, security, data integrity, and regulatory records all require stability. Products, policies, and workflows need to adapt more regularly.
A modular or composable architecture helps separate these different needs. Lenders can update one part of the lifecycle without rebuilding the entire platform, while open interfaces connect it with core systems, data providers, KYC services, customer channels, and audit tools.
Few lenders can replace every system at once, and wholesale replacement is not always the appropriate response. Addressing a specific constraint in origination or loan management may deliver more value while allowing stable capabilities to remain in place.
The practical goal is to make frequently changing parts of the operation easier to manage without destabilizing the rest.
Build governance into execution
Governance belongs inside the change process. Clear ownership, role-based permissions, separation of duties, testing, approvals, version control and effective dates should be part of how each change is made.
This also creates a reliable record of what changed, when it changed, and who approved it, rather than requiring teams to reconstruct that evidence later.
Once a change is live, lenders need to understand whether it is working as intended. Approval rates, referrals, overrides, exceptions, and customer outcomes can show whether the operational result matches the original decision.
Automate the standard path without losing judgement
Automation works best for cases that fall within clear policy. Cases involving greater complexity or uncertainty still benefit from expertise and judgement.
ASN Bank’s business-financing journey provides a practical example. Working with Akkuro Lending, the bank introduced a fully digitized process for SMEs - from application to final approval in a matter of months.³ Applications within standard policy frameworks can be processed automatically, while complex cases remain available for personal assessment.
This balance is important. Automation makes the standard path consistent. Exceptions remain visible and can be routed to an appropriately authorized reviewer with the relevant information.
“The objective is not to automate every case. It is to make the standard path genuinely standard—and give exceptions a clear, controlled route to human judgement.”
Joris Nijboer
Measure the decision-to-execution gap
Lending leaders need a clear view of how well their organisation implements change. Three questions are particularly useful:
-
How long does it take to move from approving a policy, pricing or product change to applying it in production?
-
Can the change be applied consistently across every relevant product, channel, and stage of the lending lifecycle?
-
Can the organization demonstrate what changed, when it changed, who approved it and why?
Material changes to lending policies, products, and decision rules still require careful assessment, testing, and approval. The opportunity is to eliminate avoidable delay and fragmentation when implementing approved changes.
Lenders cannot control the volume or timing of regulatory, market, and customer change. They can determine how ready their operation is to respond.
I see adaptability as both an operational capability and a risk-management capability. A good lending decision creates value only when the organization can put it into practice—with speed, consistency and control.
Sources
- Thomson Reuters, “Five steps banks can take for worry-free compliance.”
- European Banking Authority, “Simplification and efficiency of the regulatory and supervisory framework,” 2025.
- Akkuro, “ASN Bank digitizes the entire business financing journey.”