Banking OpEx Reduction: Where Cost Actually Leaks Before You Fund a Transformation

Key takeaways

‍

• Bank operating cost concentrates in three drivers: IT, personnel expenses and physical infrastructure. The share in each varies by institution, so the starting point is your own ledger.

• A target set before the diagnosis tells you how much to cut but not where, which is why so many cost programs cut the wrong line.

• Cost-to-income is the scoreboard, and it barely moves when a program cuts a line that wasn’t the problem.

• A diagnostic produces a measured picture of where cost sits today, ranked by how reversible each decision is, so the target is sized from evidence instead of a benchmark.

‍

What OpEx reduction means in banking

‍

OpEx reduction is lowering the recurring cost of running the business without harming output. In banking it concentrates on the three biggest drivers: IT, personnel expenses and physical infrastructure. The aim is to find where cost leaks before committing to a transformation program.

‍

That’s a different exercise from one-off cost cutting. A budget freeze or a percentage reduction across every department finds savings fast, but it treats a well-run team the same as an inefficient one, because it can’t tell the two apart.

‍

The savings appear in next quarter’s numbers, but often return the quarter after, once frozen spend is released or work that was dropped has to be picked up again. A blanket cut also gives no way to tell which savings will last and which will come back as cost.

‍

Run as a standing discipline, OpEx reduction keeps finding where cost concentrates as the business changes. You get a repeatable answer to where the money goes that holds up the next time someone asks you to defend it, instead of a fresh guess each budget cycle.

‍

"OpEx optimization" and "cost optimization" in banking all describe the same goal: more output from the same recurring spend. Finance teams use all these terms interchangeably, since cost-to-income is what they all focus on.

‍

Why cost-to-income is the number that matters

‍

Cost-to-income is the ratio most banks use to measure operational efficiency in banking. It shows how much of every unit of income is consumed by the cost of earning it.

‍

Cost-to-income ratio = Operating costs ÷ Operating income

‍

In this ratio, operating costs are typically administrative expenses and depreciation.


A lower ratio means greater efficiency, because less cost is consumed for every dollar of income.


Many banks target a ratio below 50%, although the right level depends on business mix and scale, so treat it as a direction more than a universal pass mark.

‍

For context, the European Central Bank’s supervisory banking statistics put the aggregate cost-to-income ratio of supervised banks at 53.06% in the second quarter of 2026. That’s the lowest since the series began in 2015. The ECB attributed the improvement to operating income growing faster than costs, which shows that the ratio responds to both sides of the division.

‍

The ratio moves slowly when a program cuts the wrong line. Trimming a visible, easy-to-defend cost such as a vendor renewal or a discretionary budget looks decisive in a board update, but changes little if that line wasn’t where cost was concentrated. The real drag on the ratio, like two departments running the same reconciliation, is often less visible and keeps running as before.

‍

Simply put, if you can't see the activity, the ratio won't move. This visibility gap usually happens if you diagnosis after setting a target. Running a bank on measured data is how you find which line needs to move before committing to cut it.

‍

The three drivers, and how much sits in each

‍

Cost concentrates in three places at most banks. The share each holds varies by institution, and no reliable cross-bank figure exists for all three. Read your own general ledger before comparing against anyone else’s.

‍

• IT, split between run and change. Run keeps existing systems operating: licenses, infrastructure and support. Change funds new projects and transformation. In its October 2024 analysis of bank technology spending, McKinsey found that run-the-bank and mandatory-change spending together often reach 70% of the technology budget at large banks that disclose the split. That leaves limited room for the new initiatives leadership wants funded, which is the tension a transformation budget meets before it’s approved.

• Personnel expenses. Usually the largest single driver. The opportunity here is duplicated and manual work, which is a different question from how many people the bank employs. Picture a commercial lending team and a risk team each pulling the same exposure figures from different sources every month, unaware the other runs the identical reconciliation. Or picture a handoff where someone re-keys data a system should have passed along. Removing that duplication changes what people spend their time on, and frees capacity for work the bank needs done.

• Physical infrastructure. Branch networks, office space and data centers still carry significant recurring cost as more work moves to remote and digital channels. This driver moves slowest of the three, because leases and facility commitments don’t unwind on a budget cycle. Exiting one early can cost more than letting it run to its next break point.

‍

The personnel driver is the one most often estimated from an org chart, which shows headcount and nothing about how time is spent. Workforce measurement in financial services is how it gets measured with evidence, across teams and locations instead of reporting lines.

‍

Why most cost programs miss

‍

Most cost programs miss for three reasons, and all three are avoidable once you know what to look for.

‍

1. Targets are set top-down, from an industry benchmark or a board-mandated percentage, instead of from the bank’s own cost base. A target built that way tells finance how much to find and says nothing about which line has that much room in it. The search starts from a number and works backward.

‍

2. Savings are then identified from org charts and general ledger lines alone. An org chart shows reporting structure and a ledger line shows what a cost category totals. Neither shows whether the work behind the line is efficient, duplicated or already half automated, so proposed savings tend to come from wherever the spreadsheet made a number easy to isolate.

‍

3. Duplicated process work is invisible in both views, which makes it the most expensive miss. Two departments running the same monthly reconciliation don’t appear as a flagged line item anywhere. They appear as two normal-looking cost centers, each defensible on its own, and comparing their workflows side by side isn’t anyone’s job.

‍

The fix for all three is the same starting point: how work runs today. Find cost that’s real and recurring, and absent from every report anyone reads.

‍

How? Measure time, tool usage and duplication across teams before a target is set, then let the target follow from what the measurement finds.

‍

Cost reduction in banking hinges on finding this pattern, among the other visibility gaps that cost banks money.

‍

What a diagnostic looks like

‍

An OpEx diagnostic is a fixed, three-week scan that measures where time, cost and tool usage concentrate, then turns those findings into a ranked list of decisions. It runs in five steps.

‍

1. Map where time and cost concentrate across IT, personnel and property.
2. Compare application usage with what’s licensed.
3. Identify duplicated or manual work between teams and departments.
4. Size each finding against its share of the cost base.
5. Rank findings by how reversible the underlying decision is.

‍

Mapping starts from what happens today, and sets a baseline you can defend if a board member challenges a number later. For property, that includes how teams use the office, which can be shown by work location insights.

‍

Usage comparison compares licensed seats to the applications people open. Surprisingly, it's often the first time anyone has seen both side by side.

‍

Duplication is where the personnel driver gets its evidence: which teams do overlapping work, and how much of the week it takes.

‍

Sizing separates findings that move the cost-to-income ratio from those small enough to leave for later. Ranking by reversibility then decides the order. A license downgrade or a tool consolidation is easy to undo, while a reorganization or a lease exit isn’t.

‍

What comes out is a short, ranked list of decisions the leadership team can take. A worked example at a large US bank shows the approach applied to one line, IT contractor spend, and how that saved them $2.5 million.

‍

In practice, the findings support three kinds of decision:

• The first is which licenses and tools to retire, downgrade or consolidate.

• The second is which duplicated processes to merge and who should own the merged version.

• The third is which property commitments to review at their next break point.

‍

Each decision arrives with the evidence behind it, which is what lets it survive a challenge.

‍

Sequencing the work

‍

Open with IT and vendor costs, before anything in personnel or property. This data is usually cleanest, because license counts, usage data, and contract terms already exist in some form. Most decisions in this driver are reversible too: downgrading a tier, consolidating two overlapping tools, or renegotiating a renewal doesn’t require unwinding anything structural if it ends up being the wrong call.

‍

Personnel and property come later, and they take longer. Changes there touch how teams work day to day, or commitments measured in years, so they need more evidence before you act and more lead time once you do. Rushing them first means impulsive additions that don't show up as wasteful for a quarter, or irreversible cuts that cause system decay without leaving any marker

‍

This order reflects which decisions you can make with confidence now, as opposed to which driver matters most. The later drivers need the evidence from the earlier findings before they’re safe to act on. Savings from IT can also help fund what comes next, since capturing them rarely needs new budget.

‍

A practical rule helps here. Act on a finding once it’s measured, sized and reversible, and queue it for later if any of the three is missing. Keeping a visible record of what was captured and when also shows the board progress early, while giving slower decisions the time they need.

‍

Where the data comes from

‍

Every step above depends on measuring where cost sits, and for most banks that measurement doesn’t exist anywhere finance or operations can see it.

‍

Insightful’s Workforce Analytics measures where time goes across teams and locations, and Workspace Security complies an auditable activity log of what applications and tools are actually in use. Together they show where cost is sitting before a transformation budget is committed.

‍

A diagnostic layer supplies the evidence an OpEx reduction runs on: where teams duplicate work, which licensed tools sit unused, and how time and office use differ across locations and departments.

‍

All of it is measured at team, department, and application level, and points to the process that needs tweaking. The decision about what to change, and how fast, stays with you and the team leads who own the work.

‍

Insightful is SOC 2 certified, HIPAA and GDPR compliant, and conforms to ISO 27001, which matters when the bank’s own security and risk teams review it. Because it measures time and application use on an ongoing basis, the same data serves the first diagnostic and acts as a control for later checks on whether the savings held.

‍

Find the leak before you set the target

‍

Run a diagnostic first. The number that comes out is one you can defend when the board asks why you cut or changed one line and not another.  If the findings point somewhere you didn’t expect, that’s a discovery you can error-correct for and feed back into the same cycle.

‍

Request a 14-day audit of where your own operating cost sits, and close the loop on ROI before the next budget cycle.

‍

Frequently asked questions

‍

What is OpEx reduction?

OPEX reduction is lowering the recurring cost of running the business without harming output. In banking, it concentrates on IT, personnel expenses and physical infrastructure, the three biggest drivers. The aim is to find where cost leaks before committing to a transformation program, instead of cutting broadly.

‍

What is the cost-to-income ratio?

The cost-to-income ratio divides a bank’s operating costs by its operating income. A lower ratio means greater efficiency. It’s the headline metric banks use to measure OPEX discipline, and many target a ratio below 50%, although the right level varies with business mix and scale.

‍

What are the biggest operating costs for a bank?

The biggest operating costs for a bank are IT, personnel expenses and physical infrastructure. IT splits between run, which keeps existing systems working, and change, which funds new projects. Personnel is usually the largest single driver, and physical infrastructure moves slowest.

‍

How long does an OPEX diagnostic take?

An OPEX diagnostic typically takes about three weeks. That’s enough to map where cost concentrates, compare application usage with licenses, identify duplicated work between teams and size each finding. The result is a ranked list of decisions a leadership team can act on.

‍

What is the difference between cost cutting and OPEX reduction?

Cost cutting is a one-off reduction, such as a budget freeze or a flat percentage across every department, that treats every team the same. OPEX reduction is an ongoing discipline that finds where cost concentrates and targets that specifically, so savings last beyond the next budget cycle

‍