Why the Future of Banking Productivity Is About More Than Headcount
How human work, decision throughput, organisational friction, specialist capability and external labour are changing the economics of the future bank.
The future workforce question is not how many people a bank employs. It is how much economically valuable output its organisational architecture can produce for every unit of workforce economic cost.
For decades, the workforce has appeared in bank strategy principally as a cost, a capability base and an organisational resource. Management monitors headcount, compensation, productivity, span of control, employee engagement and revenue per employee, while restructuring programmes frequently translate strategic change into reductions in roles, branches and administrative capacity. When profitability comes under pressure, personnel expense naturally attracts attention because it represents one of the largest controllable components of the operating cost base.
That logic remains valid, but it is becoming insufficient.
The economic architecture of bank work is changing. Routine processing is becoming easier to automate. Physical distribution continues to decline in many markets. Increasingly sophisticated technology allows more customer activity to be served without corresponding growth in administrative capacity. At the same time, banks require greater capability in cyber security, data, fraud prevention, operational resilience, regulatory interpretation, complex risk management, specialist advisory and technology governance. Work can also move outside the formal employee base into outsourced operations, cloud providers, specialist partners, managed services and other third parties without disappearing from the economics of the institution.
The result is that the workforce question can no longer be reduced to how many people a bank employs.
A smaller bank workforce is not automatically a more productive workforce. An institution can remove employees while preserving the same underlying complexity, transfer activity to contractors or service providers, automate individual tasks while retaining cumbersome processes, increase managerial spans while creating slower decision pathways, or reduce staff expense while weakening the capabilities required to generate revenue and control risk.
The opposite can also occur. A bank can increase employment in selected areas while improving its economic position because those people generate more revenue, enable faster decisions, protect capital, reduce losses or allow the institution to operate at greater scale without equivalent increases in cost elsewhere.
Recent European banking data illustrate why the distinction matters. The number of employees at EU credit institutions declined by only 0.8 percent during 2025, while the number of bank offices fell by 2.62 percent. At the end of the year, euro area credit institutions still employed approximately 1.74 million people. The workforce is therefore changing far more gradually than the physical distribution footprint, reinforcing the point that digitalisation does not translate mechanically into equivalent reductions in employment.
The economic question is becoming more precise. Banks need to determine how much economically valuable output their organisational architecture can produce for every unit of workforce economic cost, which forms of work deserve increasingly scarce human attention, and whether the composition of the workforce is changing quickly enough to support the institution the bank intends to become.
Bancly describes this as Workforce Economic Productivity: the amount of revenue, customer value, decision capacity, risk control and institutional adaptability produced for each unit of workforce economic cost.
This is not an HR metric. It is an enterprise economics question.
Headcount Is Not Workforce Productivity
Headcount is one of the easiest organisational variables to observe, which helps explain why it features prominently in restructuring programmes.
An institution employed 20,000 people and now employs 18,000. The arithmetic is visible, the cost implications can be estimated and management can communicate a clear productivity narrative. Yet the economic meaning of those 2,000 positions depends entirely on what happened to the work.
If unnecessary activities disappeared, the reduction may represent genuine productivity. If processes were redesigned so that fewer people can produce the same or better outcomes, the improvement may be structural. If work moved to technology or an external provider at a lower total economic cost, value may still have been created.
If the same tasks were redistributed among the remaining workforce, customer service weakened, control capacity deteriorated or external expenditure replaced internal salary cost, the productivity improvement can be substantially smaller than the headline reduction implies.
This distinction becomes especially important because employment changes in banking do not move uniformly with digitalisation. ECB structural indicators show that EU bank employment actually increased modestly during 2023 and 2024 before declining by 0.8 percent in 2025, even while branch networks continued contracting. The long-term decline in employment therefore appears less linear than the continuing reduction in physical offices.
The conclusion should not be that digitalisation has failed to reduce labour requirements. It is that the relationship between technology and workforce economics is more complicated than simple substitution.
Some work disappears. Some becomes more productive. Some moves elsewhere. Some becomes more specialised. Some is created because an increasingly digital bank requires capabilities that the traditional institution did not need at comparable scale.
This means that management needs to distinguish headcount productivity from economic productivity.
Headcount productivity asks how much output is produced by each employee. Economic productivity asks how much institutional value is produced by the complete architecture of human work, including employees, contractors, external specialists, outsourced providers and the managerial structure required to coordinate them.
The second measure is harder. It is also much closer to the economics that matter.
The Bank Workforce Is Being Recomposed
The future workforce is unlikely to be defined simply by fewer people. It is more likely to be defined by a different composition of work.
Traditional banking required substantial human capacity in activities such as transaction processing, branch servicing, document handling, reconciliations, data entry, administrative support and routine control. Many of these activities remain necessary, but the human effort required to perform them can decline as infrastructure, workflow systems and automation improve.
Other activities are moving in the opposite direction. Modern banks require sophisticated capabilities in cyber security, technology resilience, fraud prevention, data governance, compliance, model risk, complex credit, digital product management and third-party oversight. Customer economics can simultaneously increase the value of sophisticated relationship managers, private bankers, transaction specialists, corporate advisors and other roles where judgement, trust and commercial interpretation remain important.
The operating model is therefore moving from a workforce dominated partly by process execution towards one with a greater concentration of judgement, orchestration, specialist expertise, customer influence and control.
That transition is economically significant because different forms of work have different cost structures. Routine processing can often be standardised and scaled. Expert work is frequently more expensive per employee. A bank can therefore reduce headcount while average compensation rises because the remaining workforce contains a greater proportion of scarce specialists. Personnel costs do not necessarily fall in direct proportion to employee numbers.
This helps explain why bank cost plans need to be interpreted carefully. The European Banking Authority reports that staff expenses represented approximately 9.9 percent of equity for EU and EEA banks in 2025. Banks expect personnel expenses to decline over the next several years, partly because of automation and efficiency measures, but the EBA has cautioned that broader cost assumptions may be optimistic because technology expenditure remains substantial and continuing investment is required.
The structural implication is larger than the forecast itself. The bank may employ fewer people while spending substantially more per strategically important role and more on the technological and external architecture that enables those people to operate effectively.
The objective should therefore not be minimum headcount. It should be the right economic composition of work.
Workforce Cost Can Move Without Disappearing
One of the risks in conventional workforce analysis is that employment cost is treated as if it were synonymous with the economic cost of work. It is not.
Banks increasingly use external technology providers, outsourced operations, specialist service companies, contractors, consulting capacity, cloud infrastructure and third-party platforms. Some of these arrangements can produce genuinely superior economics because external providers possess greater scale, expertise or technology than the bank could economically reproduce internally.
The work, however, still has a cost. It has simply moved.
A bank can therefore reduce formal employee numbers and report lower personnel expense while increasing third-party administrative expenditure. The reported workforce becomes smaller while the economic workforce supporting the institution may remain substantially larger.
This distinction becomes more important as outsourcing and technology dependence increase.
European supervisory authorities have identified rising reliance on third-party providers, particularly across cloud, payment and technology services. In the EBA’s 2026 risk assessment questionnaire, approximately 80 percent of responding banks identified ICT service-provider dependencies outside the EU and EEA as a significant challenge, while around 60 percent identified payment-solution dependencies. Operational resilience requirements increasingly require banks to understand not only direct internal operations but also the third parties on which critical services depend.
The Basel Committee’s operational resilience principles reinforce the same point. Banks are expected to map interconnections and interdependencies supporting critical operations, including third-party dependencies, because responsibility for the operational outcome remains with the regulated institution even when work has been externally provided.
This has a workforce implication that receives too little strategic attention. Outsourcing can remove jobs from the payroll without removing the economic requirement to govern the work.
Third-party relationships require procurement, oversight, risk assessment, performance management, continuity planning, data governance and exit capability. The organisation may therefore exchange direct execution cost for external service cost plus internal orchestration cost.
That trade can still be economically attractive. It simply needs to be measured correctly.
Bancly would therefore define Workforce Economic Cost more broadly than salary and benefits. It should incorporate employee cost, contractors, outsourced labour, specialist services and the organisational cost required to coordinate external work where those resources substitute materially for internal activity.
Without this broader view, headcount reduction can become accounting migration rather than economic productivity.
Organisational Friction Has an Economic Cost
A considerable proportion of workforce expense does not arise because banks employ too many people. It arises because people spend too much time navigating the institution.
A credit decision passes through multiple committees. A customer request moves across several departments. One team prepares information that another team reformats. A decision requires repeated escalation because authority is unclear. Control functions request similar evidence independently. Management meetings exist partly to coordinate organisational structures that were created for a previous operating model. Employees spend time finding information, reconciling systems, managing exceptions and translating between organisational boundaries.
None of these activities appears on the income statement as organisational friction. They appear as salary.
This creates a crucial distinction between labour cost and friction cost. Labour cost is the economic price of employing capability. Friction cost is the portion of that capability consumed because the institution’s architecture makes economically useful work harder to perform.
Reducing friction can therefore increase productivity without reducing headcount. The same workforce can produce more decisions, serve more customers, manage more assets or control more risk if the organisation removes unnecessary hand-offs, duplicated governance and process complexity.
This matters particularly in regulated institutions because some controls are genuinely necessary. The objective is not to remove governance indiscriminately. It is to distinguish control from coordination overhead.
A credit committee that materially improves risk selection can create economic value. A committee that exists because decision rights are ambiguous can consume it. A second-line review that prevents losses can be highly productive. Repeated manual evidence gathering caused by poor information architecture can represent structural waste.
The bank should therefore be able to ask not simply how many people perform a process, but how much of their effort contributes directly to the economic outcome the process is intended to produce. This is where organisational design enters Future Banking Economics.
Decision Throughput Is Becoming a Productivity Measure
Industrial productivity has traditionally been understood through units produced per worker or per hour. Banking produces something different.
Much of the economic output of a modern bank consists of decisions. Should this customer receive credit? At what price? How much capital should be allocated? Should a suspicious transaction be escalated? Which customer should receive which offer? Should a limit be increased? Should an exposure be restructured? Should a risk be retained, distributed or hedged? How should excess liquidity be deployed?
These decisions determine revenue, risk, capital, customer outcomes and losses.
The economic productivity of a bank therefore depends partly on its ability to make more high-quality decisions with less organisational effort and shorter elapsed time.
Bancly describes this as Decision Throughput.
Decision Throughput is not speed alone. A bank that approves everything quickly has not become more productive if losses subsequently rise. Nor does a bank improve productivity simply by adding more controls if high-quality decisions take so long that economically valuable opportunities disappear.
The relevant objective is the number and value of decisions the institution can process at acceptable quality for a given amount of economic resource.
This can be measured across many businesses. In credit, higher Decision Throughput can reduce turnaround time, increase conversion and allow relationship managers to spend less time managing process. In fraud, faster high-quality decisions can reduce losses without unnecessarily blocking legitimate customer activity. In collections, better prioritisation can improve recovery economics. In treasury, faster information and decision cycles can improve liquidity and balance-sheet management. In customer servicing, removing routine escalations can reduce cost while preserving human attention for complex cases.
This concept connects workforce architecture directly to the economics of the bank. The value of better work should eventually appear through numbers.
Expertise Is Becoming More Economically Concentrated
As routine work becomes easier to automate or standardise, the relative value of scarce expertise can increase. This creates a different workforce problem from traditional manpower planning.
The bank may need fewer people in a function while becoming more dependent on the quality of a small number of highly specialised individuals.
Cyber security provides an obvious example. Data, regulatory interpretation, model validation, treasury, complex corporate credit and specialist wealth advice can exhibit similar characteristics.
This creates concentration risk inside the workforce. A bank can be adequately staffed numerically and still be fragile if critical capability rests with a handful of individuals whose expertise is difficult to replace.
The same issue arises when important knowledge resides outside the institution. Outsourcing can improve economics while increasing dependence on external expertise that the bank no longer possesses internally.
Future workforce strategy therefore needs to distinguish labour capacity from institutional capability.
One thousand employees do not equal one thousand interchangeable units of work. The economic value of individuals differs according to their knowledge, decision authority, customer relationships and ability to influence risk or revenue.
This complicates crude headcount reduction. Removing one routine administrative role and losing one highly productive relationship manager both reduce headcount by one. Their economic consequences can be radically different.
The bank therefore needs a stronger concept of workforce capital allocation. Human capacity, like financial capital, should be allocated towards the activities where the institution possesses a Right to Win and where judgement or expertise materially changes economic outcomes.
The objective is not simply to employ fewer people. It is to ensure that the most expensive human capacity sits where human judgement generates the greatest marginal institutional value.
The Managerial Layer Is Part of the Workforce Economics
Senior management structures are rarely treated as a productivity variable in the same way as operational work. They should be.
Managerial layers have economic value when they improve accountability, allocate resources, resolve trade-offs, develop capability and ensure control. They create economic cost when they principally relay information between other layers, replicate decision authority or increase the number of organisational interfaces required before action can occur.
This makes managerial architecture part of Workforce Economic Productivity.
A highly layered institution can create substantial hidden expense because the cost is not limited to executive compensation. Each layer creates reporting requirements, meetings, information preparation, escalation processes and coordination routines throughout the organisation.
Flattening the hierarchy does not automatically solve the problem. Excessively wide spans can weaken supervision and overload decision-makers. The relevant question is whether the architecture allows authority to sit close enough to information while retaining appropriate control.
This is why future workforce design cannot be reduced to a target span of control. The issue is decision architecture. Who has authority? What information is required? Which decisions genuinely need escalation? Which controls can be embedded rather than manually reviewed? Where is judgement economically necessary? Where is management activity simply compensating for fragmented organisational design?
These are operating-model questions, but their consequences are financial. Every unnecessary layer consumes cost. Every avoidable escalation delays economic activity. Every unclear accountability increases coordination requirements. The managerial architecture therefore belongs inside the cost curve discussed in Article 04.
Productivity Must Include Risk and Control
One of the easiest mistakes in workforce transformation is to treat revenue-generating or customer-facing activity as economically productive while considering control capacity principally as overhead. Banking cannot operate on that basis.
Risk, compliance, cyber security, fraud prevention and operational resilience do not necessarily create revenue directly, but they protect the ability to generate revenue without unacceptable losses.
The economic question is therefore not whether control functions should be minimised. It is whether they produce sufficient risk-control productivity.
The EBA reported approximately 3.9 million operational risk loss events across EU and EEA banks during 2025, an increase of 25 percent over the previous year. Materialised operational losses nevertheless declined to EUR 15.2 billion, while technology-related risk, fraud and cyber security remained major areas of concern. Operational-risk capital requirements also rose to 13.4 percent of total capital requirements at the end of 2025, partly reflecting regulatory changes.
These numbers illustrate why reducing control headcount mechanically can produce false productivity. The economic purpose of control capacity is to prevent losses, reduce tail risk, protect regulatory standing and preserve customer trust.
The relevant productivity measure should therefore ask how much risk the control architecture can govern effectively for each unit of cost. A bank whose risk organisation becomes smaller while operational losses rise has not necessarily become more productive. A bank that invests more in fraud capability while materially reducing fraud losses may have improved Workforce Economic Productivity even though direct cost increased.
The same logic applies throughout the organisation. Productivity is an economic relationship between resources consumed and outcomes produced. The outcome must include risk.
Technology Does Not Remove the Need to Redesign Work
Technology will clearly influence workforce economics, but technology alone does not determine them.
An institution can place new systems on top of old processes and end up operating both. It can automate individual tasks while preserving the hand-offs between them. It can create digital customer journeys while retaining manual exception handling behind the interface. It can introduce sophisticated decision tools while requiring the same approval hierarchy to sign off the output.
The organisation then becomes technologically richer without becoming economically simpler.
This is why workforce productivity cannot be separated from the Structural Cost Position introduced in Article 04. Technology creates economic value when it changes the amount or quality of human work required to produce an outcome. If the technology cost rises while the labour, process and organisational costs surrounding it remain unchanged, the bank may simply have created an additional cost layer.
European supervisory evidence points directly to this tension. ICT-related expenditure represented 32.1 percent of other administrative expenses at EU and EEA banks in 2025, up from 31.2 percent a year earlier. Banks simultaneously expect staff expenses to fall over the coming years, making the extent to which technology investment eventually translates into lower total operating economics a central profitability assumption. The EBA estimates that if anticipated staff and administrative cost savings do not materialise, projected sector ROE for 2028 could be approximately 10.1 percent rather than 11.2 percent under a simplified sensitivity analysis.
This is why the economic test should not be whether technology has been deployed. It is whether the operating architecture now requires less low-value human effort and produces more valuable output.
A Workforce Economic Productivity Framework
A more complete workforce framework should therefore examine several dimensions simultaneously.
The first is Workforce Economic Cost. This includes direct employee expense, benefits, contractors, outsourced labour and material external service arrangements that substitute for internal work.
The second is Economic Output. The organisation should understand how workforce capacity contributes to revenue, customer relationships, risk management, capital productivity and service outcomes rather than measuring productivity only through transactions or employee counts.
The third is Decision Capacity. Management needs to know how many economically important decisions can be processed at acceptable quality, how long they take and how much organisational effort each decision requires.
The fourth is Skill Intensity. The bank should understand where scarce specialist capability sits, how dependent critical activities are on individuals or external providers, and whether workforce investment is moving towards the capabilities the future institution actually requires.
The fifth is Scalability. A productive workforce architecture should allow customer volume, transaction activity or balance-sheet scale to increase without requiring proportional increases in labour.
The sixth is Adaptability. The institution needs sufficient flexibility to move people, skills and decision capacity as profit pools, regulation, customer behaviour and the bank’s strategic priorities change.
Together, these dimensions constitute Workforce Economic Productivity.
The CEO Workforce Economics Scorecard
The implications for CEO and board oversight are substantial because many conventional workforce metrics remain necessary while answering only part of the economic question.
Leadership should continue monitoring total headcount, but it should also know what categories of work are expanding and shrinking and whether that composition reflects the strategic direction of the bank.
Staff expense should remain tightly governed, while management distinguishes direct personnel cost from the wider economic cost of external labour, outsourcing and organisational coordination.
Revenue per employee can remain useful, but management should understand whether improvement reflects genuine productivity, temporary revenue conditions or the migration of work outside the employee denominator.
Turnaround times should be measured not merely as customer-experience indicators but as evidence of Decision Throughput.
Control functions should be evaluated through risk outcomes and loss prevention rather than being treated uniformly as administrative overhead.
Senior-management structures should be assessed according to whether they accelerate or impede decisions, not simply whether span-of-control targets have been achieved.
The bank should also know what proportion of its workforce performs work that could reasonably cease to exist if processes, products and organisational structures were redesigned rather than merely automated.
This produces six CEO questions. What does the workforce actually cost economically? Which forms of work create the greatest institutional value? How much expensive human capacity is consumed by organisational friction? Where does human judgement materially improve economic outcomes? Can the bank grow without workforce cost growing proportionately? Can the organisation reallocate capability fast enough when its economic priorities change?
The Questions at the Top Must Change
The future workforce conversation therefore needs to move beyond personnel planning.
Boards should continue asking how headcount will change, while understanding what will happen to the work associated with those roles.
Management should continue seeking productivity, while distinguishing work elimination from cost migration.
Automation should continue to be pursued where the economics justify it, but the institution should determine whether technology is removing structural complexity or merely performing existing complexity faster.
Outsourcing should be used when external providers create better economics, while the bank accounts for the internal governance, resilience and dependency cost created by those arrangements.
Specialist talent should be expensive where scarce expertise genuinely changes revenue, risk or capital outcomes. The relevant discipline is ensuring that premium human cost sits in premium economic activity.
Managerial layers should be evaluated according to the decisions they improve rather than the hierarchy they preserve.
Control functions should demonstrate risk productivity rather than being subjected automatically to the same reduction logic as routine administration.
Most importantly, workforce planning should become part of the economic strategy of the institution.
If the bank intends to move towards different profit pools, the workforce should move towards the skills required to capture them. If the institution intends to become more capital productive, its people and decision architecture should support more precise capital allocation. If customer primacy matters, economically valuable relationship capability must receive appropriate investment. If Structural Cost Position is expected to improve, low-value work and organisational friction must genuinely disappear rather than migrate between budgets.
The workforce plan should therefore prove the strategic thesis just as the capital plan should.
The Bank Workforce Is Becoming an Economic Variable
Banks will almost certainly employ people in 2035. The important question is what those people will be doing.
A future bank should require less human capacity to move information between systems, execute routine transactions, reconcile simple records and perform repetitive administrative tasks. It should require substantial human capability wherever judgement, trust, creativity, complex risk interpretation, relationship influence, institutional accountability and strategic decision-making remain economically important.
The exact balance will differ by institution and market. What matters is that the economics of work will continue changing.
The bank that treats workforce transformation principally as a headcount programme risks optimising the wrong variable.
Reducing employees can improve cost. Reducing unnecessary work can improve the institution. Those are not the same objective.
Workforce Economic Productivity requires management to understand the complete architecture through which human capacity is converted into customer outcomes, revenue, risk control, decision capacity and financial performance.
It requires the bank to distinguish the cost of capability from the cost of friction. It requires leadership to understand that outsourced work remains economic work. It requires specialist capability to be treated as scarce institutional capital rather than interchangeable labour. It requires organisational design to be evaluated according to the economic throughput it enables. It requires control capacity to be measured by the risks it governs, not simply by what it costs.
Ultimately, it requires the workforce conversation to reconnect with the numbers that govern the bank.
A more productive workforce should lower unit cost. It should improve operating leverage. It should increase Decision Throughput. It should strengthen customer economics. It should reduce avoidable operational losses. It should allow the institution to scale without proportional expansion in labour. It should improve the relationship between organisational cost and the revenue, risk and capital outcomes produced.
Those effects eventually reach CIR, customer profitability, RAROC, ROE and enterprise value.
That is why the workforce is becoming an economic variable rather than simply an organisational one.
The central question for leadership is therefore no longer merely how many people the future bank should employ. It is what work will still justify human economic cost in the bank of the next decade, what work should cease to exist, and whether the organisation is reallocating its most expensive resource towards the activities that create the greatest institutional value.
The answer will determine far more than the size of the workforce. It will determine the productive capacity of the bank itself.
