FINANCIAL SERVICES · NETWORK TRANSFORMATION
A European national bank has seen usage of its cash machines erode for three years in towns with fewer than 3,500 residents. The plan calls for removing 1,200 machines over three years — about 22% of the rural fleet — for a projected saving above €42 million a year at maturity.
The question is not whether customers still use these machines. It is what their disappearance means, place by place and situation by situation, and how that reading spreads.
THE PROBLEM
The operating data is unambiguous: rural cash-machine usage is declining 4.2% a quarter, electronic payments are rising, remote banking is becoming the norm. Each machine costs an average of €38,000 in annual maintenance.
This data measures usage. It does not measure what the point represents: an occasional withdrawal option, a fallback, autonomy, a service for local shops, a signal that services are being maintained in the area.
The question put to the executive committee was therefore not how many times the machine is used, but what its disappearance changes — and how that disappearance will be told. Two industry precedents had led one network to restore half the machines it had removed, and another to trigger a reading of condescending abandonment relayed up to the national level.
WHAT'S THE SAME FOR EVERYONE
WHAT DIFFERS FOR EACH
A VARIABLE: OPTION VALUE
The value of an infrastructure is not reducible to its usage rate. Part of that value lies in the fact that it remains available, including for those who no longer use it.
USAGE IS NOT VALUE
In the source case, 73% of rural segments had not used the machine for eighteen months, having switched to supermarket withdrawals or fully digital banking. This shift does not produce acceptance of the closure. It coexists with a structured rejection, one that the low-usage narrative reinforces rather than reduces.
WHAT A USAGE RATE MEASURES
Withdrawals made over a period. In the source case, 73% of rural segments had not used the machine for 18 months.
WHAT IT DOES NOT MEASURE
Rejection remains massive. It is not about the lost function, but about what the closure means.
WHAT CAN COEXIST
Paying entirely by card, not having withdrawn cash for a year and a half, and still opposing the removal of the point.
WHAT SHIFTS
Explaining low usage does not reduce rejection: this message validates the perception of indifference and worsens it.
A user/non-user split makes the decision impossible to inform. The case's 12 segments cross age, place of residence, banking habits, mobility, local political sensitivity and media exposure.
WHAT'S TESTED
The executive committee had 5 communication scenarios tested, combining timing, the scope of the announcement, the channel used, and the presence or absence of a compensation scheme.
2,400 synthetic customers, distributed proportionally across the 12 segments, were exposed to the 5 scenarios in randomised order to neutralise learning effects, with an average of 7 follow-up questions per person. Each scenario's media trajectory was then projected over 12 weeks, modelling likely relays: regional press, associations, local elected officials, shopkeeper unions, local social networks, national press.
SIMULATED POPULATION
The reconstructed population covers 1.2 million retail customers of the network, calibrated on public data and proprietary segments. No identifiable data enters the system.
It is structured into 12 locally coherent demographic segments, cross-referencing age, place of residence, local political sensitivity, banking habits, relationship to rurality, and media exposure.
ESTABLISHED DIGITAL USAGE
banking managed entirely remotely, withdrawals abandoned long ago, no reliance on the physical point
NON-USE WITH OPTION VALUE
no withdrawal for eighteen months, yet opposed to removal: the possibility mattered more than the use
REGULAR CASH USE
cash used for part of everyday spending, direct functional loss in case of closure
SHOPKEEPERS AND TRADESPEOPLE
cash flows tied to their activity, dependence on the local point for change and deposits
CONSTRAINED MOBILITY
no vehicle permanently available, schedules poorly compatible, dependence on someone else for transport
MULTI-BANKED CUSTOMERS
several institutions, several possible points: the closure is absorbed without reorganisation
URBAN SENIORS
no machine removed near them, yet strongly identifying with the fate of rural areas
LOW INVOLVEMENT
no settled position: the reaction will depend on the narrative available at the time of the announcement
These configurations make part of the population's heterogeneity visible. The simulation covers synthetic individuals, not a handful of persona types.
REACTIONS
REJECTION IS NOT ECONOMIC, IT IS SYMBOLIC
73% of rural segments had not used the machine for 18 months, having switched to supermarket withdrawals or fully digital banking. Yet rejection remains massive: the 47 motivations identified converge on the same structure. It is not the function that is at issue, it is the political signal of its disappearance. The rational message about low usage worsens the rejection, because it validates the perception of indifference.
THE REACTION SPILLS BEYOND THE GEOGRAPHIC SCOPE
Urban senior customers, living in cities where no machine would be removed, identify with the fate of rural populations: “today it's them, tomorrow it's us.” The reputational impact reaches 34% of the customer base, well beyond the areas targeted. Communicating only to the affected zones underestimates the real spread.
THE ALTERNATIVE'S VALUE LIES IN WHAT IT ACKNOWLEDGES
The most powerful lever identified is neither the gradual announcement nor the mobile service: it is the operational partnership with a major national postal operator, perceived as a public acknowledgment of the right to cash access. Announced last, the same scenario gets 45% acceptability; announced first, 67%. The scheme is identical, the reading is not.
THE CHANNEL IS PART OF THE CONTENT
A national announcement via the daily press gets 41% initial acceptability. The same decision announced in a cascade — prefects, then mayors, then regional press, then national press, over six weeks — gets 71%. Local elected officials relay the scheme as a negotiated regional agreement, not as a decision imposed from headquarters.
Two customers can receive the same announcement and draw two opposite conclusions. For one, “they're removing a machine nobody uses.” For the other, “they're leaving.” The bank talks about a network; part of the area hears a decision about its place.
ACCUMULATION
The same decision can be received on its own, or folded into a broader local narrative. The source case does not document the timeline of services closed in each town concerned. It does document the effect of the narrative, though: rejection is not about the lost function, but about what the closure says about the area.
BEFORE / AFTER
A conceptual diagram, with no geographic claim. Before, the withdrawal point is part of the set of services reachable from home. After, the shops and services remain, but cash access moves to a more distant point.
This same change does not create the same constraint for everyone: it depends on vehicle availability, schedules, routes already travelled, and the alternatives genuinely accessible locally.
SAME DISTANCE, DIFFERENT VEHICLE
The trip to the next point is an ordinary detour for one person, and something to organise for another.
SAME DISTANCE, DIFFERENT SCHEDULES
A point accessible during working hours is not accessible to someone working those same hours elsewhere.
SAME DISTANCE, DIFFERENT ROUTES
If the next point sits on a route already travelled, the constraint nearly disappears.
SAME DISTANCE, DIFFERENT ALTERNATIVES
The local density of other services determines what remains possible once the machine is removed.
WHAT A WITHDRAWAL VOLUME SAYS
The intensity of use over a period. It indicates what was done, not what remains possible.
WHAT AN OPTION VALUE SAYS
What the disappearance takes from those who no longer used the point: a fallback, autonomy, a presence.
SECOND-ORDER EFFECTS
USAGE
it can be near zero and the closure still stays contested
NARRATIVE
an operating decision can become a national political topic
SCOPE
34% of the customer base reacts, beyond the areas concerned
TIME
beyond 8 weeks without visible compensation, rejection becomes structural
Making these effects visible does not indicate which network policy to adopt. It indicates where an operating decision stops being read as one.
COMPARISON
A comparative reading drawn from the simulated reactions. No scenario is free of cost. A fast removal maximises economic efficiency and concentrates friction. A conservative approach protects local presence and keeps a fleet whose usage keeps declining. A differentiated strategy better protects the least substitutable situations, but requires place-by-place negotiation and pushes the targeted saving back by six months.
THE MOST ROBUST
Access alternative announced up front + prefect-mayor-press cascade + negotiated phasing + operational compensation within eight weeks
THE MOST FRAGILE
National rationalisation announcement with no visible alternative and no local relay
PROJECTION, THEN OBSERVATION
The committee adopted the leading scenario: a cascading announcement over six weeks, prefects first, then mayors, then regional press, then national press. The partnership with a major national postal operator, operational from week six, was placed at the front of the narrative in every statement. The mobile banking service, sized to cover the 1,200 towns concerned with a weekly rotation, formed the second pillar of the scheme. The removal was phased over 30 months, accelerated where the postal partnership was operational soonest.
Over the first 24 months, no national media crisis was triggered. Three local episodes of protest were identified, in towns where the postal partnership exceeded the 8-week threshold measured in simulation, and were resolved through a targeted acceleration of the compensation scheme.
LESSON
The machine was worth more than the number of withdrawals it recorded. 73% of rural segments had not used it for eighteen months and still opposed its removal: what the closure took away was not a use, it was an availability.
The measured gaps confirm it. The same plan gets 41% acceptability when announced nationally and 71% when announced through a regional cascade. The same compensation scheme gets 45% when mentioned at the end of the announcement and 67% when placed up front. The economic content is identical; what changes is what the decision says about the area.
TRADE-OFF
POSSIBLE FUTURES
RATIONALISE
Apply criteria based mainly on usage and operating cost
MAINTAIN
Largely preserve the existing network despite declining footfall
DIFFERENTIATE
Phased closure, operational alternatives and a regional cascading announcement
METHOD
This case adds a capability to the library: measuring option value. Some infrastructures matter even when rarely used, because they guarantee “I can do it if I need to.” Mobility, public services, health, finance, energy: usage data alone structurally underestimates what remains available is worth — and the signal sent by its disappearance.
This case comes from a simulation run for a European national banking network. The company is not named, no identifiable data entered the system, and the detailed results remain the client's property. The scenario comparisons published here are qualitative; the quantitative figures cited are those documented by the client.
The simulation covers behavioural reactions to different announcement scenarios. It is neither a land-use planning assessment, nor a stance on banking policy, nor an individual prediction.
Your next decision
Describe your need. We can point you to the right level of support.
State your decision. See the future it produces.
