URBAN PLANNING & MOBILITY · TRANSFORMATION
A French metropolitan authority was preparing a structuring 8 km tramway line, a €240M investment, meant to open up 40,000 residents, cut car traffic on the corridor by 30%, and carry 8,000 passengers per hour at peak times. Commissioning: 30 months after construction began.
The route ran through a historic shopping corridor of 340 shops, 60% of them independents with fragile cash reserves. The precedent set by another metropolitan area five years earlier was on everyone's mind: a 40% commercial die-off during construction.
THE PROBLEM
In the metropolitan area that set the precedent, more than a third of independent shops had disappeared during construction and never recovered after the tramway opened. A €12M compensation plan had produced no measurable result.
Evaluation reports had concluded that the corridor underwent an irreversible sociological transformation: closures, vacant units, then independent shops replaced by chains with deeper cash reserves.
So the problem wasn't building the tramway. It was preventing a temporary construction project from producing an irreversible commercial transformation. The project leadership brought in our system three weeks before the plan was due to be presented to the metropolitan council.
WHAT THE DECISION HAD TO RESOLVE
A DISTINCTION: A TRANSPORT LINE, AN ECOSYSTEM
The same stretch of road carries two different objects: infrastructure to be built, and a commercial fabric already in operation. The project was optimizing the first; the risk sat on the second.
A MECHANISM: THE SIGNAL MATTERS MORE THAN THE NUISANCE
Standard analysis attributes the commercial exodus to barriers, dust, noise and degraded access. The simulation identifies something else: an implicit signal sent to residents — “there's no point going there anymore.”
Active signage and monthly micro-events — pop-up markets, exhibitions, cultural programming — bring the simulated exodus down from 40% to 12%, with no change to the technical phasing. The most effective response is not, first and foremost, a technical one.
WHAT IS TESTED
This is not a space of seven communication variants. It is a systemic decision space: each phasing scenario is coupled with compensation and communication mechanisms in variable combinations.
2,800 synthetic residents, customers and shopkeepers were interviewed individually, with dynamic agents re-engaged at tipping points identified in real time.
The reconstructed population covers 840,000 residents and users of the target corridor, calibrated on public data — INSEE at the IRIS level, metropolitan commercial-flow data, regional retail observatories — and on proprietary data from the territorial Chamber of Commerce and Industry covering the corridor's 340 shops.
It is structured into 19 typologies crossing how long a shop has been established, its function within the corridor, its operating structure, its estimated cash reserves and its target customer base. No personal data was entered into the system.
CONTRASTING SHOP AGE
bakeries over 40 years old, family restaurants, artisans, and recent urban concepts.
FUNCTION WITHIN THE CORRIDOR
anchor shops, convenience shops, specialty shops: three roles, three exposures.
OPERATING STRUCTURE
independent, franchise, subsidiary: cash-reserve depth is not the same.
CUSTOMER BASES AND VISIT HABITS
target customers and their relationship to the neighborhood: what shifts elsewhere, and what comes back.
These configurations make part of the population's heterogeneity visible. The simulation runs on synthetic individuals, not on a handful of persona archetypes.
REACTIONS
THE REAL RISK ISN'T THE NUISANCE, IT'S THE ABANDONMENT SIGNAL
Customers don't stop coming because of barriers, dust or noise. They stop coming because the neighborhood sends an implicit signal: there's no point going there anymore. Active signage — “the shops are open” — paired with monthly micro-events in the corridor, brings the commercial exodus down from 40% to 12%, without changing the technical phasing of the works.
PHASING DETERMINES 70% OF COMMERCIAL SURVIVAL
Phasing in short 3-month segments, i.e. 12 weeks of maximum impact per shop, preserves 78% of independent shops. Phasing in long 8-month segments, technically optimal, causes 34% of them to disappear: cash reserves don't hold beyond 4 months without customer traffic. This finding contradicted the approach favored by the project's engineering firm.
HISTORIC SHOPS ARE THE MOST FRAGILE
Shops established for more than 20 years hold up less well: older customers who easily shift to other neighborhoods, owners nearing retirement who take the opportunity to close, a model built on customer habits that are quickly disrupted. Shops less than 5 years old hold up better: younger, more mobile customers, digital tools already in place, an entrepreneurial project still in its establishment phase.
THE MESSENGER MATTERS MORE THAN THE MESSAGE
Communication carried by the metropolitan authority gets 34% favorable reception: it is perceived as an institutional dressing-up of the constraint. Carried by an association of corridor shopkeepers, with the neighborhood's own visual codes and tone, the same communication gets 71%. Shopkeepers talk to shopkeepers, in the customers' own words.
SHOPS ARE INTERDEPENDENT
The closure of an anchor shop — a café-restaurant, a landmark bakery, a signature boutique — triggers a drop in footfall for neighboring shops for 8 to 14 weeks, regardless of the state of their own construction work. Targeting anchor shops, identified by their function rather than their revenue, produces a cumulative preservation effect 22 points higher than a uniform mechanism.
THE POST-CONSTRUCTION OUTCOME IS DECIDED DURING CONSTRUCTION
A corridor that has lost 34% of its independents does not recover within the 24 months following opening, even with passenger flow matching projections: vacant units are mostly taken over by chains. A corridor that has preserved 82% of its independents keeps its identity and attracts new independents within the 12 months following opening.
A MECHANISM: A TECHNICAL VARIABLE IS A SOCIAL VARIABLE
The engineer optimizes construction continuity. The shopkeeper bears the exposure duration. The same parameter — segment length — therefore carries two simultaneous functions: industrial performance and economic survival.
A decision's variables don't stay within the silo that produced them. An industrial optimum can produce a bad urban optimum: a decision perfectly rational within its own silo becomes irrational at the scale of the system.
A MECHANISM: HISTORY IS NOT ABSORPTION CAPACITY
You expect age to produce resilience. The simulation shows the opposite: it's the shops established for more than 20 years that hold up the least. This is exactly why real individuals and structures must be simulated instead of starting from intuitive assumptions.
A MECHANISM: A SHOP'S VALUE EXCEEDS ITS REVENUE
A shop carries a direct value — sales, jobs — and a systemic value: foot traffic it brings, the street's identity, a reason to come, an effect on its neighbors. The second doesn't show up in its revenue.
Compensation allocated on revenue alone therefore allocates resources poorly. Targeting anchor shops, identified by their function within the system, produces +22 points of cumulative preservation compared with a uniform mechanism.
A TEMPORAL MECHANISM: IRREVERSIBILITY
The tramway was not going to recreate what construction would have destroyed. The commercial trajectory after opening depends less on the quality of the launch than on the fabric preserved during the works — and the chain that runs from a closure to a new commercial structure cannot be reversed.
COMPARISON
Only the dimensions for which the case documents measurements carry numeric values. “Not published” is kept rather than an estimate.
SHORT SEGMENTS OF 3 MONTHS
12 weeks of impact per shop → 78% of independents preserved
LONG SEGMENTS OF 8 MONTHS
Continuous exposure → 34% of independents disappear
THE ACTUAL ROLLOUT
The metropolitan council adopted phasing in short 3-month segments, renegotiated with the engineering firm after a month of technical discussion, with a 4% industrial overcost accepted by the project owner.
Compensation was targeted at the historic shops and anchor shops identified by the project committee: €8M instead of the €12M under consideration for a uniform mechanism. An association of corridor shopkeepers was formed and given financial support to carry customer communication.
Twelve major monthly events were organized in the corridor, drawing an average of 4,000 visitors per event, complemented by ongoing social media communication and a quarterly print newsletter distributed in the neighborhood.
ARBITRATION
PROJECTION, THEN OBSERVATION
Over the first 24 months of construction, the commercial footfall trajectory tracked the projections with an average gap below 5%. Three episodes of local pushback were identified, in areas where the event calendar had fallen behind; they were resolved with reinforced support.
At 30 months into full rollout, i.e. 6 months after the tramway entered service, independent preservation reaches 82%, above the 78% projection. Post-construction footfall is up 24% versus the pre-construction period, against a projected +18%. The commercial occupancy rate stands at 96%, versus 91% before construction.
No vacant unit was taken over by a chain: the few vacancies were filled by new independent shops. The metropolitan authority has rolled the mechanism out across three other structuring transport projects in preparation.
These figures describe what was observed after the decision. They establish no exclusive causality between the strategy adopted and any of these movements.
LESSON
What needed protecting was the commercial ecosystem from the construction. Once in service, the tramway does not recreate a fabric that 24 months of construction have dissolved: a corridor that has lost 34% of its independents does not recover within the 24 months following opening, even with passenger flow matching projections.
The construction was temporary. The disappearance of the shops would have been irreversible. And the parameter that decided that tipping point wasn't in the compensation plan: it was in the technical segmentation of the construction works.
POSSIBLE FUTURES
A
OPTIMIZE THE CONSTRUCTION
Phasing in long 8-month segments, maximum industrial continuity
B
COMPENSATE UNIFORMLY
A broad €12M financial plan spread across the whole corridor
C
PROTECT THE ECOSYSTEM
Short 3-month segments, targeted compensation for historic and anchor shops, shopkeepers' association, monthly events
METHOD
840,000 synthetic residents and users calibrated on public data and on proprietary data from the territorial Chamber of Commerce, 19 typologies, 2,800 residents, customers and shopkeepers interviewed individually across 7 phasing scenarios coupled with 11 compensation and communication mechanisms, then projection of footfall trajectories over 36 months of construction and post-construction, modeling spillover effects between adjacent shops and dynamics of vacant-unit reoccupation: 133 distinct trajectories per scenario, monthly critical points and irreversible tipping thresholds.
This case adds four mechanisms to the library: a technical variable can be the most important social variable; a population reacts to a place's symbolic state and not only to its objective nuisances; the actors within a territory are interdependent, the disappearance of one changing the survival odds of others; a temporary shock can produce an irreversible transformation.
This case comes from a simulation carried out for a French metropolitan authority facing a €240 million structuring transport project. The client is not named, no personal data was entered into the system, and the detailed results remain the client's property.
The simulation does not decide the route, the budget, or the project's political calendar. It addresses the consequences, across different populations, of several phasing, compensation and communication architectures.
Your next decision
Describe your need. We can point you to the right level of support.
State your decision. See the future it produces.
