The cost of standing still
European Freight Efficiency Index — seven capital cities, one full year of data
- 7 European capitals
- Full-year 2025 baseline
- Scores 0–100 (higher = more efficient)
- Two pillars: infrastructure readiness · externalities
Overall rankings
Efficiency scores for each city, combining two pillars across two vehicle types. Higher = more efficient road network.
- 61
Berlin leads — the only city above 60 on both pillars
- 25
Madrid scores lowest — zero uncongested hours for both vehicle types
- 144%
Efficiency gap between best and worst city across the vehicles analysed
- 5/7
Cities where trucks outperform passenger vehicles on efficiency
| Rank | City | Score (Overall) | Readiness · externalities | Band |
|---|---|---|---|---|
| 1 |  BerlinThe benchmark | 61/100 | Readiness61 Externalities62 Passenger55 Truck71 | Strong |
| 2 |  AmsterdamThe externalities champion | 59/100 | Readiness55 Externalities72 Passenger55 Truck65 | Moderate |
| 3 |  DublinThe middle ground | 49/100 | Readiness48 Externalities52 Passenger43 Truck57 | Moderate |
| 4 |  RomeThe paradox | 48/100 | Readiness39 Externalities74 Passenger37 Truck63 | Moderate |
| 5 |  ParisPermanent equilibrium | 37/100 | Readiness30 Externalities58 Passenger31 Truck46 | Needs attention |
| 6 |  LondonThe dual challenge | 29/100 | Readiness21 Externalities52 Passenger29 Truck27 | Needs attention |
| 7 |  MadridBottom of the index | 25/100 | Readiness16 Externalities54 Passenger28 Truck22 | Needs attention |
- Strong (60+)
- Moderate (40–59)
- Needs attention (0–39)
How easily vehicles move through the network — congestion volume, hours of free-flowing traffic per day, and journey time predictability.
The waste produced by vehicles in the network — primarily mid-trip idling. Traffic that keeps moving, even slowly, scores higher than stop-and-start gridlock.
Each vehicle type scored separately then combined. Trucks and passenger vehicles experience the same network very differently — both are in the score.
City profiles
Full score breakdown for each city — readiness sub-components, externalities, and the standout finding from the data.
Berlin
The benchmark
61OverallStrongRank 1 of 7Readiness61Externalities62Passenger55Truck71AUC — congestion burden52Hours below threshold59Travel time variability82Passenger readiness54Truck readiness73AUC = area under the curve — cumulative congestion burden. Score 0 = congested every hour. Threshold = hours per day where congestion stays below the critical level.
225harsh events / 1,000 trips
Berlin's polycentric structure — multiple centres rather than one dominant CBD — distributes demand more evenly than monocentric cities. The wide boulevard infrastructure provides arterial capacity most European cities lack.
Truck variability score of 95 — near-perfect predictability. The only city above 60 on both readiness and externalities.
Amsterdam
The externalities champion
59OverallModerateRank 2 of 7Readiness55Externalities72Passenger55Truck65AUC — congestion burden54Hours below threshold44Travel time variability68Passenger readiness51Truck readiness61AUC = area under the curve — cumulative congestion burden. Score 0 = congested every hour. Threshold = hours per day where congestion stays below the critical level.
356harsh events / 1,000 trips
Amsterdam demonstrates that a city doesn't need to eliminate congestion to achieve high efficiency — compact form, cycling infrastructure, and decades of optimised signal timing keep traffic moving.
Second-best externalities score (72). Vehicles idle less here than anywhere except Rome, despite moderate congestion.
Dublin
The middle ground
49OverallModerateRank 3 of 7Readiness48Externalities52Passenger43Truck57AUC — congestion burden42Hours below threshold42Travel time variability65Passenger readiness45Truck readiness52AUC = area under the curve — cumulative congestion burden. Score 0 = congested every hour. Threshold = hours per day where congestion stays below the critical level.
315harsh events / 1,000 trips
Dublin's stop-and-go pattern produces the joint-worst externalities score despite moderate overall congestion — suggesting heavy reliance on signal-controlled arterials and few dedicated freight corridors.
Most balanced pillar gap in the study — only 4 points between readiness (48) and externalities (52).
Rome
The paradox
48OverallModerateRank 4 of 7Readiness39Externalities74Passenger37Truck63AUC — congestion burden24Hours below threshold28Travel time variability80Passenger readiness27Truck readiness57AUC = area under the curve — cumulative congestion burden. Score 0 = congested every hour. Threshold = hours per day where congestion stays below the critical level.
393harsh events / 1,000 trips
Rome's traffic is slow but continuous — vehicles creep at low speed rather than stopping and starting. This produces less idling despite severe congestion. Trucks benefit from designated freight windows and professional routing.
Best externalities in the entire study (74). Largest truck-passenger gap of any city: +26 points.
Paris
Permanent equilibrium
37OverallNeeds attentionRank 5 of 7Readiness30Externalities58Passenger31Truck46AUC — congestion burden16Hours below threshold7Travel time variability82Passenger readiness26Truck readiness37AUC = area under the curve — cumulative congestion burden. Score 0 = congested every hour. Threshold = hours per day where congestion stays below the critical level.
1,191harsh events / 1,000 trips
Paris has reached a permanent congestion equilibrium — always congested, but consistently so. Variability of 83 means the city is predictably slow. Professional logistics still outperform general traffic by 15 points.
Passenger hours-below-threshold: 0. Congestion exceeds the critical threshold at every single hour of the day.
London
The dual challenge
29OverallNeeds attentionRank 6 of 7Readiness21Externalities52Passenger29Truck27AUC — congestion burden7Hours below threshold0Travel time variability69Passenger readiness21Truck readiness20AUC = area under the curve — cumulative congestion burden. Score 0 = congested every hour. Threshold = hours per day where congestion stays below the critical level.
462harsh events / 1,000 trips
London struggles on both dimensions simultaneously — extreme congestion and high unpredictability. Extensive bus lane restrictions reduce available road capacity for commercial vehicles, creating uniquely hostile conditions for freight.
Only city where both vehicle types score below 30. Lowest truck externalities (48).
Madrid
Bottom of the index
25OverallNeeds attentionRank 7 of 7Readiness16Externalities54Passenger28Truck22AUC — congestion burden0Hours below threshold0Travel time variability63Passenger readiness21Truck readiness8AUC = area under the curve — cumulative congestion burden. Score 0 = congested every hour. Threshold = hours per day where congestion stays below the critical level.
1,088harsh events / 1,000 trips
Both passenger and truck AUC scores are 0 — congestion exceeds the critical threshold at every hour. For truck operators, Madrid offers zero windows of relief, zero predictability, and no ability to optimise around conditions.
Truck variability: 33 — the lowest single measurement in the entire study. Less than half of any other city.
Passenger vehicles vs trucks
The same road network delivers very different experiences depending on vehicle type. Professional fleet management produces measurably better outcomes in most cities.
Average score across all cities: 36
Passenger vehicles face the full brunt of peak congestion with no scheduling flexibility.
Average score across all cities: 51
Professional routing, scheduled departures, and dedicated infrastructure give trucks a structural advantage in 5 of 7 cities.
| City | Passenger score | Truck score | Gap | Better performer |
|---|---|---|---|---|
Berlin | 55 | 71 | Trucks +16 | Trucks |
Amsterdam | 55 | 65 | Trucks +10 | Trucks |
Dublin | 43 | 57 | Trucks +14 | Trucks |
Rome | 37 | 63 | Trucks +26 | Trucks |
Paris | 31 | 46 | Trucks +15 | Trucks |
London | 29 | 27 | Passenger +2 | Passenger |
Madrid | 28 | 22 | Passenger +6 | Passenger |
The professional fleet effect
Trucks outperform passenger vehicles in five of seven cities. Professional fleet management produces measurably better outcomes than unmanaged traffic.
Trucks benefit from structured routing, scheduled departure times, dedicated freight infrastructure, and fleet-level data analysis.
London and Madrid are the exceptions — driven by bus lane restrictions, narrow delivery zones, and time-compressed delivery windows.
Idling deep dive
Mid-trip idling measures the waste produced by vehicles in the network — time spent stationary with the engine running.
- €2.6M
Total idle fuel cost across the vehicles analysed — fuel burned with the engine running, going nowhere.
- 15.60
L/100km — London passenger vehicles, the worst fuel consumption in the study. Paris manages 6.51 L/100km.
- 2.4×
The fuel consumption gap between best (Paris, 6.51 L/100km) and worst (London, 15.60 L/100km) for passenger vehicles.
Idling efficiency score by city
Externalities score (0–100) per vehicle type — higher means less mid-trip idling.
- RomePassenger69Truck81Truck +12
- AmsterdamPassenger69Truck76Truck +7
- ParisPassenger47Truck73Truck +26
- DublinPassenger39Truck71Truck +32
- BerlinPassenger60Truck65Truck +5
- MadridPassenger49Truck62Truck +13
- LondonPassenger54Truck48Passenger +6
Sorted by truck externalities. Bars on the same 0–100 scale — directly comparable between vehicle types.
Fuel consumption by city
Litres per 100km across the vehicles analysed — lower is more efficient. Sorted best to worst for passenger vehicles.
- ParisPassenger6.51Truck30.09
- MadridPassenger6.82Truck31.14
- RomePassenger10.03Truck29.52
- BerlinPassenger11.06Truck31.82
- AmsterdamPassenger11.10Truck32.27
- DublinPassenger11.97Truck35.93
- LondonPassenger15.60Truck36.75
Sorted by passenger fuel consumption, lowest to highest. Green = lowest consumption, red = highest.
Idle fuel wasted per trip
Litres burned per trip while stationary — engine running, vehicle not moving. Sorted best to worst for passenger vehicles.
- RomePassenger0.06Truck0.22
- MadridPassenger0.06Truck0.40
- BerlinPassenger0.07Truck0.62
- ParisPassenger0.09Truck0.66
- AmsterdamPassenger0.11Truck1.27
- DublinPassenger0.15Truck0.44
- LondonPassenger0.19Truck0.46
Sorted by passenger idle fuel per trip, lowest to highest. Amsterdam trucks (1.27L) reflect extended dwell time at loading bays, not in-traffic idling.
Why Rome leads despite heavy congestion
Rome's traffic is slow but continuous. Vehicles creep at low speed rather than stopping and starting. When a vehicle moves at 15 km/h it is technically in motion — not idling.
Amsterdam achieves its second-place position through optimised signal timing on a consistent canal-grid network that keeps vehicles moving through intersections.
Why London is the only city where trucks idle more than passenger vehicles
Extensive bus lane restrictions, narrow streets, ULEZ requirements, and concentrated delivery windows force commercial vehicles into stop-and-go patterns. Passenger vehicles can use bus lanes; trucks cannot.
Safety & harsh events
Harsh driving events — acceleration, braking, and cornering — measured from GPS-derived telematics data across all seven cities.
- 225
Berlin — safest city. Harsh events per 1,000 trips, lowest in the study.
- 1,191
Paris — most events per 1,000 trips. 78.5% are harsh accelerations.
- 5.3×
Gap between safest (Berlin 225) and most events (Paris 1,191) per 1,000 trips.
- 5/7
Cities where efficiency score and harsh event rate track together — the network shapes driver behaviour.
- 225BerlinEfficiency 61/100
Dispersed across western districts — no single bottleneck. The most operationally predictable city in the study.
- 315DublinEfficiency 49/100
N7/Naas Road generates 71% of cornering events in its top hotspot cell — a specific corridor risk.
- 356AmsterdamEfficiency 59/100
13.7% of all city events concentrate at the Schiphol motorway interchange — location-specific and solvable.
- 393RomeEfficiency 48/100
Slow-but-flowing traffic is inherently safer than stop-and-start gridlock. One of the lowest harsh event rates despite heavy congestion.
- 462LondonEfficiency 29/100Anomaly ↓ safer than score suggests
Congestion charging, active enforcement and lower aggression norms produce a safety record the road network alone would not generate.
- 1,088MadridEfficiency 25/100
A driver at Madrid's average would be 4.8× above Berlin's. Most variation comes from the network, not the driver.
- 1,191ParisEfficiency 37/100Outlier ↑ highest events
78.5% of events are harsh accelerations — drivers apply maximum throttle from every standstill. The Périphérique forces constant merge-and-weave.
Source: GPS-derived harsh events (acceleration, braking, cornering) · Full-year 2025 · Sorted safest to most events · Bars relative to the study range (Berlin 225 → Paris 1,191).
Paris — the outlier
1,191 events per 1,000 trips — the highest in the study, despite a mid-table efficiency score of 37. 78.5% are harsh accelerations. Drivers apply maximum throttle habitually from every standstill. The Périphérique forces constant merge-and-weave with no policy mitigation.
London — the anomaly
462 events per 1,000 trips — anomalously low given an efficiency score of just 29. London's safety record is a policy achievement, not a network design one. Congestion charging, active enforcement and lower aggression norms compensate for poor road infrastructure.
What the data says about city-normalised benchmarking
A driver averaging 1,000 harsh events per 1,000 trips is at the city average in Madrid and well above it in Berlin. Most variation in harsh event data comes from the road network, not the driver. Performance targets need to reflect the city your drivers operate in.
The road network structures when vehicles stop, how they accelerate, what the merge conditions look like at every junction. That's why efficiency scores and harsh event rates track together in five of the seven cities.
Methodology note: Harsh acceleration, braking, and cornering events are inferred from GPS position data rather than accelerometer readings. Results should be interpreted as relative comparisons between cities, not absolute safety metrics. Source: Geotab connected vehicle telematics, full-year 2025.
Methodology
How the EMEA Urban Efficiency Score is built — data sources, weighting, and what each component measures.
The EMEA Urban Efficiency Score is a composite metric ranging from 0 to 100, built on full-year 2025 connected vehicle data from Geotab's fleet management platform. All data is anonymised and aggregated — no individual vehicle operator or driver can be identified from the results.
Passenger vehicles
Reflects the majority of road demand
Trucks
Captures commercial logistics efficiency
Overall city score
Higher = more efficient road network
Infrastructure readiness
Measures the ease with which vehicles can move through the transportation network — combining congestion volume, windows of free-flowing traffic, and journey time predictability.
Measures the cumulative congestion burden across all hours of the day. Cities where congestion exceeds 120% of free-flow at every hour score 0; cities with significant free-flowing periods score higher.
Counts the hours per day where congestion stays below the critical threshold (congestion index < 120%). Cities that offer extended windows of manageable traffic score higher.
Measures how predictable travel times are across the day. Low variability means conditions are consistent, enabling better trip planning for logistics operators.
Infrastructure externalities
Measures the waste produced by vehicles operating in the network, focusing on mid-trip idling. Higher idle ratios signal congestion, poor signal timing, and stop-and-go patterns.
Lower idling = higher externalities score. A city where traffic keeps moving, even slowly, scores better than one where vehicles stop frequently.
Data notes
Sample, not census. Geotab data represents a sample of connected vehicles, not every vehicle on the road. Scores reflect normalised comparisons between cities rather than absolute traffic volumes.
Full-year analysis. All scores are based on full-year 2025 data, smoothing out seasonal variation, holidays, and one-off events. This provides a structural view of each city's traffic efficiency.
City-level aggregation. Scores represent the overall urban road network. Individual corridors or districts may perform significantly differently from the city-wide average.
GPS-derived harsh events. Harsh acceleration, braking, and cornering are inferred from GPS position data rather than accelerometer readings. Results should be interpreted as relative comparisons, not absolute safety metrics.
Fleet composition. Trip volumes reflect the composition of Geotab-connected fleets in each market, not the full vehicle population.
Data source: Geotab connected vehicle telematics · Full-year 2025 baseline · Seven European capitals: Berlin, Amsterdam, Dublin, Rome, Paris, London, Madrid · All figures scoped to the vehicles analysed · © 2025 Geotab Inc. All rights reserved.