Speed is the holy grail of food delivery. Platforms like Uber Eats, Deliveroo and Just Eat push hard on short delivery times, with real consequences for riders: time pressure in dense urban traffic, and elevated exposure to crashes and near-misses. New research from TU Delft (Kavta, Sharif Azadeh and de Almeida Correia, published in Transport Policy) asks a refreshing question: are consumers actually willing to give something up for their rider’s safety?
The answer is more nuanced than expected — and useful for anyone working on city logistics.
The study
The researchers ran a stated-preference experiment with 350 regular users of food delivery apps in Amsterdam and Copenhagen, two cities with strong cycling cultures where riders often use bicycles or e-bikes. Respondents faced choice tasks combining price, delivery time, and an explicit crash-and-near-miss risk indicator for the rider (shown as a percentage, a color code, and a risk letter A–D).
Two types of consumers
Using an advanced hybrid latent class choice model, the researchers identify two roughly equal-sized segments:
- Insensitive to rider safety (49.7%) — barely responds to the safety label and prioritizes speed and price.
- Sensitive to rider safety (50.3%) — willing to pay on average €4.06 more, or wait 18.75 minutes longer, for a 10-percentage-point reduction in crash risk.
The safety-sensitive segment skews younger, more highly educated, and more likely to be married. Men, on average, score higher on general insensitivity to road safety.
Relevant for last-mile policy
For city logistics, this is a meaningful signal. The authors compare their findings to earlier work on what riders themselves need to accept a safer but longer route: on average about 1.85 extra minutes for the same 10% risk reduction. The safety-sensitive consumer segment is willing to accept nearly 10 times as much waiting time as that. In other words, there’s real headroom between what riders need and what a substantial share of consumers are willing to give.
That opens the door to concrete interventions: an optional “safer delivery” mode in the app, visible risk labels at checkout, or dynamic pricing and time settings that reflect this preference for safety — without necessarily eating into platform margins.
A caveat
This remains a stated preference experiment: what people say they want can diverge from what they actually do when hungry and pressed for time. And the findings come from two cycling-friendly, safety-conscious cities — not automatically transferable elsewhere. Still, the core message stands: the assumption that consumers always choose the fastest, cheapest option only holds for half of them. For policymakers and platforms, that’s a good reason to look beyond delivery time as the sole performance metric.