Research Explainer · Pooled Mobility

Who actually uses pooled rides — and what keeps them coming back?

UberPool flopped. LyftShare never scaled. Yet some pooled services thrive. A new study digs into the real psychology and demographics behind who adopts shared rides — and what predicts how often they use them.

Abouelela · Tirachini · Chaniotakis · Antoniou
TU Munich · U. Chile · UCL London
1,118 users · Mexico City · 54,175 trip records
Likelihood to be a frequent user
👩‍💼
High-income woman, full-time Car at home · commutes north→center
↑↑ High
🧑‍🎓
Young professional, no car PT user · lives in CDMX center
↓ Lower
👨‍💻
Mid-income man, car owner Long commute · avoids parking
→ Medium
👩‍🔬
Fem. survivor of harassment in PT Security-driven · van user
↑↑ High
Scroll to explore

Pooling has enormous potential — but most people skip it

The maths of pooled rides is compelling: put 8 strangers in a van going the same direction, and you remove 7 private cars from the road. Reduce vehicle kilometres travelled. Cut emissions. Ease congestion. The theory works beautifully on paper.

The practice has been more humbling. Studies estimate that only about 20% of ride-hailing users choose the pooled option when it is offered. Of those, the trips where more than one rider actually gets matched — a "real" pooled trip — account for just 2–7% of all ride-hailing journeys. Services like UberPOOL and LyftShare never cracked mass adoption despite years of subsidised fares.

Why? And what can we learn from a service that does work? This study — using unusually rich individual-level data from Jetty, a pooled-ride platform in Mexico City — attempts to answer exactly that.

20% Share of ride-hailing users who choose pooled when available
2–7% Trips that materialise as genuinely shared rides
50% Pooled share needed to meaningfully reduce VKT (Rodier et al.)
§

Four questions, one rich dataset

The researchers combined three complementary data sources: a detailed online survey of 1,118 Jetty users in Mexico City; six months of actual trip records (54,175 trips) linked to those same users; and GTFS public transit files mapping each user's proximity to Metro, BRT, and bus services.

This combination is rare in shared-mobility research, where most studies rely on aggregated data. Having individual trip records linked to survey responses meant the team could test whether stated preferences matched actual behaviour — and study not just who uses the service, but how often, and in which vehicle type.

The scarcity of user-level data has hindered the investigation of sociodemographic factors driving many dimensions of shared services.

They answered four research questions: What makes someone shift from their current mode to a pooled service? What determines whether they choose a van versus a bus? And what predicts how frequently they keep using it — including whether underlying travel attitudes (not just demographics) matter?

§

Who is likely to adopt pooled rides?

Based on the paper's hybrid choice model, explore how different user profiles affect adoption likelihood. The model found that demographics and latent travel attitudes jointly drive the shift from car-based trips to pooled services.

Adoption Likelihood Profiler

BASED ON TABLE 9 HCM RESULTS — TIRACHINI ET AL. 2022

Medium
Moderate adoption likelihood

This profile has some factors favouring adoption (car ownership, income) but neutral transit habits reduce the shift probability from car trips to pooled rides.

Car owner ↑ Mid income → PT neutral →
§

What actually drives people to switch?

The hybrid choice model revealed that adopting pooled rides — specifically shifting away from car-based trips — is shaped by a mix of hard demographics, trip context, and latent travel attitudes. Here are the most significant factors:

🚗 Car ownership

Having 1 or 2+ cars in the household significantly increases the likelihood of shifting to Jetty from car trips. These users have a real alternative to give up — and the model captures it.

↑ Increases adoption from car trips
💰 Personal income

The strongest sociodemographic predictor. High-income users (40K+ MXN/month) are far more likely to shift, use vans, and use the service more frequently than lower-income groups.

↑↑ Strongest sociodemographic factor
👩 Gender (female)

Women are more likely to shift from car-based trips to Jetty and to prefer vans over buses. Women who cite harassment avoidance as a reason are among the most frequent users.

↑ More likely to shift and use frequently
🎂 Youth (18–35)

Younger users are significantly more likely to shift from car trips to pooled rides — consistent with the general profile of shared mobility users across different cities and services.

↑ Affects shift decision (not frequency)
🚇 Frequent PT use (latent)

People who habitually use metro and buses are actually less likely to shift to Jetty from car trips. This latent attitude was the second strongest predictor in the model — bigger than most demographics.

↓ Reduces shift from car trips
📱 Multi-tasking

Users who use their smartphones, work, read, or sleep during Jetty trips are more likely to shift and to use the service frequently. Being a passenger — not a driver — is a productivity gain.

↑ Activity during trip drives adoption
🅿️ Avoiding parking

Citing parking as a reason is a significant predictor of shifting from car trips — especially for male users, who drive more and own more licences in this sample.

↑ Notably stronger for men
🏠 City residency

Users living outside CDMX proper — in the metropolitan periphery — are more likely to shift from car trips to Jetty. Jetty was specifically designed to serve the poorly-connected northern suburbs.

↑ Outside city core → higher adoption
§

Van or bus? The choice reveals a lot

Among users, 98% ride either buses (68% of trips) or vans (30%). The bus is cheaper, larger, and more "transit-like"; the van is smaller, pricier, and closer in feel to a private car. Who chooses which — and why — illuminates how users trade off cost against comfort.

  • More likely to be female, full-time employed, high-income
  • Access and egress Jetty by walking or cycling (shorter distances)
  • Work during the trip — the van is closer in quality to a private car
  • Located where Metro headways are shorter (better PT alternative nearby — yet still choose Jetty)
  • Cite security against theft as a reason — smaller vehicle feels safer
  • Replace more car trips (15% would have driven themselves vs. 10% for bus users)
  • Less frequent overall Jetty users (cost-sensitive at higher fares)
  • More likely small households (1–2 persons) and lower-income groups
  • Most frequent Jetty users overall — bus is affordable enough for daily use
  • Willing to walk longer distances to access stops
  • More likely to talk on the phone — social comfort in a larger group
  • Cite seat booking and ease of payment as key reasons (transit-like behaviour)
  • Primarily replace public transit trips — Metro + Microbus combos (23%)
  • Smaller VKT-reduction impact because they were already using shared modes

The policy implication is sharp: van users are the sustainability win. They pull car drivers out of cars. Bus users are important for equity and network coverage — but their contribution to reducing vehicle kilometres is smaller because they were never driving in the first place.

§

What keeps riders coming back?

On average, users made 1.7 trips per week on Jetty across the seven-month study period. But use is distributed very unevenly — a core group of daily commuters drives the majority of trips.

Distribution of Jetty use frequency (N = 1,118)
4+ times/week
36%
1–3 times/week
28%
1–3 times/month
23%
< Once/month
13%

The model identified several key predictors of use frequency — and some surprising non-findings:

Users with longer average Jetty trips use the service more often. This makes intuitive sense: for short trips, PT or walking might be competitive. For long, complex commutes — especially ones that would otherwise require 2–3 mode transfers — Jetty offers time savings that increase with distance. The convenience premium is larger when the alternative is genuinely painful.
Users whose nearest Metro station has longer headways (more infrequent service) use Jetty more often. This directly models Jetty as a complement to public transit gaps. Where the metro runs frequently, users rely on it. Where it runs poorly, Jetty fills the void. This has strong implications for where new pooled services should be deployed: target the gaps in formal transit coverage.
Users who sleep, study, read for pleasure, or work during Jetty trips are significantly more likely to be frequent users. This points to a "reclaimed time" effect: being a passenger rather than a driver converts dead commute time into productive or restorative time. Users who experience this time gain are more motivated to keep using the service. Service operators should lean into this — comfortable seating, stable Wi-Fi, and quiet environments are not just nice-to-haves.
The further users have to walk or travel to reach a Jetty pick-up point — or from the drop-off to their destination — the less frequently they use the service. This is a classic last-mile finding, but it has a clear operational implication: dynamically updating pick-up and drop-off locations based on actual demand patterns can directly boost frequency. The research notes Jetty already does this to some extent.
The interaction term (Female x Reason: Security Against Harassment) is statistically significant and positive: women who specifically use Jetty because it protects them from the harassment they experience in CDMX public transit are among the most frequent users. This demographic is not a niche — it represents a genuine quality-of-life improvement that the formal transit system is failing to provide. Safety is not a soft feature; it is a core service attribute driving retention.
Interestingly, citing "fare" as a reason to use Jetty is associated with shifting to the service from car trips — because Jetty is cheaper than taxis and e-hailing. But the same factor negatively predicts frequency of use, because Jetty is expensive compared to metro or colectivos. Users who joined because of cost savings relative to their previous mode will reduce usage as costs accumulate over time. This tension matters for pricing strategy.
§

Five things city planners should take from this

01
Target car users — not transit users

The pooled service only reduces traffic when it pulls people out of cars and taxis, not out of buses and metros. Route design, marketing, and pricing should be calibrated to attract car-owning, car-driving commuters — the people currently adding the most VKT to the network.

02
Deploy in transit gaps, not in transit-rich corridors

Jetty use increases where Metro headways are long. Pooled services work best as genuine complements to formal public transit, not as competitors to it. Urban authorities should map transit gaps and fast-track pooled service licences in those areas.

03
Safety is not a brand attribute — it is an infrastructure gap

The finding that women who use Jetty primarily to escape harassment are the most frequent users is both the most compelling and the most troubling result in the paper. It indicates formal public transit in Mexico City is failing a fundamental obligation. Pooled services can help — but they cannot substitute for the deeper problem.

04
Minimise the last-mile penalty

Access and egress distance directly reduces frequency. Dynamic relocation of pick-up and drop-off points in response to demand clusters is one of the highest-leverage operational decisions a pooled service can make.

05
Address equity before scaling

The service is dominated by high-income, highly educated users with cars. This is a common pattern in shared mobility globally — but it means the service currently provides a premium option for already-privileged commuters. Integration into formal transit with fare subsidies is essential before treating pooled rides as a city-wide solution.

Go deeper

This explainer covers the main findings. The full paper includes complete model tables, the hybrid choice model specification, EFA factor loadings, and a detailed comparative analysis of van versus bus users. Open access — free to read.

Read the Full Paper →

Abouelela et al. (2022) · Transportation Research Part C Vol. 138 · DOI: 10.1016/j.trc.2022.103632 · Open Access CC BY 4.0