Carpooling (ride sharing, lift sharing, covoiturage) can reduce congestion, emissions and parking space needed by utilizing empty seats. It can be an important building block of MaaS systems. This blog post is about the factors making a carpool work – or not. It combines findings from scientific analyses as well as experiences from the former Open Carpool and some other carpooling systems.
No carpool system and their setting are the same. So there is no one size fits all set of factors which will work everywhere. But being aware of commonly known supporting and resisting forces can help to apply the right measures to make it a success.
The purpose of carpooling or ride sharing is to reduce car traffic. So from a driver perspective we are only interested in the motivation of non-profit drivers to offer empty seats where the driver would drive the route anyhow and passengers join the ride in addition. Rides originated by the passenger are therefore out of scope.
This is different from a passenger perspective, where the passenger could join any ride, which already happens – be it non-profit or for profit. So in addition to non-commercial private carpools, the passenger’s motivation could be the same for carpooling options of private non-profit carpools or for commercial driver services such as Taxi Pooling or pooling options from ride hailing services, such as Uber Pool or Clever Shuttle. Out of scope are rides with fixed schedules on fixed routes, such as regular public bus services.
A third party often neglected but to be considered is the carpooling system operator. They need to earn money in order to be able to develop and operate the carpooling platform.
It is also reasonable to distinguish between two very different use cases:
- Long distance travel
- Short to mid distance daily commute
In the travel use case quite some factors are different to the daily commute use case, such as: distance and related savings, planning horizon, matching probability due to a limitation on a limited number of cities compared to covering areas with an unlimited number of pick-up and drop-off points. Given the successful of some long distance carpooling systems in the market (e.g. BlaBlaCar or Quick Ride), this use case seems to be more viable than the short distance daily commute, where I am not aware of any profitable large scale system, except for closed company- or university-internal platforms, such as flinc, LiftShare, SAP TwoGo, SPLT, Wunder Carpool or Zimride. Most of the latter companies were operating open carpools before but gave up on this use case and now only focus on closed carpools.
Carpooling Success Factors
What makes a carpool successful? Given our focus on sustainable mobility this answer is relatively simple: it should reduce car miles driven. A simplified (and overestimated) proxy for this metric could be the miles driven by carpool passengers. Another success indicator would be the number of “matches”, where passengers and drivers could successfully share a ride. To get these success metrics up, as many people as possible need to be motivated to join a carpool as drivers or passengers.
What are the supporting and restricting forces for this motivation? What should be looked at when establishing a new carpooling or ride sharing system? The answer: it depends. It depends on many external factors and cultural aspects as well on the user’s individual situations.
As an example some studies state that women are more likely to carpool than men, other studies state the opposite. So what is true? Probably both. In countries like Saudi Arabia, women will likely not carpool on their own, whereas in France they likely will. So generalized data have limited predictive value for specific implementations. But the list below may work well as a check list on what to look at when getting a carpool implemented or improved. In the style of a Force Field Analysis they are listed as supporting or restricting forces. If it these forces are mainly affecting the passenger, they are in the upper part, if they concern the driver, they are in the lower part and if the affect both, they are rather centered between them.
These factors have an influence on the willingness to carpool for the passenger or driver. They are partially given by the external situation in which the carpool operates and partially specific to the user (internal). External factors could also be set as interventions to influence the motivation to share rides, e.g. with the highly effective High-Occupancy Vehicle (HOV) lanes. The different judgement factors have been grouped into seven categories:
1) Transport Situation
If I have to catch a plane to be on time for my own wedding, my transportation need is very different than if I just want to go to for a drink on my own. In the first scenario I will take no risk, in the second one I don’t mind and “no transport” would also be an option. Another strong factor are transport alternatives. There is a good reason why you can rarely find any carpools next to a good and reasonably priced public transport. If you need to transport children, large or heavy goods, carpools are also less likely an option. And needless to say: you can only offer a seat in a car if you have one 😉 With the young generation in urban areas this is no longer as common as it was 50 years ago.
Need to ride ⇒
Car & seats available ⇒
⇐ Transport alternatives
⇐ Kids, goods to transport
⇐ Legal or insurance risk
Safety and security are “must-be” factors. This starts with a safe and licensed car and driver but must also cover potential criminal intentions both from drivers and passengers. In India as an example the government requires strict “Know Your Customer” (KYC) checks for all carpools after several cases of raping and murder. Luckily this is a rare exception. Knowing your “customer” is a benefit of so called “closed” carpools within companies or universities. Carpool companies can build trust by applying “Duty of Care” measures such as identity checks, sharing participant photos and implementing feedback systems.
Duty of Care ⇒
Trust, Feedback ⇒
Closed system ⇒
⇐ Open/public system
⇐ Crime rate
If the question “Is it safe?” is answered with “Yes“, the next question will likely be “Does it work reliably?“. And this is where most carpools struggle.
A carpool is a two-sided marketplace where drivers supply rides and passengers demand them. As with every marketplace it only works if there is sufficient supply matching the demand. Through positive cross-side network effects the value of a carpool increases exponentially with the number of users. So bottom line the more users a carpool has, the better it works. Ideally every potential driver and every potential passenger would belong to one single carpool – or a set of carpools combining their supply and demand in a “meta carpool”. Unfortunately this is rarely the case and most carpools struggle to reach a “critical mass”, i.e. a minimum number of drivers and passengers to reach a reasonable likelihood to find a matching partner for a ride.
How big does this critical mass – or more precisely “critical density” need to be? This is a difficult question and would be a great topic for further scientific research. One approach to answer it would be to look at how often a user is willing to try finding a match ride before giving up. It is also important to understand that from a passenger perspective it is not about finding a single ride but about finding rides in both directions. From our own survey and also from what we could see in our data, users try it 2-3 times before they give up. So the minimum probability to find a matching ride needs to be above 50%. The higher the better. How to increase this matching probability would be a big enough topic for another blog post. Below just a few factors are mentioned:
D Ride back guarantee ⇒
High ride frequency ⇒
High matching probability ⇒
⇐ Flexible working hours
⇐ No Shows
⇐ Spread of pickup points
“What’s in for me?” For drivers the answers depend on the rewards offered. For passengers this question has at least two answers: in addition to social benefits they can save time and money.
4) Time Benefits
There can be direct time savings of commute time through the use of HOV lanes and no need to look for a parking lot and passengers can use the commute time alternatively: instead of focusing on traffic, they can read or simply relax. These time benefits can be reduced by required time to find a matching ride or multiple stops and detours to pick up passengers.
Alternative use of time ⇒
HOV lane allowance ⇒
Parking availability ⇒
⇐ Matching effort
⇐ Multiple stops
5) Financial Benefits
Financial benefits heavily depend on regulations and how the reward system is set up. The least common denominator is a passenger’s contribution to the driver’s fuel costs. For short distances these are often minimal amounts, which make short distances less attractive for drivers – unless they do not need to awkwardly ask for small change at every ride. This is where automated billing and payment gets important.
In addition to the passenger’s direct contribution the driver’s costs, there might be corporate or public funds or exemptions from parking fees, taxes, etc., which on the other hand often lead to more bureaucracy.
Price transparency ⇒
Cashless payment ⇒
Shared fuel costs ⇒
(Tax exemptions) ⇒
(Free parking) ⇒
6) Comfort & Convenience
These terms are often excluded in scientific research as they are hard to define and measure. But they seem to be too important to be ignored. When I asked non-carpoolers for their reasons not to offer or take a shared ride, I too often heard answers like: “You’d better don’t talk to me before I had my second coffee in the morning” or “My car is my castle” – especially when you use to sing out loud in your car while driving. A commonly requested feature for Open Carpool was a “Black List” where users could block other users, whom they didn’t want to pool with. Many users also want to retain their time flexibility. The need to switch modes of transport to go to and from a pickup point can also be a stopper.
Door to door connection ⇒
Black List ⇒
⇐ Seating space
⇐ Harsh weather
⇐ On-route pickup
With a growing recognition for the need to protect our climate and reduce congestion and air pollution, altruistic motives becomes more important.
Climate change ⇒
Air pollution ⇒
Sharing attitude ⇒
⇐ “Man-made climate
change is fake news” 😉
All these factors play a role in a driver’s or passenger’s individual decision to attend a carpool or refrain from it. A good understanding on which factors are most important and potential show stoppers – and also about how they change over time – will determine if a carpool system will be successful or not.
The following criteria do not influence user’s decisions but can give indications on which target audiences might more or less open to carpool. In the scientific research not all of these factors are conclusive and some are very country-specific. So reviewing local research in preparation of a carpool implementation might be worth the effort.
- Number of people in household
- Number of cars in household (mandatory criterion for drivers)
- Marital Status
Please contribute to this post by adding your point of view to the discussion using the comment function and so help other carpooling initiatives to learn!
- Olsson, Lars & Maier, Raphaela & Friman, Margareta. (2019). Why Do They Ride with Others? Meta-Analysis of Factors Influencing Travelers to Carpool. Sustainability.
- Neoh, Jun Guan & Chipulu, Maxwell & Marshall, Alasdair (2015). What encourages people to carpool? An evaluation of factors with meta-analysis. Transportation.
- Tsao, H. J. & Da-Jie Lin (1999) Spatial and Temporal Factors in Estimating the Potential of Ride-sharing for Demand Reduction. Computer Science
- Glasnapp, James (2016) Nine requirements for a successful private ridesharing system. TechCrunch
- Heinitz, Florian (2020) Potenziale und Hemmnisse für Pkw-Fahrgemeinschaften in Deutschland. Umweltbundesamt (German language)
Icons (c) Creative Stall and Orin zuu from the Noun Project
9 thoughts on “Carpooling Lessons Learned”
In my opinion the most relevant factor is the effort/benefit ratio for the drivers. If they receive a meaningful benefit, they would offer enough rides. In consequences passengers would find enough rides and the whole thing would start to fly.
The problem with short distances is that the little money a driver could earn does often not justify the effort. E.g. for just $3 I would not be willing to take a detour, take phone calls or wait for a passenger. For $30 I would, which is what I could get for a long distance ride. Another non-monetary benefit could be time savings via the mentioned HOV lanes or dedicated carpool parking, where parking is limited or very expensive.
The same is true for car pooling platform companies. If they take e.g. 10% of the driver fee, their variable costs are likely higher than the earning on short distances. Just based on a share of the driver fees the unit economies don’t work out and it will be hard to generate a positive margin – unless they receive additional funding.
Folks, what do you think?
Thank you Chris,
I had a chat on your comment with a former CEO of one of Germany’s most successful carpooling platforms. He agreed to your comment on the need for additional funding at least for short/mid distance rides. Another problem for these platforms is the “grey market” of users who met via the platform once and then match outside it.
He also argued if private carpools will ever be able to provide the 99% fulfillment and reliability users expect from all modes of transport. Without a fallback option, it won’t work otherwise.
Very interesting points! I believe that your view on giving enough benefits for drivers to make the carpooling marketplace work is very accurate and that in short distance commuting it is still difficult to manage that with only offering monetary incentives (cost-sharing with passengers).
We at Wunder Carpool (https://www.wundermobility.com/carpool) launched carpooling communities in very different locations and cultures and came to the conclusion (similar to Clemens view on his article) that there isn’t one universal answer and that carpooling is also not going to work for every environment. In places like the Philippines, the cost of owning a car and commuting every day is enough for people to carpool for monetary incentives. But in Europe we realized this wasn’t the case. Since we don’t have HOV lanes over here yet (at least at a larger scale) we discovered that parking is a pain that is big enough for drivers to consider carpooling as an option to get either preferential, reserved or free parking space. Thats why we added an integration to our platform for companies to manage parking allocation for carpoolers and this made the activation and retention of drivers good enough to have a functioning carpooling community! This works very well for larger companies with parking issues and it is a cost-effective way for companies to optimize their parking infrastructure, while still providing benefits to their employees .
Regarding the carpooling platform companies, I can tell you from our experience from the B2C sector that it is possible to achieve a positive margin, but only once you reach a big enough scale to lower operations and technology costs enough to actually profit from a market. This scale is in the millions of rides per year and to achieve that you need the right variables in the market and very high investment in education and acquisition costs. So it is possible, but until there is more support from the government (doesn’t need to be subsidies, can also be accomplished with regulation that promotes carpooling as seen in the US) I believe that these types of platforms will not expand to become mainstream in most countries.
Wunder Carpool is a carpooling solution with over 5 million rides and I have been working on launching and scaling carpooling communities since 2013, in case you (or anybody else) wants to discuss more about the subject I’m always happy to connect through Linkedin (https://www.linkedin.com/in/philippwenger/).
P.S.: Thanks Clemens for sharing this article, it is super clear and lays out the main success factors in a very comprehensible way!
It will be interesting to see how the carpooling functions with ridesharing companies like UberPool and Lyft Line will evolve and be used as these companies scale. While transit and carpooling is decreasing in the US, these ridesharing services are booming so it seems critical that they embrace more occupants in each car. It will be especially interesting to also see how the eventual rise of autonomous cars in these services utilize carpooling. From what I’ve read, both drivers and riders don’t like pooling options because of human tendencies to be late, be loud, be annoying, etc. It may be up to governments to develop regulations making pooling more attractive.
Thanks Clemens for this very interessting post!
A very nice and systematic display of the many problems/challenges of carpooling.
For all these “issues” (=requirements for a smart carpooling) I tried to formulate a suitable answer or requirement in my #eMIT concept. (If wanted I could describe point by point how #eMIT addresses these requests and provide it to you. The Option 21 plattform http://www.O21.de displays main principles of the #eMIT concept.)
Yes, it’s not that easy – but the technical possibilities are already there today – but for Germany at least unfortunately I don’t see the necessary political support yet to build and maintain such a solution sustainably for the common good.
I’m glad to have found this valuable blog now.
Pingback:Mobility Hubs – Mobility as a Service (MaaS)
Pingback:Rural MaaS and the Next Big Thing – Mobility as a Service (MaaS)
[* Shield Security plugin marked this comment as “Trash”. Reason: Failed Bot Test (checkbox) *]
It’s really cool
Pingback:10 Ways To Stop Global Warming - The Asia Press