Think about it; Mobility as a Service (MaaS) is essentially the same service levels of each mode separately and currently offered; and as separate offers based on the same service levels, why would the offer of telling someone they can use particular modes at the same service levels through a digital platform (App) (if no relevant incentives – not necessarily financial) be of sufficient difference to change travel behaviour?
Think of MaaS as adding another attribute (maybe more than one attribute) to an already unchanged set of attribute levels on unpackaged modes? This latter feature may explain why it appears to not be attractive enough for most people. Also, we know that many people are (sufficiently) familiar with travel options and know how to combine them without having to use an App? We should give the public more credit about knowing than is typically claimed in the MaaS deliberations. I call this the framing of MaaS.
Let us use an example. This is the situation today: assume a person uses the car and the car trip is 39 mins, parking cost is $20, tolls $10 but door to door. A multimodal trip without any discounts to what is on offer as separate modal services is a10 minute walk to the train, a 5 minute wait for the train, 50 minutes on the train, a $5 fare and a 10 minute walk to the destination. What does MaaS do? We could tell you about this non-car trip if you did not already know!
But knowing about an often inferior good (in terms of attributes that matter) is not enough to change travel behaviour, a reason why someone does not use it at present (Hensher et al. 2021). So what else needs to be on offer? For a start, we need to make the private car less attractive (something many of us have been promoting for years), which seems to not be possible, and/or make the modal components of the alternative more attractive. But this latter initiative can happen outside of MaaS (planners and service providers have been doing this for years); so if the additional attribute(s) associated with MaaS that is (are) claimed to represent the appeal of MaaS is (are) not significant, MaaS may add no utility at all. So what is this missing attribute(s)? That is the crucial question. This comes down to what I refer to as differential effort and seamlessness beyond what is already known and available being the missing ingredient (Hensher and Xi 2022). Is the effort of engaging in a multi-modal interrogation via a digital app a sufficient investment in return for an expected utility gain?
Hensher and Xi (2022) suggest that effort relates to both the outlay of effort required to find the best or acceptable way of moving around and once decided, the actual human effort required to undertake such movement; and seamlessness refers to the supplied service opportunities that enable a frictionless movement from beginning (origin) to end (destination) of a movement activity. To quantify the “user effort and seamlessness” of a multimodal trip is a serious empirical challenge; for example, each travel mode might be assigned an inconvenience cost per unit of time, and users set preferences on their maximum acceptable inconvenience cost, referred to as user inconvenience tolerance, an idea proposed by Hainong Xi in her PhD thesis. Intuitively, the inconvenience cost of travel modes aims to capture discomfort in shared and public transport modes and transfer among different modes (Xi et al. 2020). The inconvenience cost per unit of time can be assumed to be mode-specific and based on vehicle occupancy (the number of shared riders in a vehicle). The broadening that comes with subscribing to a package of modes as a package of modal and non-modal services with attractive incentives (Hensher and Mulley 2021) may be a significant source of gained utility in which inconvenience tolerance is cushioned by the benefits offered through services outside of modal supply.
The inability to change the levels of service associated with available modes is often the biggest stumbling block, since we are not aware of how one might differentiate this within a MaaS offer compared to purchasing each mode outside of a MaaS offer, apart from sharing information. This is where financial and non-financial incentives will be important, but to date we have not found the ones that will really make a sufficiently noticeable difference to enable MaaS to change travel behaviour in ways aligned with achieving societal sustainability goals. The eco-system of MaaS appears to have many challenges to face and resolve before we can see a pathway to scalability, and even a business case that might have commercial legs (Hensher et al. 2020).
It is time to start taking co-design or co-creation much more seriously and broaden the potential sources of positive utility, since the testing and/or introduction of a limited number of MaaS packages (as well as pay-as-you-go (PAYG) options) may be at the centre of a misplacement with the essential conditions associated with effort and seamlessness that matter across a very heterogeneous population of travellers. To achieve scalability with this approach, we will have to create a platform that allows a significant number of users to mix and match relevant services (defined inside and outside of the modal offerings), and package them to a bundle. Persona are interesting as initial market segment guides, but are insufficient in guiding the design of an appealing set of MaaS offers.
An important, and final, comment is that many of the points raised are written from an individual traveller perspective. Looking from a household (or group) position with the ability to share / donate mobility resource to each other may be more attractive as a way forward.
Acknowledgment: I thank John Nelson, Chinh Ho, Corinne Mulley, Aitan Militão, Hainong, Xi (ITLS) and Andy Taylor (Cubic) for their comments and feedback. I have, with permission, included a number of quotes in footnotes from these colleagues as a way of recognising their views and comments. We also recognise that there is an extensive and growing literature on MaaS with Hensher et al. (2020) covering many of the contributions, but it is not appropriate to include a literature review in this paper which is designed to summarise a lot of lessons learned as I and colleagues have investigated in detail elsewhere MaaS over the last 7 years.
 In discussions with Chinh Ho (23 February 2022), he comments that “[An] App is just a digital platform to interact with users, and whether we realise the benefits of MaaS or not will depend on what we build into that app. If it is just a multimodal platform, no doubt the benefit would be very marginal. Hence the need to distinguish different levels of integration.” Corinne Mulley adds “…a superior (or not so superior) journey planning app masquerading as MaaS is likely to do little to change travel behaviour…. We have known that there is a soft benefit for having everything in one place, and this applies to journey planners too, but the evidence is that the willingness to pay, where it is positive, is small. So this is unlikely to encourage too much take up. Putting this in a choice context, as you have, makes it starkly clear.” Chinaei et al. (2022) show how distributed architecture can be developed to realise a MaaS digital network using blockchain technology; and while this is a very appealing way to “…offer benefits to the government, users and industries by enhancing their access to everybody else’s assets in a secure way” (page 7), it is still not clear how this will change travel behaviour in a non-marginal and societally sustainable compliant way, which is the ultimate reason for MaaS.
 And when electric cars appear at a scalable level under the current road pricing model, they will be an attractive purchase, with a lower price to purchase and also to use, making the challenge of MaaS even more demanding (Hensher 2020). While this action may appeal to those who want to do good things for the physical environment (a ‘tinge of guilt’ removed), it will not assist growing levels of traffic congestion, or getting people out of their private cars.
 Unless we can fold it into a MaaS offer as a corporate car choice, as set out in Hensher et al. (2022) in what we refer to as Electric Car Sharing as a Service (ECSaaS). To date, most MaaS products have focussed in particular on public transport, but sadly in many jurisdictions the adage ‘to make public transport more attractive we have to make the private car less attractive’ still applies.
 Furthermore, multi-modal offerings may be desirable through a digital platform if payment is required in advance of travel. Historically in NSW for example, the Travel10 public transport ticket had critiques, with single-ticket users questioning why Transport for NSW receive their $20 payment in advance (i.e., prepaid for 10 single tickets). All else being equal, including the level of service and no discount, packaged services might reduce rather than increase utility, at least for some people, because the subscription fee, pre-paid, could stay in an offset account / bank account to gain interest by others (just like pre-paid tolling does).
 Andy Taylor in correspondence (22 February 2022) says that “I like the concept of inconvenience tolerance. I think the applicability of this across all modes, including private car usage, would be an exciting methodology that could provide insights into how far we have to nudge someone's behaviors to get them to change mode. As a baseline, though, it can help explain the reluctance of 'MaaS' adoption of multimodal mobility.”
 John Nelson in correspondence (24 February 2022) makes the point that “… I like the focus on effort and seamlessness. This highlights the inconvenience typically associated with public transport and other shared modes as well as the penalty of transfer among different modes. But the easiest way to ease this is by enhanced information provision which takes us back to the good journey planner.”