Composition of a Mobility as a Service (MaaS) offer for university students based on willingness to pay (WTP) and its determinants

Keywords:

Abstract:

Purpose

Mobility as a service (MaaS) is an emerging technology reshaping the structure of the automotive industry (Adner and Lieberman 2021, 99–100). MaaS is characterized by the integration of several transportation modes, individualization through customized transportation offers, a digital platform and collaborative consumption, i.e. shifting from car ownership to sharing models (Jovic and Baron 2019, 185). A higher affinity for shared and multimodal transportation systems is especially given among the “generation young” (age 18 to 25), for whom the age-based life event of university studies influences mobility demand (Elder 1994, 5). Commuters of large universities and regional university alliances have been of focal research interest because their trip frequency, mode choice and the distance travelled differ substantially from those of the general population (Hafezi et al. 2019, 519; Khattak et al. 2011, 137). Furthermore, retaining students as customers increases long-term revenue as their future life earning is on average 2.7 times higher than individuals without vocational education (Schmillen and Stüber 2014, 1). However, students are a heterogeneous customer segment with significant differences in their mobility preferences (Groth et al. 2021, 6–13). Therefore, it is essential to understand the patterns that affect the importance of mobility services in general, as well as specific variables that lead to significant differences between customer segments.

The importance, or utility, is transformed into willingness to pay (WTP), i.e. the maximum price a buyer would accept for a given offer (e.g.  Günthner et al. 2021, 81). Although evidence is limited (Polydoropoulou et al. 2020, 153), preliminary research indicates that WTP for MaaS generally deviates from cost-covering prices (Liljamo et al. 2020, 110). Thus, when designing a MaaS model, its revenue potential must be considered, which is defined to a great extent by customers’ WTP. Revenue potential is increased by designing segment-specific MaaS offers (Hensher et al. 2020, 78). Even though WTP and MaaS bundling has been assessed in previous studies (e.g. Guidon et al. 2020), as well as segmentation of MaaS customer groups (e.g., Alonso-González et al. 2020), it has not been researched for the specific segment of university students. Consequently, this study aims to answer the following research questions:

1. What influences the students’ importance of MaaS attributes, and, therefore, WTP?
2. Which homogeneous student clusters result from benefit segmentation and in which variables do they differ significantly?
3. How does revenue potential change from offering one generic utility-maximizing MaaS package to creating segment-specific integrated MaaS offers?

Design

Within this research, a conjoint analysis, an indirect price query evaluating product bundles as a whole, (Eggers and Sattler 2011), is applied since it is the best method to estimate WTP for new products, such as MaaS (Baier and Brusch 2009, 3). The decisions made by respondents are transformed into part-worth utilities for attribute levels and subsequently into WTP by formulating a hierarchical bayes (HB) model, which represents a random effects model estimating parameters both at the individual and at the population level (Hein et al. 2020, 2). The influence of several variables, e.g. mobility practices and availability of transportation modes, on attribute importance and thus on WTP is assessed through multiple linear regression analysis. In a further step, benefit segmentation is applied for assessing the heterogeneous transportation offers based on attribute importance and by applying cluster analysis. The previously mentioned variables are examined in their ability to significantly distinguish clusters. In a final step, a market simulation including real competition is performed to identify the segment-specific MaaS packages that have the highest revenue potential.

The study includes the following transportation offers: car sharing, (e-)bike sharing, e-scooter sharing and on demand e-shuttles as possible MaaS components, each with three different service levels, as shown in Table 1.

Table 1: Elements of the MaaS offers and their monthly prices as an input for conjoint analysis. Own elaboration.

The research study is implemented for empirical application by the University Alliance Ruhr, consisting of the universities of Duisburg-Essen, Bochum, and Dortmund in the German Ruhr metropolitan area. Public transportation is not included in the analysis since all participants have a discounted subscription to public transportation.

Findings

A total of 1,487 students participated. The majority (n = 1193, 80.2 %) of the respondents were aged between 20 and 29, proving that university studies are indeed an age-based life event. Their income was relatively low, as 57,5 % (n = 855) had a monthly income below 1000 €. The hierarchical bayes model’s McFadden’s Pseudo R-squared is 0.599, proving the goodness of internal model validity (McFadden 1979, 313). The relative importance for each attribute is calculated by dividing the maximum range of one attribute’s level utilities (the coefficients) by the whole utility range.

The price attribute is by far the most important one, which is not surprising given the relatively low income of the respondents. For the regression analyses, first results indicate that both the price and the e-scooter importance is hardly explainable through the remaining variables as the effect of the models is only weak. However, car and bike sharing models have a moderate effect and plausible regression coefficients, but these preliminary results are still subject to validation.

The transportation attribute importance variables are used for benefit segmentation and for MaaS composition based on their differing WTP. The benefit segmentation through k-means results in a two-cluster solution with clusters only differing in WTP and in the preference of bike type included in the MaaS offer. The results indicate that although WTP in general is very small, presenting segment specific MaaS offers significantly increases the revenue potential. A more thorough report of the results will be provided at the Gerpisa Colloquium in June.

Practical and theoretical implications

The results contribute to research in the field of MaaS profitability by extending it to the promising segment of university students. As a practical implication, this means that companies that want to specifically target students with their MaaS offering will receive suggestions on what type of mode might be best integrated, what variables should be used for segmentation, and at what price MaaS should be offered.

For the theoretical implications, it is proven that more personalized, integrated offers provide students higher value, as stated by the service dominant logic (e.g. Vargo and Lusch 2004, 6), depending on the price. These findings serve as an input for ecosystem design, having clarified the focal value proposition based on the examined customer needs to create customized MaaS offers (Adner 2021, 11–16). The results are also valuable for creating the operating model on ecosystem level and the business model for the orchestrator, especially the profit model.

References

Adner, Ron (2021): Winning the right game. How to disrupt, defend, and deliver in a changing world. Cambridge, Massachusetts: The MIT Press (Management on the cutting edge series).

Adner, Ron; Lieberman, Marvin (2021): Disruption Through Complements. In: Strategy Science 6 (1), S. 91–109. DOI: 10.1287/stsc.2021.0125.

Alonso-González, María J.; Hoogendoorn-Lanser, Sascha; van Oort, Niels; Cats, Oded; Hoogendoorn, Serge (2020): Drivers and barriers in adopting Mobility as a Service (MaaS) – A latent class cluster analysis of attitudes. In: Transportation Research Part A: Policy and Practice 132, S. 378–401. DOI: 10.1016/j.tra.2019.11.022.

Baier, Daniel; Brusch, Michael (2009): Erfassung von Kundenpräferenzen für Produkte und Dienstleistungen. In: Daniel Baier und Michael Brusch (Ed.): Conjointanalyse. Berlin, Heidelberg: Springer Berlin Heidelberg, S. 3–18.

Bornstedt, Maike (2007): Kaufentscheidungsbasierte Nutzensegmentierung. Entwicklung und empirische Überprüfung von Segmentierungsansätzen auf Basis von individualisierten Limit Conjoint-Analysen. Göttingen: Cuvillier Verlag Göttingen.

Eggers, Felix; Sattler, Henrik (2011): Preference Measurement with Conjoint Analysis. Overview of State-Of-The-Art Approaches and Recent Developments. In: GfK MIR 3 (1), S. 36–47.

Elder, Glen H. (1994): Time, Human Agency, and Social Change: Perspectives on the Life Course. In: Social Psychology Quarterly 57 (1), S. 4–15.

Groth, Sören; Hunecke, Marcel; Wittowsky, Dirk (2021): Middle-Class, Cosmopolitans and Precariat among Millennials between Automobility and Multimodality. In: Transportation Research Interdisciplinary Perspectives 12 (8), S. 100467. DOI: 10.1016/j.trip.2021.100467.

Guidon, Sergio; Wicki, Michael; Bernauer, Thomas; Axhausen, Kay (2020): Transportation service bundling – For whose benefit? Consumer valuation of pure bundling in the passenger transportation market. In: Transportation Research Part A: Policy and Practice 131 (3), S. 91–106. DOI: 10.1016/j.tra.2019.09.023.

Günthner, Timo; Proff, Heike; Jovic, Josip; Zeymer, Lukas (2021): Tapping into market opportunities in aging societies - the example of advanced driver assistance systems in the transition to autonomous driving. In: International Journal of Automotive Technology and Management (IJATM) 21 (2), S. 75–98.

Hafezi, Mohammad Hesam; Daisy, Naznin Sultana; Liu, Lei; Millward, Hugh (2019): Modelling transport-related pollution emissions for the synthetic baseline population of a large Canadian university. In: International Journal of Urban Sciences 23 (4), S. 519–533. DOI: 10.1080/12265934.2019.1571432.

Hein, Maren; Kurz, Peter; Steiner, Winfried J. (2020): Analyzing the capabilities of the HB logit model for choice-based conjoint analysis: a simulation study. In: J Bus Econ 90 (1), S. 1–36. DOI: 10.1007/s11573-019-00927-4.

Hensher, David A.; Mulley, Corinne; Ho, Chin; Wong, Yale; Smith, Goran; Nelson, John D. (2020): Understanding mobility as a service (maas). Past, present and future. 1. Ed. San Diego: Elsevier.

Jovic, Josip; Baron, Paul (2019): Implications for improving attitudes and the usage intention of Mobility-as-a-Service - Defining core characteristics and conducting focus group discussions. In: ICoMaaS 2019 - Proceedings, S. 182–197.

Khattak, Asad; Wang, Xin; Son, Sanghoon; Agnello, Paul (2011): Travel by University Students in Virginia. In: Transportation Research Record 2255 (1), S. 137–145. DOI: 10.3141/2255-15.

Liljamo, Timo; Liimatainen, Heikki; Pöllänen, Markus; Utriainen, Roni (2020): People’s current mobility costs and willingness to pay for Mobility as a Service offerings. In: Transportation Research Part A: Policy and Practice 136 (2/2009), S. 99–119. DOI: 10.1016/j.tra.2020.03.034.

McFadden, Daniel (1979): Quantitative Methods for Analyzing Travel Behaviour of Individuals: Some Recent Developments. In: David A. Hensher und Peter R. Stopher (Hg.): Behavioural travel modelling. London: Croom Helm, S. 279–318.

Orme, Bryan K.; Chrzan, Keith (2017): Becoming an expert in conjoint analysis. Choice modelling for pros. Orem: Sawtooth Software.

Polydoropoulou, Amalia; Tsouros, Ioannis; Pagoni, Ioanna; Tsirimpa, Athena (2020): Exploring Individual Preferences and Willingness to Pay for Mobility as a Service. In: Transportation Research Record 2674 (11), S. 152–164. DOI: 10.1177/0361198120938054.

Schmillen, Achim; Stüber, Heiko (2014): Lebensverdienste nach Qualifikation: Bildung lohnt sich ein Leben lang. In: IAB-Kurzbericht (1), S. 1–8.

Vargo, Stephen L.; Lusch, Robert F. (2004): Evolving to a New Dominant Logic for Marketing. In: Journal of Marketing 68 (January), S. 1–17.