6. An Evaluation Concept for ETMs

Cp. Section 6.1.2.1.

Cp. Section 6.3.

Cp. Heagney (2016), pp. 1–214; Project Management Institute (2017), pp. 1–536.

Cp. Abts and Mülder (2017), pp. 515–558; Krcmar (2015), pp. 211–228; Pilorget (2015), pp. 107–128.

In a broad definition, the needs assessment deals with “[…] understanding business goals and objectives, issues, and opportunities and recommending proposals to address them” ( Project Management Institute (2017), p. 30). Besides setting objectives, the needs assessment should involve an analysis of requirements which is needed for the selection of software in general (cp. Abts and Mülder (2017), p. 526) and the selection of EMs (cp. Rickes and Scherenschlich (2018), p. 14).

Cp. Section 6.3.1 for a more detailed explanation of the implementation activities.

Not all the objectives broadly set in the needs assessment and specified in more detail in the business case might be realized as planned. Therefore, the achievement of the objectives should be critically reviewed after the implementation. In addition, the experiences from the implementation should be used for the future, e.g. by asking what went well and where improvements are wanted for the next time (cp. Heagney (2016), p. 14).

Source: own representation. This figure has been designed using resources from flaticon.com (cp. https://​www.​flaticon.​com).

Cross-functional teams allow the realization synergies because employees with different skills work jointly on achieving the objectives of a project (cp. Kalabina et al. (2021), p. 183).

A combination of managerial/logistics know-how and IT expertise is generally recommended for the evaluation of IT-based logistics solutions (cp. Hausladen (2020), p. 328).

Employees who benefit from, contribute to or will be impacted by a project are called stakeholders (cp. Heagney (2016), p. 48). For example, the accounting department is involved in the payment of the transportation services (cp. Section 2.​3.​3) and may be impacted by the ETM implementation if it affects payment processes.

The last two steps of the procedure for the evaluation of ETMs (implementation and review) will not be part of the evaluation concept. There are three main reasons for this decision. First, the research questions and objectives of this thesis are mainly related to the first three steps. Second, the implementation and review activities are highly dependent on the outcomes of the previous steps. By covering the first three steps in detail, the needed foundation for the two last steps is built. Finally, the implementation and review activities of ETMs should be quite similar to those of other IT systems. Thus, there should be a lower need to support the activities during the implementation and review steps.

Cp. Heagney (2016), p. 59.

The use of ETMs may help to reduce deadheads (cp. McKinnon et al. (2015), pp. 325–326).

It is very difficult for an ETM to identify those contract or spot market exchanges on an ETM which actually avoid deadheads. For example, when a shipper fills the truck of a transportation service provider with a transport order on the spot market it is not possible to determine with certainty whether this exchange actually avoids a deadhead because it may still be possible to fill the truck with another transport order.

Source: own representation.

Cp. Section 6.3.1 for an overview of ETM costs.

Cp. Section 4.​4.​4.

Source: own representation.

The explanations of the ETM functionalities in this work provide a good starting point for these discussions (cp. Section 3.​5). However, the range of functionalities is constantly increasing. Therefore, a first basic market research can be helpful to be up to date with the latest developments.

Source: own representation.

Cp. Section 2.​1.​2.​1.

Internal customers are relevant for internal goods movements (e.g. movements between plants of a manufacturing company). Furthermore, shippers can serve different customer groups which share a specific set of needs and expectations, e.g. hospitals or global accounts (cp. Donovan and Samler (1994), p. 80).

Some products require specific vehicles or adherence to specific standards or procedures, e.g. reefer cargo (cp. Section 2.​2.​2). Furthermore, different shipment sizes can be relevant (cp. Section 3.​3.​2).

Cp. Section 2.​2.​2 for an overview of logistics units relevant for road transportation. Products can also be aggregated to product groups (cp. Kilger and Wagner (2015), pp. 129–130), which can be a basis for the distinction of different transportation services.

Cp. Section 2.​1.​2.​2. The transportation relations can be analyzed on different aggregation levels (e.g., detailed delivery addresses or on a higher aggregation level of two-digit postal codes). Furthermore, they could be grouped according to their geographical scope, e.g. local, regional, continental (cp. Section 3.​3.​2).

Asset specificity is associated with the availability of few or no suppliers (cp. Papenhoff (2009), p. 93; Picot and Dietl (1990), p. 179) and is thereby strongly related to the contextual variable availability of alternatives. Hence, both contextual variables have been combined under “availability of transportation service providers” to reduce complexity for the evaluation.

Source: own representation based on the conceptual research framework (cp. Sections 4.​3.​2.​2.​1 and 4.3.2.2.4) and the preliminary study (cp. Section 5.​1.​3.​2).

The average demand interval reflects the degree to which demand is regularly occurring and the coefficient of variation measures the variation in demand quantities. For the calculation of these variables, cp. Sankaran et al. (2019), pp. 180–190. For example, services have a low demand uncertainty when the squared CV is below 0.28 and the ADI does not exceed 1.34 (smooth demand). The demand is uncertain if it can be classified as erratic, lumpy, or intermittent according to the CV and ADI (cp. Boylan and Syntetos (2008)).

Source: own representation.

Clearly, a competitive market only exists when there are sufficient suppliers. For example, electronic reverse auctions are preferred when there are many competing suppliers (cp. Schoenherr and Mabert (2011); Smeltzer and Carr (2003)) and when the needed supplies involve low levels of asset specificity (cp. Jap and Haruvy (2008); Stoll (2008), p. 239). The empirical study of this thesis also provides weak indications that asset specificity has a negative effect on the intended depth of ETM use (cp. results of H18 in Section 5.​2.​6.​2.​2).

The spot market is suitable for transportation services that are difficult to plan, whereas the contract market is preferred for well-predictable or frequently recurring services (cp. reasoning for H16a in Section 5.​2.​1.​2.​1). This is also evident in the results of the preliminary study (cp. Section 5.​1.​3.​2) and the results of the main study (cp. results of H16a in Section 5.​2.​6.​2.​3).

The ease of service description depends on the complexity of the service description and the degree of specifiability. It thereby combines complexity and specifiability as contextual variables to facilitate the evaluation of transportation services by reducing the number of variables that must be assessed.

Cp. Lai and Cheng (2016), p. 69.

Source: own representation based on the conceptual research framework (cp. Sections 4.​3.​2.​2.​2 and 4.​3.​2.​2.​6) and the preliminary study (cp. Section 5.​1.​3.​2).

Source: own representation based on the conceptual research framework (cp. Section 4.​3.​2.​2.​3) and the preliminary study (cp. Section 5.​1.​3.​2).

Source: own representation.

Source: own representation.

The spot market is suitable for transportation services that are difficult to plan, whereas the contract market is preferred for well-predictable or frequently recurring services (cp. reasoning for H16a in Section 5.​2.​1.​2.​1). This is also evident in the results of the preliminary study (cp. Section 5.​1.​3.​2) and the results of the main study (cp. results of H16a in Section 5.​2.​6.​2.​3).

E-tendering is used to identify and select the right transportation service providers for a contract that covers multiple transportation needs over a longer time. The functionality electronic transport order can be used to transfer multiple transport orders which are created for this contract (cp. Section 2.​3.​3.​2).

Cp. Section 4.​2.​2.​4.

Complexity increases transaction costs, ambiguity, and risks of transactions. Therefore, it is likely that shippers seek close relationships with transportation service providers for transportation services that involve a high complexity (cp. reasoning for H17b in Section 5.​2.​1.​2.​1). Furthermore, discrete exchanges (e.g., via electronic reverse auctions) are possible when the needed supply can be clearly specified (cp. Hawkins et al. (2009); Hawkins et al. (2010); Kaufmann and Carter (2004)), but relational exchanges are preferred when specifiability is low (cp. Radkevitch (2008), p. 93; Levi et al. (2003)). The results of the preliminary study also indicate that the use of a restricted supplier base is beneficial when the conditions and requirements must be defined in detail before the transport (cp. Section 5.​1.​3.​1.​2).

The importance of a supply may affect a company’s dependency on suppliers. It has been found that buyers use relational exchanges in a purchasing situation for important supplies (cp. Cannon and Perreault Jr (1999); Janz (2004), p. 299; Metcalf and Frear (1993); Werner (1997), p. 161). The results of the empirical study of this work also show that the relational orientation toward transportation service providers increases with the importance of the services (cp. results of H3b in Section 5.​2.​6.​1.​2). Furthermore, the experts of the preliminary study perceive less risks to delivery quality when a restricted supplier base is used (cp. Section 5.​1.​3.​1.​2) which may lead to a preference for the use of a restricted supplier base for important transportation services.

Source: own representation.

Cp. Abts and Mülder (2017), p. 528; Held (2003), p. 218; Brenner (1994), p. 99.

Source: own representation.

Market research can be conducted based on information from the internet, the press, trade fairs, or discussions with other companies (cp. Abts and Mülder (2017), p. 529). Furthermore, market studies provided by research institutes or consulting companies can be used (cp. Hausladen (2020), p. 323).

Cp. Abts and Mülder (2017), p. 513; Krcmar (2015), p. 214.

The required functionalities can be the main or additional functionalities of an ETM.

For example, Courier.net (cp. https://​www.​courier.​net) or Packlink.com (cp. https://​www.​packlink.​com) support transactions for CEP shipments.

There are ETMs that support transportation services for reefer cargo (e.g., https://​coolload.​com) or bulk goods in silo trucks (e.g. https://​siloladungsboers​e.​com).

For example, Transporeon (cp. https://​www.​transporeon.​com) and Timocom (cp. https://​www.​timocom.​de) offer access to transportation service providers which can move goods with standardized or more specialized vehicles.

Cp. Manke and Funder (2017), p. 13.

Cp. Sänger (2004), p. 98.

Since each further alternative must be evaluated, the efforts of the evaluation depend on the number of alternatives that are selected for the detailed evaluation. A good number of alternatives for a more detailed evaluation typically consists of 3 to 6 options.

“Scoring model” is the translation of the German term “Nutzwertanalyse” (cp. Hartel (2015), p. 83; Riedl (2006), p. 111). Alternatively, the term “utility analysis” could be used for the translation of the term (cp. Schulze (2009), p. 166.). The scoring model is a methodology for multi-criteria decision-making (cp. Schleich (2018), p. 43). It has been selected because it is a methodology that is well accepted for solving problems in logistics and supply chain management (cp. Hausladen (2020), p. 330; Hartel (2015), p. 83; Schulze (2009), p. 166). Furthermore, scoring models are considered suitable for selecting software (cp. Feyhl (2004), pp. 94–95) or EMs (cp. Held (2003), p. 221). The procedure is also very popular in corporate practice (cp. Schneeweiß (1991), p. 120). An important reason for this popularity may be that in comparison to other methods for multi-criteria decision-making (e.g. analytic hierarchy process), scoring models are easier to understand (cp. Riedl (2006), pp. 117–118). One of the inherent weaknesses of the scoring model is the subjective weighting which might be (consciously or unconsciously) guided by the interests of participants (cp. Geier (1999), p. 143; Götze (2014), p. 199; Kühnapfel (2019), p. 13). However, there are ways for reducing this weakness. For example, the method of pair-wise comparison could be applied to elicit the weights (cp. Kühnapfel (2019), pp. 13–16 Quah and Toh (2012), pp. 77–83).

These steps are derived from recommendations that have been provided in the literature (cp. Feyhl (2004), pp. 94–95; Kühnapfel (2019), pp. 5–25; Riedl (2006), pp. 111–115).

The evaluation criteria should be defined in a group with 3 to 20 members, whereby a group size between 5 and 10 people can be considered optimal (cp. Kühnapfel (2019), p. 5).

Cp. Sections 4.​3.​3.​2.​1 to 4.​3.​3.​2.​6.

Source: own representation.

Cp. Boyer and Olson (2002), pp. 490–493; Chien et al. (2012), pp. 465–466.

Cp. Adomavicius et al. (2013), p. 497; Brandon-Jones and Kauppi (2018), p. 34; Chakraborty et al. (2002), pp. 64–68; Lee and Lin (2005), pp. 169–171; Lin (2007), pp. 371–373; Kang (2014), pp. 457–460; Kollmann (2001a); Zahedi et al. (2010), pp. 310–312.

For example, employees may work against an IS which they do not accept (cp. Brown et al. (2002), p. 284) or the obligatory use of an unaccepted system leads to lack of job satisfaction which is associated with absenteeism or higher employee turnover (cp. Sørebø and Eikebrokk (2008), p. 2358).

The voluntary use extension by employees, for example, can improve outcomes for firms (cp. Hsieh et al. (2011), pp. 2027–2031.

Several studies indicate that information quality is positively associated with the acceptance of e-procurement systems (cp. Chakraborty et al. (2002), pp. 64–68; Jones and Kim (2010), pp. 632–633; Kassim and Hussin (2013), pp. 10–13; Ramkumar et al. (2019), p. 340).

There are a few indications for a positive effect of service quality on the acceptance of e-procurement systems (cp. Brandon‐Jones and Carey (2011), pp. 283–286; Devaraj et al. (2002), pp. 324–327; Lee and Lin (2005), pp. 169–171; Zhou et al. (2009), p. 331.

Clearly, the potential to increase supply assurance and flexibility for the procurement of transportation services depends on the number of transportation service providers which can be reached on the ETM. Furthermore, a larger number of transportation service providers on an ETM should also positively affect freight rates. It has been found, for example, that competition among suppliers has a positive effect on the success of electronic reverse auctions (cp. Wagner and Schwab (2004), p. 21).

For example, ETMs can implement qualification systems (cp. Section 4.​3.​3.​2.​5).

Source: own representation.

When categories are used, the start should be made by assigning weights to these first. Afterward, weights should be assigned to the criteria within the categories. The weight of an individual criterion is determined by multiplying the weight of the category with the weight of the criterion within the category. For example, when the weight of a category is 40% and the weight of the criterion within the category is 10%, the weight of the criterion is 4% (cp. Kühnapfel (2019), pp. 12–14).

In vendor workshops, the ETM operators can present the functionalities and possibilities of their ETM. It is generally a good idea to prepare such workshops and presentations with a demo script which is sent to the ETM operators beforehand (cp. Groß and Pfennig (2017), pp. 277–287).

Reference customers are companies that already use the specific IT system which should be evaluated (cp. Abts and Mülder (2017), p. 529). Typically, the ETM operators bring the reference customer and the shipper together.

Tests provide very good insights into the processes and functionalities supported by ETMs.

Cp. Kühnapfel (2019), p. 17.

Source: own representation.

More than one ETM can be needed, for example, when one ETM alone is not appropriate for purchasing all the different types of transportation services of a shipper. In such cases, it is a good idea to present in a transparent way to which degree each of the ETMs is appropriate for a certain type of transportation service. For example, this can be done by consolidating the types of transportation services and the suitability of different ETMs for purchasing them in a matrix (cp. Held (2003), p. 229; Kersten and Held (2001), p. 48).

Fee levels could also be represented as criteria that receive score values. However, this involves the transformation of precise monetary values into more abstract score values.

Source: own representation adapted from Riedl (2006), p. 115.

Cp. Abts and Mülder (2017), pp. 545–547; Wieczorrek and Mertens (2011), p. 275.

Cp. Brugger (2009), p. 13.

Source: own representation.

In general, hardware, software, personnel, and other costs represent different types of IT costs (cp. Potthof (1998), p. 7). ETM fees usually cover the software costs as well as costs for hardware (e.g., a computer server) and other costs (e.g., costs for renting space required for hardware) that an ETM operator incurs in course of the provision of its services. Personnel costs can occur for different activities during the implementation (e.g., training) and productive use of an ETM (e.g., support).

Cp. Gehra (2005), p. 176; Potthof (1998), p. 7; Schumann (1992), p. 66.

Source: own representation.

Cp. Sänger (2004), p. 82; Schwind et al. (2011), p. 3; Stockdale and Standing (2002), p. 230.

Activities for the system setup include all activities to prepare the ETM and connected IT systems for productive use. This can include, for example, configuration, programming, establishing, or changing system interfaces or migration of data.

The activities for enabling end-users comprise user-related tasks that are needed so that end-users can effectively work with the ETM (e.g., training, documentation, or granting authorizations).

Once the needed system setup and enabling of end-sers has been completed, it can be tested whether the ETM is ready for productive use (go-live). Typically, higher efforts to support users and make short-term corrections are needed shortly after the go-live.

Elimination refers to the possibility that unnecessary or non-value-adding process steps are eliminated. Automation means that manual activities are replaced by computers and software. The acceleration of processes means that processes can be performed faster (cp. Best and Weth (2009), pp. 124–157; Krickl (1994), pp. 27–29).

Source: own representation based on Best and Weth (2009), pp. 124–157 and Krickl (1994), p. 28.

Activity-based costing is an instrument for business controlling. The beginning of activity-based costing has been made in the research community in the USA (cp. Miller and Vollmann (1985); Cooper and Kaplan (1988)). In the German research community, the instrument termed “Prozesskostenrechnung” which is based on activity-based costing has been introduced in the late 1980 s (cp. Horváth and Mayer (1989)).

Cost drivers are all factors that cause a change in the cost of an activity. They are used to identify the root cause of the work and cost of an activity (cp. Emblemsvåg and Bras (2001), p. XVIII).

The process costs can be calculated based on the needed time for conducting the process for a single cost driver. For example, when a process without an ETM requires 10 minutes of an employee’s time and a fictitious rate of 90€ per working hour is used, the process costs per cost driver amount to 15€. When it is estimated that the time per cost driver can be reduced to 4 minutes with the help of an ETM, the costs per cost driver are reduced by 9€ to 6€.

When an ETM is used to support multiple business processes, the estimated process cost for all processes must be summed up to calculate the total process cost reductions.

Source: own representation.

Only these ETM main functionalities support the information and negotiation phases and can therefore reduce freight rates. In addition, the empirical study has confirmed that a higher depth of ETM use in terms of these functionalities is also associated with higher expected freight rate reductions (cp. results for H26 Section 5.​2.​6.​2.​2).

13 out of 22 respondents of companies using an ETM replied that abolishing the use of an ETM would increase freight rates (cp. Section 5.​2.​4.​2.​2).

For example, if a purchasing volume of 1,000,000€ can be conducted via spot market functionalities or e-tendering on an ETM and the use of the ETM is expected to lead to 5% lower freight rates in comparison to current ones, the total freight rate reductions amount to 50,000€.

Cp. Brugger (2009), pp. 340–348; Potthof (1998), p. 29.

Cp. results of H23a in Section 5.​2.​6.​2.​3.

The definition is based on a definition of the service level in logistics (cp. Schönherr (2015), p. 169; Reichmann (2011), p. 379).

The KPI is derived from a definition of delivery flexibility (cp. Disselkamp and Schüller (2004), p. 138). In contrast to this general definition of delivery flexibility, the flexibility defined within this thesis focuses only on changes related to transport orders.

Cp. Section 4.​4.​3.​3.​2 and Section 5.​1.​3.​1.​2.

Cp. Section 4.​4.​4.

This KPI is adapted from a definition of delivery reliability (cp. Schneider and Hennig (2008), p. 216). Quality issues of transport orders can occur in terms of timeliness, accuracy, and condition (cp. Section 4.​4.​3.​3.​2). It should be kept in mind that transportation service providers are not responsible for all problems with delivery quality (e.g., if a delay is caused by long waiting times in the plant of a shipper).

No existing possibilities for measuring information capability have been found in the literature (cp. Böttcher (1993), p. 229; Kaminski (2003), p. 123; Rennemann (2007), pp. 185–186; Seeck (2010), p. 10). Therefore, the KPI has been self-developed.