Standardizing social pickup points to mitigate risk perception and improve
consumer adoption
Master thesis
Technology & Operations Management University of Groningen
Faculty of Economics and Business
Supervisor: prof. dr. K.J. Roodbergen Co-assessor: dr. N. B. Szirbik
January 27, 2020
Jimmy Smith S3000494
p.j.smith@student.rug.nl
Abstract
Purpose – This research studies the influence of perceived risk and risk relievers on consumer adoption of social pickup points. Next, the appropriate boundaries of social pickup point design and process standardization are investigated by researching which risk relievers are relevant and feasible to implement. The overarching goal is to investigate social pickup points as a last-mile solution.
Design/methodology/approach – Both a questionnaire and a multiple case study have been conducted.
The questionnaire quantitatively measures consumer perceived risk and effectiveness of risk relievers on social pickup points. The multiple case study qualitatively studies pickup point firms perspectives on both managerial and operational level. Consequently, the views of consumers, managers, and operators can be compared and set in context with current literature.
Findings – This study prioritized risk dimensions and corresponding risk relievers for social pickup points.
Performance, privacy, time, and physical risk are considered most important, followed by financial risk, and image risk is least important. Another finding is the mismatch between what carriers and consumers expect from social pickup points with what the social pickup point firms can offer in the current business format. Modular standardization that can adapt to different consumer needs can help to resolve the mismatch.
Research implications/limitations – Future research on social pickup points should focus on the social aspect because that is the most significant difference from instore pickup points. For future research, it must be taken into account that right now only a tiny fraction of the parcels are processed by social pickup points, and firms indicated it can only serve as an extra layer to the instore pickup point network.
Originality/value/contribution – This study proposes design standardizations for social pickup points that can improve consumer adoption, and make it a more interesting last-mile solution.
Keywords: Social pickup points; Perceived risk; Risk relievers; Design standardization
Acknowledgment – First, I would like to thank my thesis supervisor prof. dr. K.J. Roodbergen for guiding
me during the complete process. Also, I would like to express my gratitude to the managers and operators
at pickup point firms that I interviewed for their time and input, even in the busiest time of the year. Lastly,
I would like to thank all the people that conducted the questionnaire, and especially the people that
distributed the questionnaire.
Table of contents
1. Introduction... 5
2. Literature research ... 8
2.1 Social pickup points ... 8
2.2 Consumer adoption ... 9
2.3 Perceived risk ... 10
2.4 Risk relievers ... 11
2.5 Standardizing design ... 13
3. Methodology ... 15
3.1 Questionnaire ... 15
3.2 Multiple case study ... 16
3.3 Data collection ... 16
3.4 Data analysis ... 18
3.5 Coding tree ... 18
4. Findings ... 21
4.1 Questionnaire findings... 21
4.2 Interview findings ... 24
5. Discussion ... 30
5.1 Risk dimension prioritization ... 30
5.2 Shared responsibility ... 31
5.3 Modular standardization ... 33
6. Conclusion ... 35
7. References ... 37
Appendix ... 43
A. Questionnaire results ... 43
B. Interview guidelines ... 49
1. Introduction
The rise of pickup points in the Netherlands has been researched by Weltevreden (2008), showing in 2006 there were approximately 900 locations that processed only 1,4% of all online orders. Nowadays 9% of the Dutch consumers prefer collection from a manned pickup point (Postnord, 2018). The pickup point market has become a mature market with carriers PostNL, DHL, UPS, DPD, and GLS that together have approximately 8500 instore pickup points (ACM, 2018). However, sometimes the instore pickup point capacity is not sufficient to meet the demand (Corré, 2019), and it is interesting to consider alternatives to increase pickup point capacity. In 2016 a new type of pickup point has been introduced by startups ViaTim and Homerr (Dijkhuizen, 2017). These are so-called social pickup points that are located at individuals’ homes and open outside office hours. Since it is a new concept, it is not yet clear how social pickup points should be designed and what their role will be in the last-mile (Weerd, 2019). Therefore it is interesting to investigate the social pickup point design and its opportunities.
Research from the Amsterdam University of Applied Sciences on city logistics suggests that standardization and reliable processes are necessary for further development of logistics hubs (HvA, 2019). The social pickup point concept was one of the hub types they researched, and therefore social pickup point standardization and reliability of the process are interesting topics to research. Brouwer (2019) conducted explorative research on social pickup points and suggested that consumer perceived risk possibly influences their choice, leading to not adopting the new concept. Perceived risk is the degree to which individuals believe that if they purchase products in a specific way, they will suffer losses (Lim, 2003). To improve consumer adoption, the usage rate of a new concept, it is necessary to mitigate consumer perceived risk on social pickup points. Risk relievers are design and process measurements that can influence consumer choice and stimulate a specific decision (Kunze & Mai, 2007). When certain risk relievers turn out to stimulate consumer adoption of social pickup points, it is beneficial to add them to the standard design. Consequently, it is interesting to research to which boundaries it is beneficial to standardize social pickup point design and processes. Currently the concept of social pickup points is in its growth phase (Dijkhuizen, 2017), and therefore there is still room to adjust the social pickup point concept by applying design and process standardization.
Literature on last-mile solutions is very extensive with research on many different options. Multiple
solutions still focus on front door delivery, such as delivery in the evening, shorter delivery windows, and
a personal reception box (McKinnon & Tallam, 2003). However, the current paper focuses on pickup
points, especially on social pickup points, on which Brouwer (2019) and Wissink (2019) conducted explorative research, and the current paper extends this new field of literature.
In this research, we analyze the influence of perceived risk on social pickup point adoption. Additionally, it researches which risk relievers are effective to mitigate perceived risk, and which of them should be standardized in the social pickup point design and process. Therefore the following research question is addressed:
What risk relievers are effective in mitigating consumer perceived risk and should be added to the standardized social pickup point design and process?
To answer this research question the following sub-questions are addressed:
- What are the sources of perceived risk on social pickup points and how important are they?
- Which risk relievers are effective to decrease perceived risk on social pickup points?
- What are appropriate boundaries for social pickup point standardization?
A combination of a questionnaire and a multiple case study is used to answer these research questions.
The questionnaire focuses on the first two subquestions by quantitatively researching the presence of perceived risk and the effectiveness of risk relievers among potential social pickup point users. The multiple case study mainly answers the third subquestion by researching the relevance and feasibility of standardizing risk relievers. For the multiple case study, interviews are conducted at both social and instore pickup point firms on a managerial and operational level.
Academically this paper mainly builds upon papers in three different categories. First, it contributes to last-
mile solution literature by investigating the construct perceived risk that influences adoption. Similar
research is conducted by Zhou et al. (2020), who research the adoption of the last mile solution lockers,
by applying an extended UTAUT model of which perceived risk is a construct. Second, it contributes to
general pickup point literature. The paper of Xu et al. (2011) standardizes the design of pickup points, and
the current paper extends that by suggesting social pickup point standardizations. Third, this paper
generalizes and extends explorative single case studies on social pickup points by Brouwer (2019) and
Wissink (2019) with more generalizable results from a multiple case study and quantitative results from a
questionnaire. The added business value of this paper are standardization recommendations that
managers can use to formulate their social pickup point design.
The remainder of this paper has the following structure: Section 2 provides an overview of current literature, including general information on the social pickup point concept, evaluation of the different perceived risk dimensions with corresponding risk relievers, and franchising and standardization literature.
Section 3 describes the methodology of the research, and Section 4 presents the findings. Section 5
discusses the findings, and Section 6 presents the conclusions of this research.
2. Literature research
The new concept social pickup points has been briefly introduced in Section 1, and Section 2.1 describes the concept more extensively. Since the current paper researches social pickup point adoption, the adoption model of Zhou et al. (2020) on last-mile solution lockers is described, which includes the construct perceived risk. Subsequently, perceived risk is further investigated, resulting in six different perceived risk dimensions: financial, performance, physical, time, image, and privacy risk. After that, possible risk relievers are described per risk dimension. Lastly, standardizing and franchising literature is described to find appropriate standardization boundaries for social pickup points. The literature from this section serves as a base for the questionnaire and interview guidelines.
2.1 Social pickup points
First, the social pickup point aspects that this section elaborates on are briefly introduced. The distinctive characteristics of social pickup points as a last-mile solution will be described. Therefore the not-at-home problem is introduced, which is a significant problem for front door delivery. Pickup points are a possible solution for the not-at-home problem, and this section describes the differences between social and instore pickup points.
Homes are being empty for longer periods a day due to inflexible working patterns, long commutes, an increase in working women, and an increase in single-person households resulting in a relatively high proportion of first-time delivery failure (Park & Regan, 2004). This ‘not-at-home’ problem causes disadvantages to the carriers with higher operating costs and inconveniences to consumers leading to lower satisfaction (Ferrand et al., 2008; Lee & Whang, 2001). Unattended delivery turns out to be a good last-mile solution to lower carrier operating costs and simultaneously satisfy consumers that are unwilling to wait for hours for a delivery to arrive (Ferrand et al., 2008).
Pickup points are considered an unattended delivery method since the final consumer does not need to
be present at the delivery and can pick it up later (Zenezini et al., 2018). Instead of the final delivery to the
front door of the customer, the parcel ships to a preselected pickup point that receives the parcel and
stores it until the customer picks it up. Benefits of pickup points are the increase in delivery quantity per
hour due to reduced road traffic, delivery is more likely to be attended by the customer, and the routing
planning becomes less complex (Ferrand et al., 2008; Zenezini et al., 2018). Pickup points can be divided
into manned and unmanned pickup points, of which the focus will be on the former. Unmanned pickup
points do not involve employees, and the most common formats are lockers in public space and a reception box at home (Halbauer, 2018).
Manned pickup points can be divided into instore pickup points and social pickup points. Instore pickup points are longer in business than social pickup points, and therefore more literature is available and the concept is defined more extensively. Instore pickup points are located in post offices, petrol stations, and any other retail shop (Ferrand et al., 2008). Main concerns of instore pickup points as a last-mile solution are the distance between the instore pickup point and the customer’s home, and the time available for collection (Ferrand et al., 2008). With social pickup points, the parcel is delivered at individuals that open up their homes in certain hours, accepting parcels for the neighborhood (Paazl, 2019). Social pickup points have the potential to solve the distance and time problems of instore pickup points since they can potentially startup close to everybody who has neighbors and the opening times are outside office hours (Wissink, 2019). Next to that, social pickup points can be beneficial because of two social aspects. First, it can be environmentally friendly (Brouwer, 2019). Second, it can improve the social life of both the operators and consumers of social pickup points (Wissink, 2019).
This concept is firstly introduced in the Netherlands, and currently the main actors in this business are Homerr and ViaTim, with partnerships with carriers DHL and DPD. Although the potential benefits of new concepts are often enticing, it cannot be realized until the consumers accept and embrace the new concept (Wang et al., 2018), and therefore the next section describes adoption theory of a last-mile solution.
2.2 Consumer adoption
This section discusses the paper of Zhou et al. (2020) that researches the adoption of last-mile solution
lockers, using an extended UTAUT model that includes the construct perceived risk. A UTAUT model is a
technology acceptance model that aims to explain user intentions to use a new concept (Venkatesh et al.,
2003). Technology acceptance models can be useful to influence usage behavior. E-consumers are more
aware of risk (Tandon et al., 2018), and therefore the construct perceived risk is adopted in the conceptual
model in Figure 2.1. The paper of Zhou et al. (2020) researches the negative effect of perceived risk on
behavioral intention and usage behavior of lockers.
Figure 2.1: Extended UTAUT conceptual model to research self-service parcel services adoption (Zhou et al., 2020)
All relationships in the conceptual model of Zhou et al. (2020) were empirically validated; thus only perceived risk has a negative effect on behavioral intention and usage behavior. Therefore perceived risk can be a relevant research topic for other last-mile solutions since risk mitigation can lead to an increase in behavioral intention and usage behavior.
2.3 Perceived risk
In this section the concept of perceived risk will be described. To provide a better, more detailed understanding it is divided into six dimensions: financial, performance, physical, time, image, and privacy risk. After that, the relationship between sources and consequences of perceived risk is described.
The construct perceived risk is considered relevant because Tandon (2018) emphasized the importance of perceived risk among e-consumers, and Brouwer (2019) suggested to improve social pickup point requirements to mitigate perceived risk. In the current paper perceived risk is defined as the degree to which individuals believe that if they purchase products in a specific way, they will suffer losses (Lim, 2003;
Zhou et al., 2020).
In the current paper, perceived risk is divided into six dimensions by recombining the definitions of Lim
(2003) and Zhou et al. (2020). Financial risk is defined as the potential monetary loss associated with the
payment price of the service (Lim, 2003; Zhou et al., 2020). Performance risk is defined as the possibility
that the service cannot achieve the expected performance (Lim, 2003; Zhou et al., 2020). Physical risk is
defined as the chance that the service is not safe and might be harmful or injurious to your health (Lim,
2003). Time risk is defined as the perceived risk of wasted time due to disutility of the service or waiting
time at the social pickup point (Lim, 2003; Zhou et al., 2020). Image risk is defined as the potential loss on your self-image because of using the service and a combination of psychological and social risk (Lim, 2003;
Zhou et al., 2020). Privacy risk is defined as the probability of having personal information disclosed as a result of the service (Lim, 2003).
Lim (2003) indicates that only identifying these perceived risk types is not useful since it does not provide clear guidance on which steps to take to reduce consumers’ perceived risk. The perceived risk dimensions represent ‘consequences’ referring to a type of loss consumers perceive to suffer as a result of their actions. These results fail to show the sources of such perceived risk, while identifying the sources of perceived risk is useful to target their resources in the right places. Therefore Lim (2003) identifies four sources of perceived risk in relation to online shopping: technology, vendor, consumer, and product. The relationships of perceived risk sources and consequences are presented in Figure 2.2 and can be useful to develop effective measurements.
Figure 2.2: Matching of sources and consequences of perceived risk (Lim, 2003)
To change design, aiming to mitigate the negative effects of perceived risk on behavioral intention, the effect of perceived risk sources needs to be reduced. Risk relievers are measurements to reduce perceived risk and the topic of interest in the next section.
2.4 Risk relievers
This section first describes risk relievers and their role in mitigating perceived risk to stimulate adoption of a product or service. Next, a list of risk relievers is constructed based on the perceived risk dimensions from Section 2.3.
The level of perceived risk is built up from the ‘amount of uncertainty’ and the ‘consequence of purchase’
(Kunze et al., 2007). To reduce the perceived risk level, there are probable actions called ‘risk relievers’
that can be implemented to stimulate a specific decision (Kunze et al., 2007). Risk relievers are instruments employed by suppliers to mitigate perceived risk and thus stimulate product or service adoption (Roselius, 1971). Firms face a trade-off between the benefits of risk relievers that lead to higher sales volume by mitigating perceived risk that caused hesitancy to buy, and the costs of offering these risk relievers. Firms need to decide which risk relievers can be most effective for relieving the specific perceived risk dimension that the consumers experience with the service (Roselius, 1971). Risk relievers can be applied on strategy and design changes, mitigating consumer perceived risk, consequently leading to improved adoption of the technology.
In Table 2.1 risk relievers are categorized according to the perceived risk dimensions from Section 2.3. In the third column a definition of the risk reliever is given. The risk relievers are obtained by literature research on risk relievers in e-commerce and risk-reducing strategies for last-mile solutions.
Risk type Risk reliever Definition References
Financial Free sample Use a free sample of the product on a trial basis before buying.
(Roselius, 1971; Tan, 1999)
Money-back guarantee
Buy whichever brand that offers a money-back guarantee.
(Derbaix, 1983; Roselius, 1971; Tan, 1999)
Performance Major brand image
Buy a major, well-known brand of the service and rely on the reputation of the brand.
(Derbaix, 1983; Roselius, 1971; Tan, 1999)
Store image Buy the brand that is carried by a store that you think is dependable and rely on the reputation of the store.
(Derbaix, 1983; Roselius, 1971)
Brand loyalty Buy the brand you have used before and have been satisfied with in the past.
(Cases, 2002; Derbaix, 1983; Roselius, 1971)
Reviews Read online reviews to obtain more information on the service and its performance
(Racherla & Friske, 2012;
Zhang, 2018)
Word of mouth Ask friends and family for advice about the service.
(Cases, 2002; Derbaix, 1983; Roselius, 1971;
Tan, 1999) Speak to a
salesperson
The possibility to contact a
salesperson by phone or the internet.
(Cases, 2002)
Product in advance
Seeing the product in advance in real life.
(Cases, 2002;
Korgaonkar, 1982)
Process
documentation tool
The presence of a tool that documents the service process at the pickup point.
(Xu et al., 2011)
Physical Lighting The presence of proper lighting facilities in and around the pickup point.
(Kedia et al., 2017)
Monitored area The presence of video surveillance in and around the pickup point.
(Halbauer, 2018; Kedia et al., 2017; Kelli De Oliveira et al., 2019) Openness of
space
The openness of space at the rooms of the pickup point.
(Kedia et al., 2017)
Container box The presence of a container box to store the parcels.
(Xu et al., 2011)
Time Time ambiguity The uniformity in opening times. (Brouwer, 2019) Image Environmental
impact
Show environmental benefits to stimulate green options.
(Abid & Latif, 2015)
Sense of community
Show the opportunity to meet new people and make new friends.
(Möhlmann, 2015;
Tussyadiah, 2016) Privacy Privacy disclosure Presence of a privacy disclosure in the
store or webshop.
(Featherman & Pavlou, 2003; Pan & Zinkhan, 2006)
Table 2.1: Risk relievers, their definition, and reference for each risk dimension
When risk relievers turn out to be effective and improve adoption behavior it is beneficial to implement the risk reliever in the product or service. Then it is useful to standardize it into the concept and standardization literature is described in the next section.
2.5 Standardizing design
When results turn out a specific risk reliever is effective and mitigates perceived risk, it can be beneficial to add it to the standard design of the concept. This section discusses an important paper on standardization and franchising by Kaufmann and Eroglu (1999) to study the appropriate boundaries of standardization. Four format components are introduced: service deliverables, benefit communicators, system identifiers, and format facilitators. The importance of the centrality of these format components determines whether they should be core elements whose standardization must be enforced or peripheral elements of which a trade-off should be made between standardization and local adaptation.
Standardization and adaptation literature in franchising focuses on generalizing a business format
(Kaufmann & Eroglu, 1999). The franchise concept is taken as an example for designing social pickup
points. Standardization of a design depends on the extent to which design changes can be enforced by the franchisor, and the willingness and ability of franchisees to comply with it. Franchisees can contribute to standardization by sharing their expertise in the local markets, and the franchisors need to consider to which extent let franchisees work according to their own way.
The essence of franchising is capitalizing on both the economies of scale associated with large systems and the benefits derived from small, localized operations (Kaufmann et al., 1999). Four format components are introduced to organize that. First, service deliverables reflect the unique factors of the format franchise and define its unique competitive niche. Second, benefit communicators imply the existence of intangible or unobservable benefits to the customer by making the intangible, for example, protection or prestige, tangible. Third, system identifiers are visual branding elements at a specific location that link customers to the overall brand. Fourth, format facilitators are the policies and procedures that enable efficiency for both the complete franchise and specific locations.
Not all elements of these four format components are equal in terms of their centrality, and a distinction between ‘core’ and ‘peripheral’ elements is made (Kaufmann et al., 1999). Core elements have to be standardized across all franchisees. Peripheral elements are optional, and system-wide benefits of standardization must be weighed against the benefits of local adaptation. Figure 2.3 presents a table to categorize elements based on their format components and corresponding centrality.
Figure 2.3: Examples of format component elements relative to centrality (Kaufmann et al., 1999)
3. Methodology
For this research first a questionnaire, and after that, a multiple case study has been conducted. This sequence was possible because sufficient literature, discussed in Section 2, was available to set up the questionnaire. Research method questionnaires is a useful research method to reach a big group of respondents and obtain a clear set of results (Denscombe, 2015) and is therefore selected to acquire consumer preferences quantitatively. Explorative research on social pickup points by Brouwer (2019) and Wissink (2019) were single case studies that conducted qualitative research using interviews on a small group of respondents only. The multiple case study has been selected because of its ability to acquire in- depth knowledge and increased generalizability of results compared to single case studies (Gustafsson, 2017). The questionnaire combined with the multiple case study remedies the generalizability issues of qualitative interviews, and the problem of generalizability of a single case study. The multiple case study focusses on design standardization preferences from both managerial and operational perspectives. The results from the questionnaire and multiple case study are compared to investigate similarities and differences in social pickup point standardization preferences and opportunities between consumers, managers, and operators.
3.1 Questionnaire
The goal of the questionnaire was to investigate the presence of the different perceived risk dimensions and the effectiveness of risk relievers among potential social pickup point users. Therefore the questionnaire was separated into six blocks, which all researched a separate risk dimension from Section 2.3. These blocks are further divided, respectively measuring the presence of risk, and the effectiveness of risk relievers from Section 2.4. Figure 3.1 displays the overall structure of the questionnaire.
Figure 3.1: Questionnaire outline
Method
To guarantee the correctness of the questionnaire it has been checked by a supervisor and a group of 5
respondents multiple times to preclude spelling and grammatical mistakes, unclear definitions, and other
factors that could cause mistakes and misinterpretations. Since the target group was Dutch consumers,
the questionnaire has been conducted in Dutch to prevent misinterpretations, and a language barrier to
participate.
3.2 Multiple case study
Whereas the goal of the questionnaire was to investigate the consumer perspective on social pickup points, the multiple case study aims to investigate social pickup points from the perspective of managers and operators. The goal of the interviews was twofold. The first goal was to evaluate the results from the questionnaire at social pickup point firms and check whether the risk relievers that turn out to be most important are in line with the needs of the firms and whether implementation is feasible. The second goal was to find appropriate boundaries for social pickup point design and process standardization.
To accomplish the first goal, interviews are conducted at both social and instore pickup point firms to discuss the importance of risk dimensions and corresponding risk relievers on a managerial level. To accomplish the second goal, interviews were held at managerial level and with the operators of social and instore pickup points to investigate their view on standardization. Table 3.2 presents the firms and people that have been interviewed.
3.3 Data collection Questionnaire
Since there is no literature yet on a specific social pickup point user group, it is assumed that all people that use e-commerce are potential users of social pickup points. To obtain a diversified group of respondents, five people with different social demographic characteristics have distributed the survey among their acquaintances. The questionnaire was conducted in December 2019. In Table 3.1 the social demographics of the distributors are listed.
Distributor Age Work situation Living situation
1 23 Student City
2 31 Fulltime City
3 58 Part-time Village
4 31 Fulltime Village
5 60 Part-time Countryside
Table 3.1: Social demographics of the questionnaire distributors
Interviews
In this section it is described who is interviewed at which firm to guarantee the accomplishment of the
goal to obtain multiple views. The multiple case study can generally be distinguished into social and instore
pickup point firms. The interviews were conducted in December 2019.
To obtain multiple views of carriers with instore pickup points, PostNL and DHL are interviewed. PostNL has instore pickup points only but conducted a social pickup point pilot in 2019. DHL holds its own instore pickup points and has social pickup points in cooperation with ViaTim and Homerr. The interviewed social pickup point firms are ViaTim and Homerr which are the only two companies in the Netherlands. Whereas ViaTim focuses on social pickup points only, Homerr focuses on both social and instore pickup points. To obtain multiple views at each company, at least two people on a managerial level have been interviewed, except at DHL.
To obtain a diverse view on operational level, instore pickup points of PostNL, DHL, UPS, DPD, and GLS are interviewed. Where possible, locations partnering with multiple carriers were interviewed to compare experiences. Social pickup points operators have been interviewed at both ViaTim and Homerr, and to diversify, interviews have been conducted in different regions. Table 3.2 presents the interviewed companies and the positions of the interviewees.
Company Position Name Abbreviation
Managerial level
ViaTim Founder
Marketing manager
Coordinator account managers
Michiel Verkerk Robby van Eekeren Pouwel Kingma
SPP1.1 SPP1.2 SPP1.3
Homerr Co-founder
On boarding manager
Mark-Jan Pieterse Stijn Hutschemaekers
SPP2.1 SPP2.2
PostNL Business development manager
Management trainee
Rens Roelvink Anouk Oude Vrielink
FC1.1 FC1.2
DHL Coordinator pickup points Pieter de Vree FC2
Operational level
ViaTim (DHL) Social pickup point operators SPPO1,2,3
Homerr (DPD) Social pickup point operators SPPO4
Ofisa (PostNL) Instore pickup point owner IPP1
Integra (PostNL) Instore pickup point operator IPP2
Benny (DHL) Instore pickup point owner IPP3
Korenbeurs (DPD) Instore pickup point operator IPP4
Vulpunt (GLS) Instore pickup point owner IPP5
Handyman (DHL, UPS) Instore pickup point owner IPP6
Copyright (DHL, DPD, UPS)
Instore pickup point owner IPP7
Table 3.2: Interviewed companies, the position of the interviewees, and their abbreviation
3.4 Data analysis Questionnaire
The Likert scale is a useful method to measure respondents' attitudes (Emerson, 2017), and the scale used in the questionnaire is displayed below.
Not important Slightly important Moderately
important Important Very important
○ ○ ○ ○ ○
1 2 3 4 5
This answer results in scores from 1 to 5, with corresponding meaning on top of it. Where necessary, the statistical significance is calculated by computing the P-value using a T-test. The fulfillment of the T-test depends on the available data. When the direction of the difference is already known from literature, a one-tailed distribution is used, and when it is unclear a two-tailed distribution. When significance is tested on different variables for the same sample it is paired. In other situations it is a two-sample with unequal variance (heteroscedastic) since the social demographics between the groups in this dataset turned out not to be equal. For this research a P-value scoring 0.05 or lower is considered significant, 0.05 to 0.10 probably significant, and bigger than 0.10 is not considered significant.
Interviews
Interviews are recorded, and transcribed as soon as possible to secure the validity of the transcripts. When required by the interviewees, either the full transcript or a summary was shared. After that, the interviews are analyzed and coded resulting in the coding tree in Table 3.3.
3.5 Coding tree
The coding tree presented in Table 3.3 results from the coding process of the interview transcripts from the multiple case study. Right now the coding process will be explained. The findings in Section 4 further elaborate on the themes presented in the coding tree.
For coding the interviews two different approaches are conducted: top-down and bottom-up (Rauss &
Pourtois, 2013). In the first instance, a qualitative deductive approach is used with a list based on the
available literature from Section 2, which can be seen as a top-down approach. However, to guarantee the
completeness of the coding process a bottom-up approach is also conducted. Therefore a qualitative
inductive approach is applied when results were not within the boundaries of the already available coding
tree, and new codes were introduced based on quotes from the interview. These codes are second-order
codes. Next, these second-order codes are grouped as third-order codes, and finally as a general theme.
2
ndorder code 3
rdorder code General theme
Compensation only No store Concept
Individual
Carrier Adjust expectations
Consumer
Environmental impact Social aspect Social contact
Addition to instore network Extra layer Volumes, opening times,
stability
(Non-)logistical Extra services Scalability
Core business
Pickup only Instore pickup point
Servicepoint
Employees Instore Standardization
Store premises Retail chains High volumes
Rental properties Homes
Individuals (no employees)
Minimalistic measurements Corporate identity Improve recognizability
Own initiative Process design
Help available
Multiple partners Partnerships Exclusivity clause
Resources Standardization level
Coordination time
Best practices Community
Buddy system
Pickup point firm Cost allocation Pickup point operator
Opening time Core criteria Selection
Location (multiple) Relative location Language
Human interpretation Peripheral criteria Motivation
Process fit
User fit
Representativeness
Scalable Growth
Online
Consumer feedback
Set criteria Stability
Physical visit
Law Financial Perceived risk dimensions
Review Performance
Training
Process optimization
Store vs home Physical
Selection
Process optimization Time Relative location vs traffic
Environment Image
Social
Privacy operators Privacy
Personal choice
Table 3.3: Coding tree multiple case study
4. Findings
This section first presents the findings from the questionnaire, and after that, the results from the interviews are exemplified following the structure from the coding tree in Table 3.3.
4.1 Questionnaire findings
This section first describes general statistics and then the most important results from the questionnaire, which are briefly introduced in this paragraph. The complete statistics from the questionnaire are available in Appendix A. One of the main findings is the difference in the presence of perceived risk and effectiveness of risk relievers between people living in a city and people living in villages and rural areas (non-city). The most relevant differences between these consumer types are discussed, among which motivation to use social pickup points, the importance of reviews, and differences in perceived risk dimensions in general.
Next, the perceived risk dimensions are prioritized from most important to least important: performance, privacy, time, physical, financial, and then image risk. Lastly, the most significant risk relievers are discussed for each risk dimension.
Descriptive statistics
In total 181 participants started the questionnaire, of which 137 completed it successfully and are ready for analysis. However, not all 137 finished experiments are used for the final results. One respondent was left out because the time to complete was unrealistically fast, and one respondent was left out since he/she had never heard of any last-mile solution and is therefore not a potential user. The final dataset consists of 135 respondents.
Descriptive results are displayed in Figure A.1 to Figure A.7. Although multiple distributors shared the questionnaire to diversify the group of respondents, it must be noted that the social demographic statistics of the respondents do not correspond with that of the Dutch population in general. This follows from respondents that are 65% male, 66% between 20 and 30 years, and 90% has a bachelor's degree. These differences must be taken into account when results are generalized.
City and non-city
One of the findings from the questionnaire is the difference between people in and outside the city. The
living situation in the questionnaire has been divided into cities, villages, and rural areas. In the analysis it
is divided into cities and non-cities, where non-city comprises villages and rural areas. There are two
arguments to combine the results from villages and rural areas. First, from the interviews with the pickup
point firms it became clear that currently there are already undocumented guidelines and differences
between cities and non-cities. Second, the group of respondents for villages and rural areas had similar results but were relatively small sample sizes (35 and 11, respectively). These two arguments led to the decision to combine villages and rural areas and analyze the results making a distinction between city and non-city people.
There are multiple important differences between city and non-city people. Motivation to use social pickup points has been measured, and from Table A.3 it can be concluded that non-city people care more about the social aspect and less about the location than city people.
When measured perceived performance, it is found that non-city people have more trust in social pickup point performance regarding parcel delay, parcel loss, and parcel damage, presented in Table A.4. Non- city people consider reviews from acquaintances more important, whereas city people think general reviews are more important, shown in Table A.5. Both city and non-city people consider reviews on individual locations more important than on social pickup points in general. From Table A.9 it can be concluded that non-city people consider a parking area much more important than city people. Multiple time factors are considered of less importance by non-city people, among which the travel time, possible waiting line, and registration time, which can be seen in Table A.12.
Prioritization
The questionnaire measured risk dimension prioritization in two ways. First by scoring it after the set of corresponding questions on a 5-point scale, and second by prioritizing it at the end of the questionnaire by ranking them in order. Because performance risk is a combination of multiple risks, it is not asked for separately in the prioritization question because of potential confusion. In Table 4.1 the scores of the different risk types are presented, leading to the definitive prioritization in column 3.
Prioritization based on 5-point scale (Table A.19)
Prioritization based on
prioritization question (Table A.20)
Final prioritization
Privacy (4.36) Physical risk (320) 1. Privacy
Time Physical
Time (4.14) Privacy risk (339)
Physical (3.87) Time risk (357)
Financial (3.02) Financial risk (364) 4. Financial
Image (1.68) Image risk (645) 5. Image
Table 4.1: Prioritization of the perceived risk dimensions
The scores of physical, privacy, and time risk are very close to each other and therefore grouped. Since
performance risk is a combination of privacy, time and physical risk it is added to this group. Financial risk
scores a little less high, and image risk is considered of much less importance.
Effectiveness of risk relievers
All the risk relievers from Table 2.1 have been researched in the questionnaire. Analysis of the results led to risk relievers that turn out to be either ineffective and should not be implemented in the social pickup point design, or effective and therefore proposed for standardization. Only the most significant results are presented for each risk dimension.
Financial
Pickup point firms indicated that risk reliever ‘money-back guarantee’ is regulated by law and is therefore obligatory to apply when conducting business in parcel delivery services. Risk reliever ‘free sample’ is not really applicable for social pickup points since it currently already is a free service, but otherwise considered moderately important (3.56), which can be seen in Table A.2.
Performance
Risk reliever ‘reviews’ (3.80) is considered important and has been further investigated by making two distinctions: first between reviewing social pickup points in general and specific locations, and second between general reviews from all people and reviews from acquaintances. These differences are already described in ‘City and non-city’.
The questionnaire researched nine risk relievers for mitigating employee performance risk, and Table A.7 shows that none of these risk relievers is considered effective. Seven risk relievers have been investigated to improve findability, accessibility, and recognizability. Most effective measurements turn out to be showing directions on the firm’s website (3.70) and Google (4.56), showing a photo in the checkout process (4.06), and a billboard at the location (3.44) presented in Table A.9.
Physical
Six out of seven measured risk relievers to mitigate physical risk are considered effective. Table A.11 shows that all types of lighting (4.15, 4.08, 3.99) and a closable cupboard (4.03) are very important. Cameras in the street (3.14) and at the entrance (3.44) are considered moderately effective for mitigating physical risk.
Time
New time risk relievers were researched in the questionnaire. Table A.13 shows that both automatic
updates (3.90) and personal updates (3.54) are effective. The content of these updates is very important
and should include when locations are unexpectedly closed (4.48) and show busy times (3.66).
Image
The questionnaire measured the importance of social pickup points usage on the self-image of the user (2.08), the importance of how friends and family feel about the user (1.53), and the importance of how neighbors feel about the user (1.64) presented in Table A.15. Showing the environmental impact (3.11) turns out to be the most effective image risk reliever, which can be seen in Table A.16.
Privacy
Two privacy risk relievers have been measured and the results are presented in Table A.18.
Communication to consumers on signing a privacy contract by the operator (4.59) and ensuring only operators have access to personal data (4.42) turn out to be effective risk relievers for mitigating privacy risk.
4.2 Interview findings
Whereas Section 4.1 presented the consumer perspective, the current section presents social pickup points from a managerial and operational point of view. This section follows the coding tree structure from Table 3.3 and discusses the following themes: the social pickup point concept, pickup point standardization, the selection process, and the perceived risk measurements.
The social pickup point concept
First the theme ‘concept’ is elaborated on by presenting codes that describe the social pickup point concept. It was found that there is a mismatch between what social pickup point firms can offer in the current business format, and what carriers and consumers expect. This mismatch is described and exemplified, followed by elaborating on the differences between the social and instore pickup point concept.
To introduce the mismatch between social pickup point possibilities and carrier and consumer
expectations, a quote of a co-founder of Homerr is presented: “Carriers expect social pickup points to be
similar to a shop. They want it to be self-representative and present the core values of the carrier. We do
not think that is the concept of a social pickup point. It can be, but then they need to get a salary, wear
company clothes and meet with more strict criteria. In our concept, we just ask people to be at home at
certain times and make use of their hospitality and storage space. That is all we desire. If we would desire
all kinds of requirements, then we would not be looking for homes but stores. It is not possible to demand
too much from somebody that only gets €0,28 compensation per parcel. Consumers need to start to
understand that as well. They need to appreciate the social aspect, which is, for example, lowering the
number of vans in the street and stimulating contact between neighbors. That is the difference between social pickup point at homes and servicepoints in stores.” (SPP2.1). This quote corresponds with the first three third-order codes from the coding tree by mentioning social pickup points are no stores, expectations need to be adjusted, and focus should be on the social aspect.
Action is required to adjust expectations since a ViaTim employee also indicates the limitations of the social pickup point concept: “Some things we can demand, but we have to keep in mind that social pickup points are not firms with employees. We do not pay rent or salary, only a small compensation. Therefore we are limited in what we can demand, although there are more things that we would want to demand when possible.” (SPP1.3). Next to the difference between employees and individuals, there is a fundamental difference between the primary business of social and instore pickup points: “For instore pickup points it is standard that they are representative because their own goal is for consumers to visit their place. That is a difference with social pickup points that are places where people live.” (SPP2.2). The different primary purposes lead to a different service level: “At shops, the employees have chosen to become a shop assistant and have contact with customers for work. Therefore you can expect a higher level of customer care from them than from somebody that signed up for a social pickup point. I think there is a difference between them.” (SPP1.3). This difference is exemplified by a store operator that extended its opening times the day before Boxing Day (IPP6) and social pickup point operators that unexpectedly closes because of daily activities such as walking the dog and bringing their children to sports (SPPO1, SPPO4).
Homerr does not believe the social pickup point network can be a self-sufficient network but states it can have value in areas with fewer shops: “We think it is a nice extra. Look, we do not believe the complete network can run on social pickup points. Not in terms of volumes, opening times, and stability. But they can be of value at places where it is difficult to find instore pickup point locations. In certain areas, villages actually, it is difficult to start instore pickup points, and then social pickup points are a good addition.”
(SPP2.2).
Both PostNL and DHL mentioned the limitations of social pickup points compared to instore pickup points.
The main limitations are, among other things, scalability of the number of points, scalability of the number of parcels per point, and stability of the points. The carriers do see advantages such as opening times outside office hours, geographical proximity, especially in areas with fewer shops, and the social aspect.
The social aspect can be extended when social pickup points offer extra services (SPP1), and the more
extra activities, the more stable the pickup point (SPP1, SPP2). However, offering these services requires
more coordination and requires a trade-off between benefits and extra investments. There is a distinction between Homerr that only focuses on scalable logistical services and ViaTim that includes the social aspect with its extra services: “I do not believe you can organize and execute all these extra services properly.”
(SPP2.1). Whereas ViaTim coordinates the non-logistical services where possible: “Extra services that can be easily centrally coordinated by ViaTim will be uniform.” (SPP1.2).
Pickup point standardization
The codes that are elaborated on in this section are first briefly introduced to show the findings on social pickup point standardization. Standardization is divided into two types: corporate identity standardization and process standardization, and for these standardization types differences are described between social and instore pickup point companies. Corporate identity entails the ‘benefit communicators’ and ‘system identifiers’ from Kaufmann et al. (1999) and process standardizations comprise the ‘system facilitators’.
Social pickup point operators indicated they would appreciate improving the currently low corporate identity standardization to improve recognizability (SPPO1, SPPO4), but remark that they are restricted in making definitive changes to their locations because they rent their homes (SPPO1, SPPO2). The social pickup point operators indicate to appreciate the freedom to design the workspace and process themselves, with assistance from the social pickup point firm when they need it (SPPO1, SPPO2, SPPO4) Instore pickup points had a higher level of standardized corporate identity. The points of DHL, UPS, DPD, and GLS had a similar level of corporate identity with signing inside behind their service desks, and branding that was also visible in the dark. PostNL locations had an increased level of corporate identity and process standardization since they received a branded service desk, and cupboards to store parcels out of sight and reach from the customers. There is a difference in the number of parcels pickup points process. PostNL locations stored 150 to 250 parcels, other instore pickup points had a maximum of 75, and social pickup points indicated to store 5 to 15 parcels.
PostNL is the only carrier that conducts an exclusivity clause. Therefore PostNL has full control over a
location and can apply standardization without the influence of other carriers. Having full responsibility
has also been one of the reasons for PostNL not to cooperate with social pickup point firms but to run their
own pilot (FC1.1). Most instore pickup points cooperate with multiple carriers and therefore it is hard for
these carriers to standardize according to their specific needs. Currently social pickup point firms only
conduct parcels of one carrier, but they have the intention to partner with multiple carriers.
In the interview PostNL indicated that they have the resources to facilitate their pickup points with the tools they need and that they have specialized teams that improve the customer journey and thus optimize and standardize the process. DHL indicated to consider more standardization in their pickup points with the highest capacity and upgrade these with a DHL branded desk and full services. Thereby DHL emphasized their prudence because of the corresponding high investment costs. The degree to which social pickup point firms can affect the concept depends on the amount of time and money they can invest:
“At ViaTim we think it is important to give the social pickup points freedom. We give them the responsibility to work out their own initiatives. If there would be resources to facilitate and manage those initiatives that would be ideal. Then it would be possible to always shape it according to the company values, but that is time and money intensive. Then you need management to coordinate that, but currently we do not have time for that.” (SPP1.2). The social pickup point concept is thus restricted in standardization by time and money and depends more on the execution of individual operators.
Instead of enforcing standardization using resources, Homerr focuses on obtaining standardization by using their ‘community’. Homerr tries to inspire social pickup points to improve by showing best-practices and instructions via social media and newsletters. Also, they introduced a buddy system in which social pickup points are coupled to learn from each other. Standardization caused by copying from other points can decrease the coordination effort for social pickup point firms. A ViaTim point operator also mentioned the need for learning from others: “It would be useful if they organized a meetup day or get you in contact with other operators. Right now we operate alone and have to invent everything ourselves, while other operators have probably more experience.” (SPPO1). It can be concluded that a community can partly take over standardization activities from firm management and ViaTim indicated to startup a buddy system as well.
Social pickup point selection process
Right now the third-order codes of the theme ‘selection’ are briefly introduced, and these are further elaborated on in the upcoming paragraphs. The level of standardization is relatively low at social pickup points, and in this section, it is described that the selection process is the most important tool to influence social pickup points. It is concluded that current social pick up point firms have two evident distinctions in the selection process that lead to different networks: the strictness of selection and physical visits or not.
The selection process and the corresponding type of growth influence the social pickup point concept. At
the start of instore pickup points DHL was in the same position as social pickup point firms now: “When
we started with our instore pickup point network, we were in the same position as ViaTim and Homerr right
now. Then you have to start up the network. First you focus on quantity, and when that is sufficient, you narrow it down and focus on quality. If you want to focus on sky-high quality right away, it is possible, but then the network grows very slowly.” (FC2). Homerr focuses on growth with less strict selection criteria and continues selection while operating using customer feedback: “Some criteria are difficult to check in the selection process and hard to monitor ourselves. However, customers pick up parcels at the points every day, and feedback is given by them. When it turns out that something is not right, we hear it from the customers.” (SPP2.1). ViaTim aims for sky-high quality right away using stricter selection criteria: “We also want to grow, but only when growth is possible while staying true to our core values.” (SPP1.3). The firms have to make the trade-off between pursuing a social pickup point network quickly or qualitatively.
The type of growth depends on the strictness of the selection criteria. All pickup point firms indicated to make a distinction between strict core criteria and optional peripheral criteria. Peripheral criteria are location-dependent, and human interpretation is required to judge relevance. This quote from the interview at PostNL exemplifies the importance of human judgment: “We have account managers that are responsible for recruiting new locations. In their selection process they have some freedom; there is some grey area. We have uniform rules, but most are open for personal interpretation. In the selection process the need for capacity is central. When capacity is more urgent in a specific area, the guidelines are less strict.” (FC1.1).
In essence, the social pickup point firms have the same philosophy in selecting their locations and their basic criteria are the same, following from the interview with DHL: “Social pickup point criteria that we have at DHL are equal for ViaTim and Homerr.” (FC2). Next to the basic criteria also the room for human interpretation is similar: “The main goal of all questions in the selection process is: Would I pick up a parcel here myself? Would I feel safe? Would I feel comfortable to visit this location?” (SPP1.2) and “You need to be able to send your mother there.” (SPP2.2).
A big difference between selection methods is the choice for physically visiting potential locations or not:
“One of the key differences between ViaTim and Homerr is that we always physically visit potential points and consequently reject a lot. Only one out of fifteen registrations eventually becomes a ViaTim point.
Homerr does not physically visit potential locations and one out of two registrations becomes a Homerr point, leading to less stability. If a webshop complains about a point at Homerr they close it, but at ViaTim we have ensured the quality of the points and therefore not just shut down a point of which we think it has the right quality.” (SPP1.1). At Homerr it is explained why they switched from physical visits to video calls:
“From the video call we obtain sufficient information. We decided not to visit all potential locations
physically. In the past we did that, but it is not scalable, and therefore we do not do that anymore. Next to that, it also does not give a good representation of reality. They can clean the house, buy fresh fruit, and put it on the table and all that for the day of the visit. But the parcels go there on a daily base and feedback we receive from the consumers. Therefore we will hear it automatically when it is not right.” (SPP2.1). This choice for physical visits, or not, is in line with the firms’ intended growth types.
Social pickup point perceived risk measurements
Most perceived risk measurements are consequences of the selection and standardization strategy of the social pickup point firms and already described. However, the six perceived risk dimensions will shortly be described in this section and when already discussed in previous sections it will be referred back.
Pickup point firms indicate that financial risk for consumers is already regulated by law and therefore they cannot influence it. There are three measurements highlighted to mitigate performance risk. First, process optimization is achieved by the selection and standardization strategy. Second, training is provided to operators, although it is considered an extra since the focus is on process simplification leading to a natural process for which a training is redundant. Lastly, general reviews on the social pickup point firms are available to consumers. Individual reviews on points are not available: “We only show the rating of Homerr in general. With reviews, especially for pickup points, we need to watch out that people base their choice on a single review. That is because people can write a negative review, also when the mistake is not in the hands of that specific point and operator. Therefore we only show a general score for customers to base their choice well-grounded.” (SPP2.2).
For mitigating perceived physical risk the selection process is used as described in the ‘selection’ theme.
PostNL points out an elementary difference between stores and homes concerning standard safety measurements: “Most retail locations already take safety measurements seriously to protect their own customers and goods. They can, for example, have cameras or an alarm system. That is probably not the case for individuals and their homes.” (FC1.2).
Social pickup point firms did not mention extra measurements to mitigate time risk, apart from the criteria
they demand in the standardization and selection process. No social pickup point firm indicated to have
taken specific measurements to mitigate perceived image risk, although on their websites, Homerr
communicates the social impact and ViaTim communicates both environmental and social impact. ViaTim
and Homerr indicated to protect the privacy of their operators but did not mention specific measurements
to mitigate the perceived privacy risk of the consumer.
5. Discussion
This section discusses how the findings from Section 4 contribute to literature. First, the risk dimension prioritization from the current paper is compared with risk dimension prioritization in digital service literature. Next, the importance of shared responsibility is discussed to solve the current mismatch between the social pickup point concept and consumer expectations. Lastly, modular standardization is discussed to improve shared responsibility by modularly standardizing based on customer needs, the number of parcels processed at a point, and the operator type.
5.1 Risk dimension prioritization
Risk dimensions were not yet prioritized in last-mile literature. The current paper prioritized risk dimensions for social pickup points and therefore fills a gap in the literature. To validate the prioritization the findings of the current paper are compared to prioritization of risk dimensions in digital service literature.
By comparing risk dimension prioritization literature with their own findings, Mustafa and Kar (2019) select privacy and performance risk as the main driving factors for the adoption of digital services. However, this conclusion is considered premature. Therefore in Table 5.1 risk prioritizations of seven papers have been listed and compared to the findings from the current paper.
Performance Privacy Time Physical Financial Image
(Cases, 2002) H H M - M L
(Forsythe & Shi, 2003) H - M - M M
(Featherman et al., 2003) H H M - H L
(Lee, 2009) M H H - H L
(Crespo et al., 2009) H M M - H L
(Luo et al., 2010) H M M L H L
(Mustafa & Kar, 2019) H H M M H M
Findings current paper H H H H M L
Table 5.1: Prioritization of risk dimensions in literature and current paper (H = high prioritized, M = medium prioritized, L = low prioritized)
