Trading off convenience versus control : a study on the impact of annoyance and time pressure on someone’s decisions for convenience or control

(1)

Master Thesis

Trading off convenience versus control:

a study on the impact of annoyance and time pressure on someone’s decisions for

convenience or control

Remko de Bruin

Student number: 10589341

22-06-2018

Business Administration - Marketing

Supervisor: Joris Demmers

(2)

Statement of Originality

This document is written by Remko de Bruin who declares to take full responsibility for the contents of this document.

I declare that the text and the work presented in this document are original and that no sources other than those mentioned in the text and its references have been used in creating it.

The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.

(3)

ABSTRACT

Background – People constantly trade off convenience versus control in online service

settings. Some research has been done about this trade-off, but little is known so far about how companies are (mis)using and nudging parts of this trade-off to steer customers into making decisions which are in the company’s benefit instead of the customer’s interest. This could lead to customers who are perceiving annoyance, which could affect people’s decision for convenience or control.

Purpose – This study investigates how online service companies can influence someone’s

convenience orientation or desire for control with the design of the digital customer journey. Moreover, the impact of annoyance and time pressure on someone’s decision for convenience or control is investigated. Accordingly, the main research question is: how do customers trade off convenience versus control in online service settings and how could online service companies profit from this? Based on several psychological theories like reactance theory, it is expected that people will behave differently in their decisions for convenience or control when they perceive a certain amount of annoyance or time pressure.

Methodology - Data was gathered by means of a digital survey that was distributed via

snowball sampling. Respondents (N = 183) had to go through a simulated digital customer journey in which they had to make decisions for convenience or control, while being exposed to different annoyance-provoking stimuli.

Findings – Different sources of annoyance in the online environment have been found.

Moreover, statistically significant results have shown that interfering a digital customer journey with pop-ups and steering questions lead to perceived annoyance. However, no significant results have been found that this annoyance impacts someone’s decision for convenience or control. Forcing people to complete a customer journey within a given time frame did not show to have significant impacts.

Implications – Although only 2 hypotheses have been supported, the results of this study do

provide interesting insights. For example, someone’s level of reactance has been found to significantly impact someone’s perceived annoyance and desire for control. With regard to practice, implications are that organizations should consider segmenting customers on the basis of situational factors that influence someone’s preference for convenience or control, like the importance of the service outcome and the level of reactance. Companies should consider as well that some nudging features like pop-ups and steering questions could lead to annoyed customers, which could have negative implications. Overall, these and other findings provide several directions for future research.

(6)

1

1. Introduction

Customer journeys are getting more complex, given that nowadays customers are connected all

the time and have more choices and decisions to make than ever before (Lemon & Verhoef,

2016). It is likely that every decision that the customer has to take, consists of some choices

which are more beneficial to the firm than other choices. It is therefore highly important for

firms to know what their opportunities are to influence the customer decision process at the

right time in the right context. One way to do this is by nudging. A nudge can be seen as an

indirect suggestion to influence the behaviour and decision making of groups or individuals

(Kosters & Van der Heijden, 2015). These nudges are a commonly used mechanism in the

digital environment. For example, in many situations, customers have to decide if they want to

obtain a service in a fast and easy way (convenience), or if they want to be more involved in

the service process which might take more effort and is more time-consuming (control). A good

example of how a nudge can influence this decision between convenience and control is on the

online service platform Airbnb. This company tries to influence its customer to give away a

certain amount of control by offering a service in a more convenient way. It is so that hosts on

Airbnb can decide to let guests book instantly, without any pre-screening of the guest’s profile

(What is Instant Book?, n.d.). This may sound frightening to hosts who rent out their places,

but this decision is framed in such a way that makes it hard for customers to not take advantage

of the option. The decision process takes customers into a course of action in which the website

shows the customer the value of turning on this option. For example, the platform shows their

customers how they could gain way more bookings if they would switch on this option, shows

them all of the advantages, and tries to counterargument every reason a customer might have to

not use the option.

In this case, the customer has to make a decision about giving away a certain amount of

(7)

2 Service convenience can be seen as the convenience that reduces customers' time and effort in

obtaining a service (Berry, Seiders & Grewal, 2002). Customers nowadays are expecting this

service convenience and want companies to be able to satisfy them by offering services without wasting the customers’ time and effort (Womack & Jones, 2005). On the other hand, customers

want to keep as much control over a situation as possible (Phelps, Nowak & Ferrell, 2000). Control can be seen as the customer’s competence, superiority, and power within the sales

relationship’ (Mullins, Bachrach, Rapp, Grewal, & Beitelspacher, 2015, p. 1073). It seems that

more and more service platforms are trying to influence decisions about parts of this trade-off.

The example of Airbnb shows that this firm is using the fact that customers want service

convenience by making firm-profiting decisions to use one of their services very convenient to

its users. As for Airbnb, it would be profitable to have more Instant Book listings on its

platform, to grow its share on the traditional vacation rental market (Ting, 2017). There are

numerous examples of companies using convenience as a way to make more profit. Lots of

companies are charging extra money for the privilege of paying for a service using an

alternative, possibly more convenient payment. However, many customers do not see this

alternative payment option as a convenience, but just as a way for companies to make more

profits (Holmes, 2017). Another example of companies trying to make more profit by forcing

customers to trade off convenience versus control are login walls. Many websites still force

users to log in before presenting them with any real content. This requires customers to decide

if they want to trade control over the situation by giving away their personal data to be able to

enter the website. Service providers, routinely collect and use this individual specific customer

information in their benefit (Budiu, 2014; Phelps et al., 2000).

Another strategy could be to make firm-harming decisions very inconvenient to its

users. This is, for instance, applicable to social network platforms that make it almost

(8)

3 with strategies to complicate this opt-out process. Examples are banishing the ‘unsubscribe’

option to the bottom of the page or changing the delete option to a more misleading ‘Change Preferences’. This forces the customer to think if it is really worth the effort to go through the

exhausting process of deleting their account to get back some of their control or to just save the

effort and keep their account. Customers deleting their account would be harmful to the firm,

as this will decrease the amount of time and attention customers are devoting to the network,

and thus lowers the profit that the firm is making through targeted ads and in-app purchases

(Crump, 2017).

These examples of retaining the customer through the creation of inconvenience or by

the provision of a lot of convenience are called lock-in strategies. These are strategies in which

the customer is so dependent on a company for its products and services that the customer

cannot move to another seller without substantial switching costs (Khosrowpour, 2013).

The situations that are described here show that providing convenience to customers

could make the online experience more straightforward, but it could also lead to annoyance and

disruption (McCoy, Everard, Polak & Galletta, 2007). Customers might have the feeling that

the company is nudging them towards firm-profiting decisions, or on the other hand, customers

could get the feeling that the company is not providing them enough convenience. It is

interesting to research which role annoyance plays in the decision between convenience or

control.

Some research has been done about this trade-off between convenience and control, but

not about how companies might influence parts of this trade-off in their benefit. Little is known

so far about how companies are (mis)using and nudging parts of this trade-off to keep customers

from leaving the firm or steer customers into making decisions that are in the company’s

benefit. Besides that, there is still minor conformity about the motivations behind choosing for

(9)

4

How do customers trade off convenience versus control in online service settings and how could online service companies profit from this?

In order to find an answer to this research question, more knowledge is needed about how the

trade-off between convenience versus control is driven by diverging motivations. Next to that,

it has to be found out what provokes customers’ annoyance in a digital environment and if this

annoyance drives people to take more control, or if it drives them to go for more convenience.

This knowledge could have practical implications for both companies and customers.

Service companies could use this knowledge in order to create their online platform in a way

that is the most beneficial to them. On the other hand, customers could learn how online service

companies are nudging them to make decisions that are more in the company’s benefit, instead

of their own benefit.

This research will have theoretical contributions as well. It will add knowledge to the

prioritizing marketing topic about the changing decision-making processes. The more complex

customer journeys are, the more it gives firms the opportunity to influence the customer

decision process by nudging the choice architecture. This concerns both the trade-off between

convenience versus control as well as the trade-off between annoyance versus effectiveness

(Marketing Science Institute, 2016). This research will address both of these trade-offs and will

provide valuable knowledge about the motivations behind the exchange of the two factors.

First of all, I will elaborate on the concepts of choice architecture, convenience, control,

and annoyance. Specifically, I will go into more detail about the diverging motivations of

customers to choose for either convenience or control and how annoyance might influence this

decision. An online survey experiment is conducted to derive the relevant results. The paper

(10)

5

2. Literature review 2.1 Choice architecture

Choice architecture describes the fact that there are many ways to present a choice to a

decision-maker, and that what is chosen often depends on how the choice is presented (Johnson et al.,

2012). It turns out that choice architecture has a significant influence on the final decision of

someone. Influencing the choice architecture can be done in many ways, for example by varying

the order of the choices, the number of choices presented and the selection of defaults. Defaults

are settings or choices that apply to individuals who do not take active steps to change them

(Johnson et al., 2012). An example of this default which influences someone’s decision for

convenience or control is organ donation, a topic which is in a lot of debate in the recent years.

Many people who avoid making an active choice about being a donor, are passively accepting

the default option (Goldstein, Johnson, Herrmann & Heitmann, 2008). This means that in

countries where inhabitants are organ donor set by default, have relatively way more organ

donors than countries where the inhabitants have to make an active choice to become a donor.

This example shows that a specific choice architecture, called a nudge, is not only used to

influence customers in their decision process but could be used on any level to improve

decisions about health, wealth and happiness amongst others.

2.1.2 Nudging theory

A nudge is any aspect of the choice architecture that changes someone’s behaviour in a

predictable way without forbidding any options or significantly changing their economic

incentives (Kosters & Van der Heijden, 2015). An influencing aspect can be seen as a nudge

when the intervention is cheap and easy to avoid. Nudges are therefore not the same as

mandates. Nudges are seen to be effective because people are unable to make economically

optimal decisions. The reason for this is that people do not have the capacity to store a large

amount of information that is needed for such decisions, as well as the cognitive ability to

(11)

6 not act in their own best interest from an economics point of view and are likely to get

influenced by these nudges. Although nudges can be used in both positive and negative ways,

this study looks primarily at the more negative influence of nudges on online customers. These

customers can be influenced during different moments of a service delivery process. As made

clear, many of the decisions that an online customer has to take, are about obtaining

convenience or about keeping control. For instant, online customers might have to decide if

they want to search through just a part of a database which is selected by use of a personalized

algorithm or if they want to search through the entire database. Searching through just a part of

a database which is already narrowed out on the customer’s preference would be more

convenient, but this would mean that the customer has to give away valuable personal data.

Choosing to search through the entire database would probably be more time-consuming, but

would keep the customer more in control. The decision which makes it able to accomplish a

task with the least expenditure of human energy can be seen as a convenience-oriented decision.

2.2 Convenience

Collier & Sherrell (2010) define convenience as the customer’s ability to find and facilitate

their need with the least amount of time and effort. This includes the physical and mental effort

that someone has to take, as well as the financial expenditure. A difference can be made between

convenience goods and service convenience. Convenience goods are goods that can be

purchased with minimal time and physical and mental effort. Service convenience can be seen

as the convenience that reduces customers' physical and cognitive effort in obtaining a service.

This could contain a combination of both selling goods and providing services (Berry, Seiders

& Grewal, 2002). In this study I will be focussing on this service convenience, and in particular

in the context of online service organizations. These organizations are increasingly using online

technology to facilitate marketplace exchanges in a variety of service settings, which can be of

beneficial value to both customers and service providers (Walker, Craig-Lees, Hecker &

(12)

7 progress, the socioeconomic changes, the more competitive business environments and thus

higher opportunity costs. Outcomes of these changes are that customers have to choose between

an increasing amount of online service companies, while the time they want to devote to the

service process is decreasing (Berry et al., 2002). Therefore, the customer nowadays is looking

for the place where the total costs or the effort and time spent in the entire service process are

the lowest (Bhatnagar, Misra & Rao, 2000).

The research that has been done on service convenience so far can be divided into two

parts; research related to time convenience and research related to effort convenience. Most

early research studied service convenience mainly as the amount of time that a customer has to

invest to realize a certain need. Gehrt, Yale & Lawson (1996) were one of the first to suggest

that time is just a part of the overall effort that a customer has to take to realize a certain want.

Customers' effort is a big part of the nonmonetary costs that influences perceived convenience

and satisfaction (Berry et al., 2002). One possibility for making a service more convenient to

customers is to make the online experience more or less difficult by choosing how much

information a company supplies to the customer during a decision making process. If

information is given in a very detailed way, the customer has to put quite some effort to process

this information. If less information is given, the customer may has to put some effort to obtain

more information elsewhere (Collier & Sherrell, 2010). So when the right balance in providing

this information is missing, the customer has to make a considerable amount of extra effort.

This can be seen as service inconvenience (wasting effort) (Berry et al, 2002).

2.2.1 Inconvenience

Inconvenience is a concept that has received a way smaller amount of attention than

convenience (Farquhar & Rowley, 2009). Whether a situation is perceived as convenient or

inconvenient depends partly on the choice architecture. For example, inconvenience can strike

up in situations where customers are estranged by aggressive and unnecessary procedures,

(13)

8 benefit (Farquhar & Rowley, 2009). This inconvenience might be an important aspect of the

decision-making process of the customer. One of the reasons that there has been done less

research on this subject could be that most research has been written from the perspective of

the producer, and there are not many producers who acknowledge that they design their

products or services in an inconvenient way. However, inconvenience is a common experience

in the digital customer journey, and more understanding of inconvenience is required in service

economies that include so many time-poor customers (Berry et al., 2002). These customers

expect fast and efficient processing of their online activities. However, long registration forms

to be filled in, complex procedures, or no availability of alternative payment methods are all

considered to be major inconveniences that make digital journeys complex and cause customer

dissatisfaction (Rajamma, Paswan & Hossain, 2009). Convenience, or inconvenience, can

strike up in different moments of the customer journey. Hence, different types of service

convenience can be distinguished.

2.2.2 Types of service convenience

Berry et al. (2002) describe different types of service convenience. Decision convenience

describes the effort customers have to devote to decide if they want a certain service or not.

Aspects of decision convenience consist of the amount of provided useful, clear, and easy to

read information (Colwell, Aung, Kanetkar & Holden, 2008). Next, access convenience

involves customers' required actions to request service and, if necessary, be available to receive

it. Examples of this type of convenience are the availability of the service provider, the

accessibility of the service provider through various ways, the time it takes to take a certain

action in the service process and the easiness to contact an employee of the service provider

(Colwell et al., 2008). At last, transaction convenience is about the amount of effort customers

have to take to realize a transaction. This depends for example on the easy and simple check

out process and the availability of different payment methods (Jiang, Yang & Jun., 2013). The

(14)

9 customers’ expenses on, and experience of, a service of a particular company (Moeller,

Fassnacht & Ettinger, 2009). One reason for this is that these types of convenience are factors

that come into play prior to a certain action. It seems that service providers are using these types

of convenience more and more to nudge their customers, and even weaken their customers’

position (Walker et al., 2002). The examples in the introduction show how companies are

influencing their customers with different kinds of service (in)convenience. A very high

decision convenience could lead to customers taking decisions which might be more beneficial

for the firm than for themselves, and a very low access convenience could result in customers

giving away valuable personal information or might keep customers from leaving the firm.

2.3 Control

On the one hand, customers want more convenience, on the other hand, customers want to keep

as much control as possible (Phelps et al., 2000). Control can be defined as ‘the customer’s

competence, superiority, and power within the sales relationship’ (Mullins, Bachrach, Rapp,

Grewal, & Beitelspacher, 2015, p. 1073). Control plays a big role in the customer’s decision

process and it is including customers’ approval, modification, and opportunity to in or

opt-out (Malhotra, Kim & Agarwal, 2004). In general, control can be seen as the customers’ ability

to determine the outcome of the service experience by handling the service offering in such a

way that it matches their ability, needs, and desires for the service outcome (Collier & Sherrell,

2010). Before going into more detail about the different types of control and the incentives of

having control, it has to be emphasized that there is a difference between actual and perceived

control.

2.3.1 Actual and perceived control

Control is an important construct in psychology literature and has been conceptualized in

different ways (Guo et al, 2016). In the current literature, a difference is made between actual

and perceived control. Actual control can be seen as the user’s ability to modify the (online)

(15)

10 levels of actual control provided to them. This means that even though a website may be

designed to provide a certain type and level of actual control, customers can still perceive this

level and type of control that the website provides very differently (Manganari, Siomkos &

Vrechopoulos, 2014). Although research demonstrates that customers make imperfect

evaluations of their own level of control, many academic papers do not emphasize the

difference between perceived control and actual control. Measures of actual control are

desirable, but they are difficult to obtain (Faranda, 2001). For that reason, in this study, ‘control’

mainly refers to perceived control.

2.3.2 Incentives of having control

There are many incentives for companies to provide customers with a feeling of being in

control. Existing research on the role of control within online service platforms showed that customers’ need for control was one of the biggest motivators of using online platforms

(Korgaonkar & Wolin, 1999). Next to that, service customers seem to prefer options that foster

perceived control even when there are no monetary incentives involved (Collier & Sherrell,

2010). Perceiving a certain amount of control could even mean the difference between

customers using a service or not using a service (Howard & Worboys, 2003). In general, control has shown an influence on customers’ overall evaluation of service quality and may lower the

risk perceptions (Collier & Sherrell, 2010).

Someone’s sense of being in control can also be influenced by the choice architecture.

The perception of having a choice, or even multiple choices over a certain situation can have a positive impact on someone’s sense of control. This can lead on its turn to the experience of

more positive emotions. This works the same the other way around. Having a sense of control

over an unpleasant situation reduces negative feelings, such as anxiety. For example, having a

choice in how someone is punished can effectively reduce the negative feelings toward the

(16)

11 are likely to experience more customer satisfaction when their increased perceived control is

high (Chang, 2008).

2.3.3 Control on websites

The notion of control is in particular important for websites because online environments may

create a feeling of limited control to customers. For example, the inability of the customer to

feel or touch the products, and the limited interaction with the personnel may decrease the customers’ perceptions regarding their level of the control (Manganari et al., 2014).

The degree of control which is provided to customers on websites may have a big

influence on the online customer behaviour. Giving customers control over the information they

encounter can lead to lower levels of uncertainty. In general, control has a positive influence on

customers’ responses. On the other hand, perceiving a lower amount of control during the online

customer journey can result in negative emotions, lower website attitude ratings, and intentions

to avoid the online store (Manganari et al., 2014).

Because of some of the inabilities of websites compared to physical stores, and the way

the choices are architected, it is not unusual that mistakes are made during the service delivery.

For example, a customer wanting to book an economy class flight was eventually charged for

business class because that was the option that was set as default. Customers desire to exercise

control at all stages of the service process, so providing the customer with control plays also an

important role in any of these mistakes made during or after the service delivery. Service

mistakes may even grow customers’ need for control during the service recovery, because of

the perceived loss of control during the service failure. This may motivate them to have a need

for a choice in the service recovery in an effort to regain control (Guo, Lotz, Tang & Gruen,

2016). Guo et al. (2016) researched this and found out that, after a service fails, involving the

customer in resolving the service failure plays a critical role in composing their perceptions of

(17)

12 process when resolving service failures, companies can improve customers’ evaluations of

service recovery and increase their repurchase intentions. For example, giving customers a

choice in determining the outcome of service recovery, may regain their perception of control,

reduce negative feelings about the company and push their repurchase intentions (Guo et al.,

2016). This is also related to the ‘service recovery paradox’, which means that customers have

more positive feelings about a firm after the firm has corrected any service failures, compared

to their feelings regarding this firm if no problem had occurred. This is because correctly

solving a mistake in the service delivery leads to more trust and confident feelings about the

firm (Krishna, Dangayach & Sharma, 2014).

2.3.4 Types of control

In one of the earlier papers, Hui & Bateson (1991) distinguished different types of control;

behavioural control, cognitive control, and decisional control. Behavioural control is defined

as the availability of a response which may directly influence or modify the characteristics of a

situation. This type of control can be seen as high when someone can, for example, complete

an online form in any sequence or order that he or she wants. Cognitive control can be seen as

someone’s ability to predict and interpret a certain situation. Cognitive control is perceived as

high when a company provides the right amount of information during a certain situation. At last, decisional control refers to someone’s choice in the selection of outcomes or goals.

Decisional control is high when a customer believes he or she has the decision to enter a

situation or has perceived alternatives (Esmark, Noble, Bell & Griffith, 2016).

2.4 Trade-off between convenience versus control

Since people expect services to be convenient while at the same time they are giving a sense of

control, it is desirable to create a service process which provides both of these factors. However,

in many cases in the online environment, the gain of more control goes at the expense of a loss

of convenience and vice versa (Ball & Callaghan, 2012; Sifry, 2014). One reason for this is that

(18)

13 account for some of the tasks and responsibilities of the customer (Berry et al., 2002). This

limits the customer’s superiority over the relationship with the firm and thus feels like a loss of

control for the customer. Oppositely, acquiring more control over the sales relationship is

usually associated with filling in longer forms, reading and adjusting service conditions or

having to compare alternative options. These are all time-consuming tasks which lowers someone’s sense of convenience. That is why the decision-making process in online service

settings is often characterized by a trade-off. This trade-off between providing convenience to

customers and maintaining customers’ control is getting more complex and differs widely per

customer (Ball & Callaghan, 2012). Individual customer differences are for example the

convenience orientation of a customer (a person's general preference for convenient services)

or the customers’ level of empathy with service firm employees (Berry et al, 2002). People are

constantly making trade-offs between control and a number of factors such as money,

convenience, and capabilities (Tam, Glassman & Vandenwauver, 2010). However, the way

how someone divides the amount of control with these other factors is highly individual and

depends on a range of variables. These different variables that have an effect on someone’s

preference for convenience and control will be discussed in the next section.

2.4.1 Variables affecting convenience orientation

A convenience-oriented customer can be seen as one who seeks to accomplish a task with the

least expenditure of human energy (Farquhar & Rowley, 2009). Several scholars aimed to

research the variables that affect convenience orientation in many different industries, but

consistent results are still leaking. Berry et al (2002) found that total household income is

positively correlated with convenience-oriented decisions. Next to this, Bernués, Ripoll &

Panea (2012) found that age was correlated with convenience-oriented decisions within the

food industry. They found that younger people (<45 years old) were more likely to choose for

options which led to convenience than older people (>45 years old). In another research in the

(19)

14 negatively correlated with convenience-orientation, meaning that a lower education level may

result in a higher convenience orientation. Also other demographic variables, such as

occupation and working hours, and lifestyle variables, such as devotion to work have all been

shown to have some impact on convenience orientation, but most of these findings are derived

from old papers and not consistent (Farquhar & Rowley, 2009). Since recent research into

variables affecting convenience orientation within the (online) service industry is still leaking,

it is difficult for marketers to segment customers according to their convenience-orientation.

Besides these demographic and lifestyle factors, there are also situational factors which

could play a role in someone’s convenience orientation. For example, when someone is

purchasing a service with a highly valued outcome, customers would likely be more tolerant of

inconvenience. This is because most high involvement purchases include relatively high levels

of perceived risk, and customers typically use more cognitive effort when making

high-involvement purchase decisions (Berry et al., 2002). Another situational factor that affects someone’s convenience orientation is the amount of time pressure that is perceived. Farquhar

& Rowley (2009) found that when people feel that they have limited time, they are more likely

to have a preference for convenience. There are also situations in which customers would be

more control-oriented.

2.4.2 Variables affecting desire for control

Mullins et al. (2015) researched in which situations customers desire control or are willing to

settle for less control. They found that when customers feel socially connected with the service

provider, they experience less desire for control in the sales relationship. Companies can build

this social connection with customers by affording the customer access to resources and

information and showing the customer that the company is acting in the customers’ best interest.

On the other hand, if customers feel that the company is not acting in their best interests, the

customer will have more desire for control (Mullins et al, 2015).

(20)

15 out that customers are less willing to relinquish control when the service is of greater

importance (Chang, 2008).

Next to that, gender might have an influence on desire for control as well. Prior research

has demonstrated that women make more negative evaluations of their skills in the online

environment than men. They also feel less in control as online users, experience higher levels

of risk and are more interested in risk-reduction (Manganari et al., 2014). Taking this into

account, this might mean that women could behave more towards control as compared to men

regarding their decisions for convenience or control.

Lastly, Ball & Callaghan (2012) found that, for an online service customer, also the

customer’s experience with the service and even the customer’s personal mood are of a high

influence. When online customers are in a good mood or have a good experience with a service,

they are more likely to give away control over a situation than when they are in a bad mood or

have a less good experience with a service. The concepts of convenience and control are

summarized in Table 1.

Taking all these factors into account, providing convenience to customers could make

the online experience more straightforward, but it could also lead to annoyance and disruption

(McCoy et al., 2007). When a certain sweet-spot between control and convenience is not offered

to the customers, this may lead to sincere annoyance (Ball & Callaghan, 2012). For example, a

very inconvenient service could be the simple process of requesting a new visa online. This

may contain countless loops of miscommunication and long waiting times. This process is very

irritating and is often reinforced by help desks and customer support centres that are not helping

and not supportive (Womack & Jones, 2005). On the other hand, it could be that the process of

requesting a new visa could be done by one click, since the customer is asked to log in, to verify

all the information the company already has about the customer and to accept all the conditions

(21)

16 the way the company gathers and uses personal information and this perceived loss of control

could lead to annoyance. The question which then arises is whether this annoyance drives

(22)

17

Table 1: Convenience and control conceptualized

Definition

Affecting variables (positive (+) or negative (-)

impact) Types Aspects

Convenience The customer’s ability to find and facilitate their want with the least amount of effort

Total household income (+), age (-), education (-), occupation, working hours, time pressure (+), devotion to work and importance of the outcome of the service (-)

Decision convenience: the effort customers have to devote to decide if they want a certain service or not

The amount of provided useful, clear, and easy to read

information

Access convenience: customers' required actions to request service and, if necessary, be available to receive it

The availability of the service provider, the accessibility of the service provider through various ways, the time it takes to take a certain action in the service process, and the easiness to contact an employee of the service provider

Transaction convenience: the amount of effort customers have to take to realize a transaction

The easiness of the checkout process and the availability of different payment methods

Control The customer’s competence, superiority, and power within the sales relationship

The social connection with the service provider (-), the importance of the service (+), gender (W+), the customer’s experience with the service (-) and the customer’s personal mood (-)

Behavioural control: the availability of a response which may directly influence or modify the characteristics of a situation

The customer’s perception of what he or she can do to influence the situation or event

Cognitive control: someone’s ability to predict and reinterpret a certain situation

The right amount of information provided by a company during a certain situation

Decisional control: someone’s choice in the selection of outcomes or goals

The customer’s belief that he or she has the decision to enter a situation or has perceived alternatives

(23)

18

2.5 Annoyance

Bell & Buchner (2018) stated annoyance as the distraction from important or pleasant activities,

or even the entire disruption of these activities. The issue of annoyance has gained more

attention in the digital age due to the growing amount of evidence that digital experiences could

lead to significant annoyance for customers. In their research about annoying advertising, Bell

& Buchner (2018) found that even when an online experience was being perceived as annoying, this could still lead to the customer taking a decision which is in the firm’s benefit. Therefore,

it is important to understand what are the sources of customers’ annoyance in a digital

environment, and how those could be used by companies to perceive a certain goal (Ghose,

Singh & Todri, 2017). For example, Tan & Chan (2015) found that annoyance for online

customers may disrupt their information processing, which would lead to poorer decisions.

Companies could use this consequence to nudge customers towards a certain position that the

company wants. Another approach, less potentially harmful to the firm, could be that annoyance

sets up a certain way of thinking. A certain amount of annoyance could result in awareness,

leading to interest, and eventually resulting in acceptance (Ghose et al., 2017). This would be a

highly effective way for companies to nudge their online customers towards a certain direction.

These consequences of annoyance could be summarized in a phenomenon which is called

reactance.

2.5.1 Reactance

Reactance is a social psychological theory that explains how people react to the perceived loss

of freedom in a situation. When a person's freedom is threatened, that person will attempt to

regain the freedom by showing opposing or resisting pressures to conform (Edwards, Li & Lee,

2002).

Brehm (1966) was the first researcher to look into this phenomenon. He explained that

reactance is most likely to occur when (1) the threatened behaviour is important, (2) the

(24)

19 the person ever actually enjoyed the freedom.

Reactance can result in different conformations. On the one hand, the effect can

encourage a desire to engage in the threatened behaviour even more strongly (rebellion), on the

other hand, the effect can result in an attitude change in the person's belief that the activity is

important (acquiescence) (Edwards et al., 2002). For example, the process of changing an

internet subscription online could be very difficult due to complex procedures and not having the ability to obtain a subscription which meets the customer’s requirements. This could

frustrate customers, and may even lead to hostile customers (Walker et al., 2002). Out of

rebellion, these customers might than decide to cancel the subscription entirely, and sign a

contract with a new internet provider. For this reason, it is useful for service companies to know

what might be sources of annoyance for their customers.

2.5.2 Sources of annoyance in a digital environment

Smith (2011) did research into digital marketing strategies that millennials find annoying. She

found that customers do not like messages that are distracting, disturbing, forced, or interfere

with their work. Specifically, customers are most irritated by pop-ups on websites, mandatory

downloads, and un-closable windows. Li, Edwards & Lee (2002) researched the notion of

annoyance of online advertisements and found that these advertisements are perceived as

annoying if their content is untrue, exaggerated, confusing, or insults the viewer's intelligence.

Although most of these researches into annoyance has been done in the context of digital

advertisements, it is assumed that these sources of annoyance are representative for other

contexts in the digital environment as well. For example, many researchers found that the

differences in the choice architecture on websites are often misleading and annoying. In

specific, a subject which is getting more and more attention in the last years is the notion of ‘dark patterns’.

Dark patterns are tricks or manipulations used in websites and apps that make customers

(25)

20 every word on every page, but they skim read and make assumptions. Companies can take

advantage of this by making a webpage or online decision look like it is saying one thing when

it is in fact saying another (Brignull, n.d.). An example of a dark pattern is the ‘sneak into basket’ technique, where a retailer automatically adds products (like a magazine subscription

or travel insurance) to customers’ shopping carts and makes it hard for them to remove the

unwanted items. Another example is ‘misdirection’, in which a certain website design may

distract customers from seeing that checkboxes are set as default, which automatically signs

them up for a newsletter or membership, spam their contacts or change their home pages.

In this research, a combination of the traditional annoying advertisement techniques and

these more novel dark patterns are being covered. In specific, the following annoying stimuli

are treated; pop-ups, confirmshaming/steering questions, and content that insults the customer’s

intelligence. These three stimuli are chosen because they are prominently used by online service

companies and are factors that influence the choice architecture of decisions for convenience

or control.

Websites where the above-mentioned factors are prominently visible are travel booking

websites. Many of these websites are designed in such a way that tries to nudge the customer

in any decision within the customer journey. Next, each specific annoying stimulus is further

explained with examples within the online travel booking industry.

Pop-ups

Pop-ups are one of the popular techniques to deliver important information or information by

which a company tries to nudge a customer’s decision on a website. Edwards et al. (2002) refer

to pop-ups as a form of rich media advertisements that automatically launch in a new browser

window when a web page is loaded.

(26)

21 creating a sense of urgency which would stimulate customers to finish the booking process

quickly. Examples of the content of these pop-ups are:

 ‘In high demand - only 4 rooms left on our site!’

 ‘33 other people are looking this accommodation right now, so finish your booking

quickly’

 ‘Last chance! Only 1 room left on our site!’

In many cases, these pop-ups appear a few seconds after entering the page, making it seem like

the message is a real-time notification. This effect is reinforced by an icon of an alarm clock.

Besides creating an urgency, the messages that someone else just booked a certain

accommodation may also reassure a hesitating customer that the property must be good.

People perceive pop-ups as annoying because they interrupt their online task and

interfere with their service process (Li et al, 2002).

This leads to the first hypothesis:

H1a: Interfering a customer journey with pop-ups will lead to a higher perceived annoyance than no interference of pop-ups during a customer journey

Confirmshaming/steering questions

Confirmshaming is the act of making the customer feel guilty into opting into something

(Confirmshaming, n.d.). This phenomenon is most visible in creating a question which presents

the customer with two options. Instead of providing a simple 'yes' or 'no', the customer is instead

presented with one option that is stated as desirable, and one option that is stated as undesirable.

The goal is to steer and shame the customer into going for the option which is most beneficial

for the company, by making the alternative seem undesirable by presenting it in shame-inducing

(27)

22 customer is not willing to activate the ‘instant book’ option is a good example of this

phenomenon. One consequence of these steering questions could be that customers are

delighted to be facilitated by the company in this way. Because of the ‘help’ of the company,

customers could know better what do to and where to go, which could bring them in a flow. Being in a flow or ‘in the zone’ is being in a positive mental state of mind because someone is

enjoying overcoming challenges (Schaffer, 2013). However, these suggestive questions and

answers are more likely to feel like a threat to the customer’s freedom and therefore provoke a

perception of annoyance.

H1b: Influencing a customer journey with steering questions will lead to a higher perceived annoyance than no influence of steering questions during a customer journey

Content that insults the customer's intelligence

Many companies are trying to nudge customers by showing them content of which the customer

knows it is untrue or explaining the customer something in many words which could have been

said in way fewer words. This way of presenting a choice architecture could lead to a high

decision inconvenience. The goal of companies is that by presenting customers with a lot of

(useless) information, they hope that it will decrease the amount of (important) content a

customer will read later on (Kirsh, 2000). This is because people are less willing to process

additional information after they read much useless information, and this limits the amount of

information that they process and understand (McCoy et al., 2007; Tan & Chan, 2015). A

common example on travel websites of presenting content of which the customer knows it is untrue is ‘This is the cheapest price you’ve seen for the dates you selected!’. However, these

untruthful and exaggerated messages and ‘information overload’ are experienced as a big

(28)

23

H1c: Influencing the customer journey with content that insults the viewer's intelligence will lead to a higher perceived annoyance as compared with no influence of the customer journey

2.5.3 Influence of annoyance on the decision for convenience or control

Derived from the literature review, arguments can be made that annoyed customers would

choose for more convenience, but there are also reasons to say that annoyed customers would

choose for more control.

The reason that a certain amount of annoyance could lead to customers choosing for

more convenience is that according to researchers in cognitive psychology, when people are

interrupted from an online task, they react negatively to the need to expand mental effort to

process additional information. The interruption also disturbs the person’s attention and limits

the amount of information that is received and understood (McCoy et al., 2007; Tan & Chan,

2015). Therefore, the customer might want to obtain a service with the least amount of effort.

This would result in a higher convenience orientation:

H2a: Annoyance will result in a higher convenience orientation

On the other side, it is also possible that customers would choose to take more control over a

situation after a certain amount of perceived annoyance. Annoyance as a result of complex

technologies that do not make the customer experience easier may lead to hostile customers and

reactance (Walker et al, 2002). These website design features could feel like a threat to the

customer's freedom, which he or she might try to restore by showing rebellion and thus choosing

for options which give them a way to restore the control over the situation (Edwards et al.,

2002). Next to that, perceived annoyance may worsen someone’s mood, which makes it less

likely that this person gives away control over a situation (Ball & Callaghan, 2012). These

theories lead to hypothesis 2b:

(29)

24 This means that because of website design features that provoke perceived annoyance, it is

expected that people will react differently in regard to their decision for convenience or control.

However, it is also important to test whether the website design features itself have an impact on people’s decision for convenience or control. It could be that people are influenced by

steering questions or pop-ups without perceiving any annoyance, for instant by getting in a

flow. For that reason, the following hypotheses are tested:

H3a: Interfering the customer journey with pop-ups will result in a higher convenience orientation (as compared with no interference)

H3b: Influencing the customer journey with steering questions will result in a higher convenience orientation (as compared with no influence)

H3c: Influencing the customer journey with content that insults the viewer's intelligence will result in a higher convenience orientation (as compared with no influence)

H3d: Interfering the customer journey with pop-ups will result in a higher desire for control (as compared with no interference)

H3e: Influencing the customer journey with steering questions will result in a higher desire for control (as compared with no influence)

H3f: Influencing the customer journey with content that insults the viewer's intelligence will result in a higher desire for control (as compared with no influence)

Since it is argued that annoyance is one of the main reasons why people would tend to choose

for convenience or control, it is expected that annoyance will act as a mediator within any of

the relationships between the website design features and the decisions for convenience or

(30)

25

H4: Annoyance has a mediating effect on the relationship between website design features and someone’s decision for convenience or control

2.5.4 Time

Interesting is if the amount of time that a customer has plays a role as well in the amount of

annoyance that someone perceives and in the decision for convenience or control. It is

imaginable that a customer who is under time pressure gets even more annoyed by any

interruptions of a task. This is confirmed by Baethge & Rigotti (2013) who found that if time

pressure goes beyond a certain level, it has negative effects on performance and increases

annoyance. Li et al. (2002) mention that when people are under time constraints and faced with

ads that interrupt a given task, customers may feel greater annoyance than when not under time

pressure. A reason for this might be that interruptions of a certain task might lead to an even

higher time pressure and higher mental demands, which leads to increasing annoyance (Baethge

& Rigotti, 2013). To test this, the following hypothesis has been stated:

H5: The amount of time that someone has will moderate the effect of the relationship between website design features and perceived annoyance such that under a time constraint people will perceive more annoyance, and under no time constraint people will perceive less annoyance

Furthermore, as mentioned before, time scarcity may also have a direct effect on a customer’s

decision for convenience or control. It turns out that when individuals experience high levels

of time scarcity, they are likely to have certain ways of thinking about time that may reinforce

their experience of time shortage and which may then induce a preference towards convenience

(Farquhar & Rowley, 2009). Quite some researchers hypothesized a direct relationship between

a lack of time and a desire for convenience, but consistent results are still leaking. For this

(31)

26

H6: Having a time restriction will result in a higher convenience orientation compared to having no time restriction

2.5.5 Conceptual model H4 H6 H5 H2 H1 Website Design Features Annoyance Convenience/control Time H3

(32)

27

3. Research method

3.1 Research design and procedure

This study used an online survey experiment in the context of the online travel industry. This

industry is chosen since many people are well known with online travel websites and these

websites are usually characterized by the earlier explained potentially annoying design features.

Two surveys with four different conditions have been used, which were randomly

divided to a participant who is part of one of the above eight groups. Due to the survey design,

it was impossible to include the conditions with a time constraint and the one without a time

constraint in the same survey. Therefore two different surveys had to be distributed to place

participants in the condition with a time constraint or the condition without a time constraint.

For all conditions, the participants were presented with different simulations of a digital

customer journey in which they had to imagine that they were obtaining an accommodation for

their travel destination. During this customer journey, the participants had to make a choice

between a convenience option or a control option for six different topics. Topics covered in this

customer journey were;

 the request for the acceptance of cookies (access convenience vs decisional control)  the way of searching through the database (decision convenience vs cognitive control)  the way of use of filters (decision convenience vs cognitive control)

 adjustment of booking conditions (access convenience vs behavioural control)  addition of a travel insurance (access convenience vs cognitive control)

 the way of recovery of a service failure (access convenience vs decisional control) These topics are chosen since these are common topics in a customer journey and often consist

(33)

28 Different pages of a website for a fictitious online travel accommodation company, named

Codex.com, were designed to eliminate the effects of prior experience and brand effects. After

completing the questions in the customer journey, participants had to answer some questions

about the level of annoyance they might have experienced and about some personality traits

related to their usual (online) behaviour. The experiment had a 4x2 between-subjects design:

The survey was distributed over the internet, using the program Qualtrics. For practical reasons

and because of the limited amount of time for this research there is chosen to simulate a

customer journey instead of using an actual customer journey on an existing service website.

The data is collected within a period of three weeks, whereafter it is analysed. The participants participated anonymously in the research and three Amazon gift cards worth of €25 have been

raffled among all the participants who completed the survey and left their e-mail address.

3.2 Sample

The sample consists of participants that would be most representative of customers of online

service companies. Nowadays almost everyone is confronted with online customer journeys

and since it is valuable to see potential differences between older and younger customers, the

sample contains participants ranging from 16 to 62 years old. These participants are assembled

by social media, e-mail, worth of mouth and through snowball sampling. Out of the 246 people

that started the survey, 183 finished the survey completely, which equals a completion rate of Customer journey No time constraint Time constraint

Control group (not influenced by any potentially annoying website design features)

A B

Interfered with pop-ups C D

Influenced by steering questions E F

Influenced by content that insults the viewer's intelligence

(34)

29 74.4%. The sample size of participants in the condition with a time restriction was considerably

lower than the sample size of the participants without a time restriction (48 vs 130). This is due

to two reasons. The first reason has to do with the operationalization of the time restriction,

which resulted in many missing responses because of the automatically forwarding to the next

question when the countdown reached zero. The second reason is that the survey design limited

me to distribute two different surveys, one with a time restriction and one without a time

restriction. This made it difficult to get evenly distributed sample sizes. However, this only

affects the reliability of the hypotheses with the time variable, for the other hypotheses, there

were enough respondents per condition to detect medium-sized effects (Bonett, 2002). The

mean age of the participants was 26.7 years (SD = 9.9). Around 80% of the participants were

between 20 and 30 years old. 50.6% of the respondents were male. 61.2% of the participants

were from the Netherlands, the other participants were from mostly European countries. All

demographics can be found in appendix 1.

3.3 Measurements

3.3.1 Independent variables

Website design features: Operationalized by interfering the customer journey (or the

welcoming screen with the instructions for the survey) with (0) no treatment, (1) pop-ups, (2)

confirmshaming/steering questions, and (3) content that insults the viewer's intelligence (Li et

al., 2002; Smith, 2011).

(0) Control group

Participants in this condition were not influenced by any potentially annoying website design

features. In this way, the effect(s) of the three different potentially annoying website design

(35)

30

(1) Pop-ups

Pop-ups have been operationalized by interfering the customer journey in three different

moments, with pop-ups that are similar to the ones earlier explained in the theory. After two

questions, a screen appeared with a pop-up presenting that a certain amount of people were

booking the property at that right time and that the property was just booked. After another

question, a pop-up presented that the property was in high demand and that there were only

limited rooms left. Examples of these pop-ups can be found in appendix 6.

(2) Confirmshaming/steering questions

This feature influenced the customer journey in two different moments. In two questions,

participants were steered into taking the convenience option or the control option by strongly

advising and giving reasons to choose this specific option instead of the other option. If

participants still chose the non-recommended option, a second confirmation question was asked

in which they could choose to still go for the recommended option, or stay with their choice for

the non-recommended option. For one question the convenience option was presented in a very

desirable way whereas the control option was presented in shame-inducing language. For the

other question, this was the other way around. An example of this confirmation question is:

Are you sure you don’t want to make use of the opportunity to be secured for many potential

problems by choosing for our travel insurance?

- No, please add the travel insurance

- Yes, I would like to continue without a travel insurance and risk a lot of costs

(36)

31

(3) Content that insults the viewer's intelligence

This stimulus has been implemented in the customer journey in three moments. Participants in

this condition were facing long explanations about customer journeys and cookies, which were

not necessary to explain or could have been explained in fewer words. Also, these participants

were faced with the following banner, just before they would proceed to the booking conditions

page:

Today is your lucky day! You chose the cheapest room. Don’t miss out on our best price - Book

now!

Time:

Time has been operationalized by two conditions:

(0) High

These participants were not forced to complete the customer journey in a given time.

(1) Low

To these participants a time restriction was given which forced them to complete the customer

journey in a given time. For every question in the customer journey, a countdown was placed

on top of the screen which showed the participants how many seconds they had to answer the

question. The participants were told before, that if they would not answer the question in the

given time, they would automatically proceed to the next question. To determine how much

time participants were given for each question, a small pre-test has been done in which was

recorded how long it took participants to answer every question. The average response time

was added up with five seconds, to make sure that all the participants would have enough time

(37)

32 perception (Hornik, 1984). The time given to respondents in this condition differed from 29

seconds to 55 seconds per question.

3.3.2 Dependent variables

Perceived annoyance:

Perceived annoyance is measured after completing the customer journey, by a simplified scale

designed by Li et al. (2002). Perceived annoyance was measured by five questions which had

to be answered on a 7-point Likert scale ranging from strongly disagree to strongly agree. The

question was; during the customer journey I felt: distracted; disturbed; forced; interfered;

obtrusive. This variable is measured by the average of the five answers (Low perceived

annoyance = 1, high perceived annoyance = 7). Two of the five items were phrased as

counter-indicative to be able to detect acquiescence bias.

Convenience versus control:

In the current research, there are no scales developed to measure the decision for convenience

or control as a dependent variable in a reliable way. Since a validated scale is missing, this

variable is operationalized in a way which is backed up by the existing theory about

convenience and control. The variable is operationalized by the simulation of a customer

journey of the search for an accommodation for a travel destination. In six different options,

participants had to choose for either convenience or control. The convenience and control

options were as follows:

(0) Convenience orientation: The option which reduced customers' effort in obtaining

the service. An example is obtaining the service while searching through just a part of the whole

database, which is determined by a personalized algorithm, instead of searching through the

Trading off convenience versus control : a study on the impact of annoyance and time pressure on someone’s decisions for convenience or control

Master Thesis