User Loyalty Articles

The second survey was conducted three weeks after the first survey.

The second-survey questions about the loyalty of users were derived from indices used in the industry (Heater, pers. com. Feb. 19, 2005; Stratigos, pers. com. Feb. 19, 2005). To understand how efficiently the personalisation system works one has to measure the accuracy and the coverage (Mobasher et al. 2001)(Sheperd et al. 2001). Accuracy measures how precisely the personalisation system can deliver content that the user likes and coverage measures how likely the user is to accept the content.

Questions about the credibility of the site were important because without credibility a website cannot ‘persuade users to change their attitudes or behaviours’ (Fogg 2003, p. 148).

Unlike the first survey, the second survey was not promoted, except on the site itself and through emails to members. This was done because this survey was designed for long-term users of the site.

Data from the five most active B+users and the five most active B-users who completed the survey were evaluated, because this group of users experienced the site over a longer period of time. Every participant was a site member for at least 15 days, had at least seven visits to the site and read at least six news items distributed on different visits. During the time they were members the sum of all visits of the B+user group was 166 while the B-users visited 152 times.

In the first section of the survey participants had to rate aspects of every site in which personalisation features were included. The part of the survey in which users had to rate the tutorial was not evaluated because an insufficient number of users did the tutorial.

According to Shapiro (1999, p.44), companies recognise the individuals’ desire for control and start to give customers the ability to personalise their experience with the company. He argues that personalisation on the Internet does not only help customers to interact with companies or persons, furthermore it helps to filter useful information.

Cunningham (2001, p.118) describes the task of personalisation as ‘delivering the right content to those who need it when they need it’.

A study conducted by Karat et al (2003) identifies a function of two variables in the value of personalisation for customers and a function of two variables for the provider: For the customer it is, ‘the cost of divulging personal information and the perceived resulting benefits’, while it is, ‘the cost of gathering information and the perceived benefits’ for the provider. The benefits for providers are usually measurable, while ‘the customer’s value proposition is more complex and can involve factors such as security, privacy, trust, and the value of business relationships’.

According to Kasanoff (2003, p.121), value through personalisation can be provided by remembering information about a person and using this information to deliver unique benefits to that person. One way of delivering unique benefits is by providing content-based recommendations (Adomavicius et al., 2003). These recommendations are similar items to the ones the user preferred in the past.

The data for personalisation is derived from web mining.

The process of web mining for personalisation is divided into three steps, data acquisition, data analysis and data output (Markellou 2004; Albanese 2004).

In web mining, several fields of data are defined. User data is separated into explicit data, which is gathered with knowledge of the user through manual input, and implicit data, which is gathered without direct interaction with the user by utilising web-usage mining to record and accumulate data about user interactions and behaviour whenever a web server receives a request for resources (Zhu, 2004). Both of these forms of data have their drawbacks. Explicit data can be influenced by negative attitudes of the user and implicit data can raise privacy concerns and thereby lead to loss of trust (Scime 2004, p.27 ff; Schubert et al. 2000; Eirinaki et al.2003).

The data is scanned for patterns and rules about users’ navigational behaviour, user and page clusters and can also be combined with other data, such as data from a database with additional information. Discovered rules and patterns can be used for personalising a website or are integrated in a user profile for a different purpose. (Adomavicus et al., 2001)

Examples for these rules can be:

Content Rules to select, sort or modify the information on a website.

Navigation Rules to add, remove, activate or sort any links in the user navigation. Presentation Rules to modify the structure of the published website or acquisition rules that determine how data is collected (Garrigós et al. 2003; Eirinaki et al. 2003).

Adomavicius et al. (2003) describes three delivery methods for personalised information. The ‘pull’ method notifies the user that there is personalised information available but displays it only when the user requests it. The ‘push’ method sends the personalised information to the user and the ‘passive’ method provides the information along with other information without interaction with the user. He also describes de-personalisation, which is the status when a personalisation system is not producing valuable results for a customer any more and therefore they stop using it. The effectiveness of personalisation is a topic of an ongoing debate. (Business Wire 2000) One[EB2] opinion is that ‘You can’t reduce a person to a rule’ (Calvacca 2001; Kastner 2003) because data does not show the reasons why a person acted as they did. The Jupiter Research report, “Beyond the Personalization Myth (2003),” confirms the statement of usability guru Jakob Nielsen (1998) that ‘Web personalisation is much over-rated and mainly used as a poor excuse for not designing a navigable website’. The report also concludes that it is cost ineffective to operate a personalised website because it costs four times more than a normal website (Festa 2004).

The success of online retailer Amazon shows a different picture. Amazon, which is the 74th most valuable brand, according to Businessweek (2004), and has the highest rating of 88 in the American Customer Satisfaction index (Allen, 2004), relies heavily on personalisation. It has spent $800 million since it was founded in 1997 on technology and, according to Jeff Bezos ‘enabled products to find customers’ (Kohavi 2004; Bezos et al. 2002).

In a recent study about on-line customer experience it also had the highest ranking of 8.0 among all companies (Britt 2005). Furthermore, Amazon also has the most loyal customers of all on-line bookstores (Brand Keys 2004).

Scope of the research

The scope of the research was to measure the impact of dynamic personalisation features on the members’ level of loyalty to the site. Members were chosen because, according to the German “Online-Datenschutz-Prinzipien” (transl: Online-Privacy-Principles) a website provider has to inform a user if identifiable data about them is collected. This is done in the disclaimer of the site when a member registers.

Another aspect was that members were able to use all features of the site and therefore benefited from all personalisation features.

Because this research was based on an emotional definition of customer loyalty, the goal was to fulfil higher-level needs (Chak 2002, p.2). The effect of the personalisation features also benefits from that because higher-level needs are more personal and therefore they have a bigger impact on the user (Kasanoff 2001, pp.113-119).

The hierarchy of needs and the personalisation ladder both derive from Maslow’s hierarchy of needs (Huitt 2004) and can be merged therefore. The result shows that most of the levels of the personalisation ladder are already in the highest level of user needs (fig. 1)(Kasanoff 2001, p.113).

Figure 1. Framework for this study

That is the reason adjustments to the website, that were made during the development process, focussed on creating a desirable, useful, credible and valuable user experience. The aspects of user experience that a website must be usable, accessible and findable were only dealt with to the extent that they fulfil the expectations of the users (Moreville 2004).