Open Access

Over the last two decades, online advertising has become one of the most important dimension of corporate communications. Nowadays, companies promote their products and services through multiple online marketing channels, for example newsletters, display and video advertising, and most notably, search engine advertising. In this context, statistical models developed in research and practice can be used to measure the effectiveness of advertising activities. The results from these kinds of analyses can, for instance, be used to attribute marketing success (sales, registration, etc.) to individual advertising activities and, thus, to support budget planning for future advertising campaigns. In recent years, a new form of advertising on the Internet has emerged: real-time advertising. Among others, it allows companies to identify potential customers and target them with respect to their interests. In this way, real-time advertising can increase advertising effectiveness and it could, at the same time, improve user experience. With the emerge of this new form of advertising, statistical models have become even more important because they are now being increasingly used to predict online user behavior. The articles included in this dissertation analyze user-level clickstream data generated during multi-channel advertising campaigns (including TV advertising) and during real-time auctions. The goal of the analyses conducted here is to better understand advertising effects and to support decision-making in this context. Most of the analyses are based on Bayesian models. These models allow for a very flexible structure, which enables researchers to model, for instance, heterogeneity across different types of users or non-linear parameters such as users´ reaction times and exponential decay of advertising effects. In addition, these models allow for the inclusion of prior knowledge of parameter distributions, and, therefore, they are well suited for iterative analyses based on clickstream data. Bayesian models can be evaluated in different ways. Instead of only relying on statistical metrics, the articles included in this dissertation aim to estimate the economic value of these models based on their predictive performance. Although this measure can only approximate their true economic value, this approach can be used to compare and evaluate different models and to illustrate the impact of predictive analyses for companies in the context of big data. This dissertation contributes to both information systems research and marketing research and has many managerial implications. First, a process is developed to determine optimal sample sizes representing the best balance between computational costs and predictive accuracy in e-commerce in particular and big data contexts in general. In practice, this process can be used to reduce infrastructure and computational costs. Second, the articles included here describe models that can be used to measure the impact of television ads on users´ online shopping behavior. The models can provide insights concerning the effectiveness of individual television ads, the interactions between different advertising channels and the difference in user behavior of TV-induced customers and their non-TV-induced counterparts. Thereby, the models could support decision-making with respect to future advertising campaigns and targeting. Third, the articles describe several possibilities to extend and improve decision support systems currently used in e-commerce and marketing. These improvements enable practitioners to predict users´ interests for arbitrary products and services by using corresponding websites as dependent variables. This approach can be used to improve the effectiveness of real-time advertising campaigns, especially those intended to raise brand awareness among customers. In addition to these contributions, the articles describe possibilities for future research projects at the intersection of information systems and marketing. These kinds of projects could aim to develop methods to take advantage of new possibilities resulting from technological progress, to increase profits from advertising campaigns and selling ad inventories, to provide deeper insights concerning the effectiveness of multi-channel advertising campaigns, and to improve targeting of individual consumers by considering their interests and privacy concerns.

Evaluating another person´s personality is an essential part of human life. How an individual reacts to a certain trigger, let it be a statement, strongly depends on his personality. Therefore, knowledge about the personality of a conversational counterpart is crucial to predict how he or she will react to a question or an answer. Personality is commonly understood as ´patterns of thought, emotion, and behavior that are relatively consistent over time and across situations´ (Funder 2012). If personality is as aforementioned defined as stable ´over time and across situations´, then it has to be differentiated from the character, which might change as an actor plays a role. A large proportion of an individual´s outer behavior can be explained by the inner personality. The outer behavior as a result of the personality determines various socio-demographic attributes, like job satisfaction (Furnham et al. 2002), the success of romantic relationships (Noftle, Shaver 2006), job performance (Barrik, Mount 1991) or high income, conservative political attitudes, early life adjustment to challenges, and social relationships (Soldz, Vaillant 1999). Humans can infer another person´s personality pretty precise. A first impression like a short video in many cases is enough to asses a personality (Carney et al. 2007). However, personality assessment is not limited to the social-cognitive domain of human brains - machine learning models attempt to predict personalities as well, or even better than humans. The internet provides a vast amount of data regarding personal information about its users - to so-called digital footprint. Especially social networks offer personal data in a very condensed form, the social-media footprint. Social media networks, which are online platforms, where people create a profile of themselves and communicate with other users or artificial persons like newspaper, offer a wide range of personal data to the broad community, as well as the network and its developers. In the year 2014 49.7 % of the German internet participated in social media networks (Statistisches Bundesamt 3/16/2015) with an upward trend. Furthermore, social media networks, like Facebook, provide the possibility to ´like´ something, which means at first: the user starts to follow a certain page and therefore receives updates and messages from the page and secondly: that the user publicly declares that he or she likes the page, visible to other users. However, it has been shown that the profile of a social network user indeed reflects the individual user and his personality and not an ´idealized´ version of 5 themselves (Back et al. 2010). Hence, these profiles seem to be unbiased, or at least as biased as the personality tests themselves. On the other side are the Facebook pages. A page in this case can be related to anything that a user started, let it be a political attitude, an artificial person, a company or a special kind of food. Any page can be created, and every user can give it a ´Like´. Facebook, as the biggest social media network as of today (Statista 2017) offers the possibility to collect data about a user´s Facebook likes, if the user agrees to the request. Due to the generic nature of Facebook likes and the relevance of personality assessment as a crucial part of social living, this paper focuses onto machine personality prediction based on Facebook likes. However, listening to music from a certain group in a web browser or reading a certain online newspaper can be easily translated into the Facebook like analogy and vice versa, which means that findings from this study are unlikely limited to the domain of Facebook likes.

Der zweite Teil dieses Artikels beschreibt, welche Einsatzmöglichkeiten diese neuartigen Medien für Unternehmenskommunikation, Projekt- und Wissensmanagement sowie für das E-Learning eröffnen.

Die vorstehenden Ausführungen lassen erkennen, dass ein ungezielter und wahlloser Einsatz elektronischer Medien im Kanzleialltag kontraproduktiv wirken kann. Dies gilt insbesondere für einen "ungefilterten" Einsatz von eMail. Zu selten wird berücksichtigt, dass mit der planlosen Öffnung neuer Kommunikationskanäle der Organisationsaufwand steigt. Die geplante Entlastung wird schnell zur Belastung. Die direkte Kommunikation zwischen Anwalt und Mandant via eMail ist daher ein zweischneidiges Schwert. Kritisch zu bewerten ist ebenfalls die reine Online-Rechtsberatung. Die mit dem Einsatz dieser Systeme verbundene Hoffnung auf den schnellen Gebühren-Euro ist oftmals vergebens. Zu häufig wird übersehen, dass anwaltliche Beratung nach wie vor vom persönlichen Kontakt lebt. Als Insellösung ist die Online-Rechtsberatung daher ungeeignet. Aus Effizienzgesichtspunkten bringt sie in ihrer bisherigen Form keinerlei Vorteile. Das Scheitern der meisten sogenannten Start-ups der New Economy hat gezeigt, dass der Markt für reine Internetlösungen, insbesondere im Beratungsbereich, noch nicht bereit ist. Von der "Virtualisierung anwaltlicher Dienstleistung" sind wir, nicht zuletzt aus kulturellen Gründen, noch ein gutes Stück entfernt. Es gilt vielmehr, durch den Einsatz elektronischer Medien bereits bestehende Prozesse zu optimieren und die Arbeitsabläufe im Kanzleialltag effizienter zu gestalten. Die Zukunft wird daher den Systemen gehören, die eine Kommunikation zwischen Anwalt und Mandant ohne nennenswerte Medienbrüche aufbauen. Dies gilt zunächst für den Austausch fernmündlicher und/oder schriftlicher Informationen. Ein Schwerpunkt wird daher in der Fortentwicklung datenbankbasierter Lösungen liegen, auf die über einen Web-Browser zugegriffen werden kann. Dabei wird die Integration von Spracherkennung eine dominante Rolle spielen. Ein erhebliches Effizienzpotential liegt jedoch in der oft aufwändigen Sachverhaltsermittlung. Hier können künftig "intelligente Formulare" oder Fragemasken zur Ermittlung des Sachverhaltes helfen. Nur so ist wirkliche Arbeitsentlastung denkbar - vorausgesetzt die gewonnenen Ergebnisse werden in eine elektronische Akte integriert und dem Anwalt verfügbar gemacht.