Targeting Using Augmented Data in Database Marketing

Decision Factors for Evaluating External Sources

verfasst von: Bettina Hüttenrauch, geb. Krämer

Verlag: Springer Fachmedien Wiesbaden

Enthalten in: Springer Professional "Wirtschaft+Technik" , Springer Professional "Wirtschaft"

Einloggen, um Zugang zu erhalten

Über dieses Buch

This study delivers insights on which external sources – e.g. website click behavior, surveys, or social media data – can and cannot be used for data augmentation. A case study is performed to test the suitability of different sources in order to create a generalized practical guide for data augmentation in marketing. Data augmentation is a beneficial tool for companies to use external data, if the internal data basis for targeting is not sufficient to reach the customers with the highest propensity. This study shows that augmenting data from feasible sources leads to significant conversion lifts.

Inhaltsverzeichnis

Frontmatter

Chapter 1. Introduction to data augmentation in marketing

Abstract

Although marketing specialists spend a considerable amount of time, money, and know-how on relevant marketing campaigns, everybody is confronted with more less than well personalized advertisements every day. Relevance in this context is defined by attracting the positive attention of the recipient to the content or offer. While it can take weeks, if not months, to create these campaigns, the customers receiving the offer only need seconds to decide whether an offer is relevant or not.

Bettina Hüttenrauch

Chapter 2. Strategic motivation for data augmentation

Abstract

When visiting a marketing conference, it is very likely that at some point the allegory of the “Tante-Emma-Laden” is mentioned. People in the middle of the 20th century shopped in such a small general store. “Tante Emma” stands for customer focused individual information.

Bettina Hüttenrauch

Chapter 3. Literature review on data augmentation

Abstract

The need to combine data from different sources is almost as old as the ability to store data itself. Some interesting information is collected at one place by one instance, other interesting information in another place by some other instance. If there is an identifier to serve as a key, joining tables is a matter of technology.

Bettina Hüttenrauch

Chapter 4. Methodological framework for data augmentation

Abstract

Data augmentation in database marketing is a special field of data fusion. No general framework for this kind of data augmentation has yet been established. The scientifically regarded and practically approached augmentations in marketing are limited to representative sources, which have convenient features comparably easy to handle. The theoretical contribution of our study is to establish a methodological framework for augmenting external sources.

Bettina Hüttenrauch

Chapter 5. Test design for evaluating the source characteristics

Abstract

From the variety of external sources available, the question arises which sources are suitable for data augmentation. There are formal requirements which need to be met, as described in the previous chapter. Beyond those requirements, it is desirable to know whether the augmented values are “good”.

Bettina Hüttenrauch

Chapter 6. Analysis of data augmentation KPIs

Abstract

In order to answer the research question of which sources are suitable for increasing conversion probabilities in direct marketing, we have developed a conceptual model and identified factors influencing the conversion probability. Accordingly, we have created a test design in order to test the hypotheses derived from the conceptual model. A case study is carried out, performing multiple data augmentations, while varying defined parameters.

Bettina Hüttenrauch

Chapter 7. Analysis of results and test of hypotheses

Abstract

The KPIs established in chapter 6 are used in this chapter to test the hypotheses stated in chapter 5.1.2. All of the tests are only possible in the case study setting, where the true target values, as well as the source data mechanisms, are known. We validate the source data mechanism antecedents (hypothesis 1).

Bettina Hüttenrauch

Chapter 8. Limitations of data augmentation and outlook

Abstract

Data augmentation sources are manifold and oftentimes easily available. The main question arising from these sources is whether using them for augmentation purposes will lead to an increase in information and a better basis for decision making. The decision for or against data augmentation is simplified with our guidelines.

Bettina Hüttenrauch

Chapter 9. Summary

Abstract

External sources like web tracking, volunteer surveys, and social media are rarely used for data augmentation in database marketing yet, even though such information can lead to significant conversion probability lifts when used for targeting and personalization. Many external sources are suspected to lead to biased augmentation results, because the available data is not representative for the customer group to which it is augmented. With our study, we deliver insights on which sources can be used beneficially for data augmentation and which augmentation methods should be used for different kinds of sources.

Bettina Hüttenrauch

Backmatter

Titel: Targeting Using Augmented Data in Database Marketing
verfasst von: Bettina Hüttenrauch, geb. Krämer
Verlag: Springer Fachmedien Wiesbaden
Electronic ISBN: 978-3-658-14577-4
Print ISBN: 978-3-658-14576-7
DOI: https://doi.org/10.1007/978-3-658-14577-4