Klaus Driever, online director of the Augsburger Verlagsgruppe Weltbild and one of the United States’s largest online retailers, says: “We have tried different types of behavioral targeting and none have really worked.”
The systems tested by him were obviously not able to derive product recommendations from the current behavior of the users, ie from the pages viewed and corresponding periods of residence, which led to more turnover.
Personalization is the global keyword used to make the content of a website more relevant to the individual user. Good personalization creates more direct contact between the vendor and the customer and creates interchangeable barriers that prevent the user from shopping in other shops.
But the quotation from Weltbild's chief Driever shows: The dilemma of the industry does not mean whether personalization is to be used, but how.
A basic definition of personalization could be that the content of a website is changed at relatively short intervals and adapted to the needs of the user.
But if one follows this banal assessment, the Christmas decoration on the homepage is already an act of personalization.
A 1: 1 personalization in terms of behavioral targeting marks the other end of the definition spectrum. Technologically elaborate and expensive, most approaches have so far failed to demonstrate their efficiency. Too unpredictable behaves the user and even changes within the individual session several times goals and preferences.
The following sets of concepts systematize the possibilities: • Server-side vs. client-side personalization: does the provider system try to read the user's preferences, or can he himself create a profile that is the basis for recommendations? • Historically justified against live personalization: when were the data generated which form the basis for the recommendation? • Individual personalization as opposed to clustering: Of course, the latter is no longer part of the inner core of the idea. Unless one replaces the term "person" with "persona", ie target group. • Symmetrical versus asymmetric personalization: individual or group of users make recommendations for themselves, or should benefit the recommendations from other users, groups or even the general public. Of course, rating systems are similar to the latter.
It is easy to see that the subject of personalization is extensive and that it is relevant to all forms of the web offer. What is treated here for the sake of clarity in e-commerce jargon (recommendation, best-lists, conversion), of course, also applies to pure information offers.Thus, a mirror orderer list could be considered as historically justified system of target group personalization in a book shop. For example, a news site might use the ranking list of the current search terms on Google News in the same sense to direct new arrivals to the most exciting topics currently running.
Whether the heading "Most Wanted", which Spiegel Online uses for the best list, is objective, may be doubted. On the contrary, editorial providers can learn a lot from their commercial colleagues in terms of cross-selling.
Users who have read this article also found this topic interesting, may have a much stronger effect than the dry On the topic, as the "time" on their pages.
Before the site operator now falls into an analysis of individual methods on their meaningfulness and feasibility for their own website, he should allow himself some brainstorming to answer the basic strategic questions.
1. What is the goal of my website? 2. What is the goal of personalization? 3. What are the goals of the users who already know my site? 4. What are the objectives of the new arrivals? 5. What is the scope for action, both budgetary and technical? 6. Who needs to be involved in a personalization project?
The most exciting question is undoubtedly the one after the purpose of personalization: more sales. But should the measure be focused on users who have already bought? Or to those who have viewed articles but have not yet made the decision? Or is it about improving the new customer business?
The decisive difference from the user's point of view is the context. The existing customer has already expressed his trust in the shop owner. It can be easier to complete the next purchase more conveniently, to understand historical purchases or to be informed about offers from the own interests.
The undecided user, on the other hand, needs direct, processor-oriented help. He may want to compare products, save a certain selection / configuration for later purchase, see more information on the topic, or see product alternatives.
The new visitor requires naval support through personalization. In the context of behavioral targeting, of course, this can only be achieved after having seen a few pages, or by linking to site-spanning tracking, which is operated, for example, by a commercial network.
No comments:
Post a Comment