Purchasing decisions: with help instead of alone?

July 2013

Imagine you want to book a hotel for a vacation in a town you have never visited before. The hotel brochure boasts a magnificent pool area and the photographs display a splendid buffet with lots of choice. One problem: the photographer has shrewdly cut out the motorway and building site right next to the resort. How can such a disappointing surprise be avoided? In the past, consumers almost exclusively relied on the recommendations of travel agencies and travel guides. Every now and then a friend might have been able to offer a specific tip. Today, the internet is an invaluable source of information. On sites such as Tripadvisor and Trivago, holidaymakers can read up on hotels and see the ratings given by previous guests. Their descriptions and reviews provide an initial impression of the pros and cons for other visitors.

In recent years, such reviews have received increasing attention due to being the focus of research. While once the only source for advice aside from friends and acquaintances was professionals, the internet has now made it possible to access recommendations compiled by other reviewers not working in the business who are most likely strangers. These social reviews have the potential to exert great influence on the purchasing decision of consumers. However, there has been little research into the actual degree of influence until now. To investigate, a research project was carried out by the GfK Verein in collaboration with the General Psychology: Cognition faculty at the University of Duisburg-Essen.

As part of the project, a group of 263 predominantly young, well-educated volunteers participated in experiments in two product categories. The first comprised “search products”, which can predominantly be effectively described using typical characteristics. Specifically, mobile phones were chosen because, with the exception of haptics, they can be defined through formal criteria such as display size, memory and other features. The second category encompassed “experience products” which consumers ideally need to use themselves in order to give evaluations and recommendations. For example, whether a bottle of wine would be good can only partly be determined through defined product characteristics such as vintage and grape variety. Hotel stays were selected for the experience category experiment in this project as they are also difficult to describe on the basis of formal characteristics.

Four recommendation sources: who helps decide?

Participants were not simply asked questions but instead completed a choice task (Choice Based Conjoint Analysis, CBCA) picking between different mobile phone models and hotels for a vacation on computers. The products/offers were not described in much detail as participants were to consider them equal on the topic of price and formal characteristics. However, there was variation with regard to the (randomly generated) reviews from four different sources: one comprised evaluations from an expert who was personally known to the participant and one from someone they were close to personally who did not have any knowledge of the product category from which they were choosing (personally known non-expert).

The other two sources of recommendations were not people of flesh and blood, but information found on different media. One was “online buzz”, so ratings on a travel portal and online shopping page for mobile phones respectively. These two sources represented the online community in the experiment. The other (specialist source) included professional reviews in travel guides (hotel bookings) and product test magazines (mobile phones). Before the task, participants had already selected media from a range of options (e.g. DuMont, Marco Polo, Stiftung Warentest, Connect, Amazon.de, Ciao.de, Trivago.de) and evaluated them as trustworthy and useful. One after the other, participants were asked to choose between many different options that only differed with regard to the configuration of positive and negative reviews from the four information sources (c.f. image). In this way, the power of influence of each information source could be determined in the analysis.

Competent acquaintance: strongest influence

Experts respondents knew personally quite clearly exerted the greatest influence on choice of product with their recommendations. If the importance weighting of all four sources as a whole equates to 100%, personally known experts account for 34% of purchasing decisions for mobile phones and 37% of hotel selections. The rating of specialist sources was identified as the second most important information source for mobile phones. In contrast, for hotel bookings, reviews in online communities (26%) were more important than specialist sources (24%). The influence of online social networks therefore evidently carries some weight for experience products. However, comments in communities play a lesser role for mobile phones (23%) when compared with reviews from specialist sources such as Connect and Stiftung Warentest (32%).

What is it that makes personally known experts so popular with consumers? The explanation is most likely the combination of knowledge and trust volunteers place in this information source. Along with their expertise, the personal relationship is an added bonus. However, another reason is also applicable: professional evaluations can only take into account and rate a limited number of details. Consequently, specific information that is important for individuals might be lacking, particularly when it comes to experience products. After all, the travel guide cannot know what value readers place on the choice of vegetarian food, natural light in the bathroom or availability of their preferred beer in the hotel bar, for example. In contrast, somebody who knows them well is able to offer the information that might be particularly important to them in their recommendation and therefore cut out a considerable amount of hassle in the decision-making process. When reading a review by a specialist source, consumers need to sift through various criteria and then also decide whether they have the same priorities and requirements.

Knowledge more important than connection

While expertise without a personal relationship plays a particularly big role in the purchasing decision process for search products, the reverse is by no means true. The advice of personally known non-experts carried no great weight, and this was true for both product groups. Only 11% of mobile phone and 14% of hotel decisions were attributable to the recommendation of someone respondents were close to but who had little knowledge in the particular area. Expertise is therefore considerably more important than a personal relationship.

In the second stage of the research project, respondents were shown a further four product attributes in addition to the four recommendation sources. For mobile phones, these were brand, shape, camera and media player features. Hotel bookings were more closely defined through hotel type, number of stars, bathroom and catering type. In a preliminary study with 100 participants, these characteristics were mentioned most as decision-making criteria and therefore selected accordingly.

Product information: lower influence

The surprising result of this test is that consumers base their decisions on recommendations to a much greater extent than they do on objective features and characteristics. This was true for both the hotel booking experience product and the mobile phone search product. For mobile phones, the influence of product features was a little greater by comparison, but it still only accounted for one third of purchase decisions (compared with two thirds for recommendations). However, the influence of recommendations was also dependent on the knowledge participants had of the product category: the greater the knowledge, the more attention they paid to objective product attributes and the less note they took of recommendations (c.f. image 3).

This phenomenon appears in writing on the subject under terms such as own-based and other-based decision-making processes (e.g. King & Balasubramanian, 1994). If a decision is own-based, consumers must know all the product characteristics, their potential effect and the use they will ultimately have for them. Those who rely on a recommendation above all save time by acquiring the knowledge from other sources, which makes it easier to come to a decision. Then they can focus on the actual intention of this process: relaxing by the hotel with a glass of delicious wine enjoying the sunset.


Data source: University of Duisburg-Essen and GfK Verein, Mutual research project

If you have any queries concerning this article, please contact Claudia Gaspar from GfK Verein (Tel. +49 911 395-2624, E-Mail: claudia.gaspar@gfk-verein.org) or Jens Hofman from University of Duisburg-Essen (E-Mail: Jens.Hofmann@uni-due.de).