Website design: What kind of evaluation method should be used for different objects?

Intermediary transaction http://www.aliyun.com/zixun/aggregation/6858.html ">seo diagnose Taobao guest cloud host technology Hall

A variety of evaluation systems are often seen in social, application downloads, and electronic business websites. They are evaluated by different objects, such as the evaluation of watercress for music, books and movies, Taobao evaluation is for customers to buy goods, pea pod Evaluation is for users to download the App. Generally, however, there are two main types of evaluation: scoring evaluation and two-dimensional evaluation (such as liking/dislike, top/step, approval/opposition, etc.).

So, what kind of evaluation method should be used for different types of evaluation objects?

This paper studies Taobao, Amazon, Beijing-east, poly beautiful products, the public comments, Zhongguancun, watercress, Youku, embarrassing encyclopedia, know, Google play and pea pod of the evaluation methods of these sites, the above-mentioned sites involved in the evaluation of the object includes: App, music, movies, books, video, electronic products, general merchandise , catering, answering, jokes.

The author finds that it can be analyzed from the aspects of the complexity of the object, the value of the object, the number of participants, the convenience of evaluation and the design intention of the evaluation system:

Evaluating the complexity of an object

The complexity of the evaluation object should be better understood, mainly the complexity of the evaluation standard, for example, to evaluate a joke in the encyclopedia. It is much easier to judge a movie than to evaluate a film (shown in Figure 1 and Figure 2 below), because for jokes, it's funny on the top, not funny on the step, relatively simple and direct evaluation criteria, But for the movie, involves the movie narrative way, the actor's acting, the story splendid degree and so on. Especially for Poly-Mei excellent products such vertical type of electric business, because the user needs more specific, users need to refer to the information more professional and meticulous, so for different types of products set a different evaluation dimension. As shown in Figure 3 and Figure 4 below, the evaluation dimension of an eye care product and a health food set by Poly-Mei excellent products is respectively.

Fig. 1 The result of a user's evaluation of a joke in the encyclopedia is "top" and "Pat".

Fig. 2 Evaluation results of a film by users in Watercress, including scoring and detailed distribution

Fig. 3 The evaluation result of an eye care product by users in Poly-Mei excellent products

Fig. 4 The evaluation result of a healthy food by the users in Poly-Mei excellent products

Obviously, the object with low complexity is suitable for two dimensional evaluation, and the object with high complexity is suitable for scoring evaluation. Because for the evaluators, the five-point system, the very system of evaluation has more space to accommodate the reviewers of the evaluation of the comprehensive impression of the object, to avoid the black-white evaluation results.

Value of the object of evaluation

The value of an evaluated object is usually reflected in its price and frequency of purchase. For example, you might think that the value of an App is less than the value of a book, because the vast majority of apps are free, even if they are not usually higher than the price of a book, the value of a general commodity, such as a commodity, is lower than the value of an electronic product, because the cycle of electronic products is generally longer, So the purchase frequency is also low.

High value things, that is, high prices and low frequency of things, users will be more willing to spend time and energy to carry out evaluation, so the form of evaluation can be more diverse, including scoring, multidimensional grade or even text evaluation. Therefore, the evaluation of the APP in Pea pod uses two dimensional evaluation method, while Amazon, Jingdong and Zhongguancun have adopted the scoring evaluation method (shown in the following figure).

Fig. 5 The results of a user's evaluation of an App in pea pods

Figure 6 The results of a user's evaluation of a book in Amazon

Fig. 7 The results of the user's evaluation of an electronic product in Jingdong Mall

Fig. 8 The results of the user's evaluation of an electronic product in Zhongguancun

Number of participants in the evaluation

Generally speaking, the distinction and reliability of the two-dimensional evaluation results are better than the scoring evaluation in the same way, especially when the number of people is small. Because scoring methods produce multiple levels of evaluation, and some reviewers choose intermediate scores (such as "3 points" in a five-point system), more evaluation is needed to produce a reliable result with a differentiated degree. For example, the results of the evaluation shown in Figure 9 below: Using the five-point evaluation method, the APP's average is divided into 3.6 points. Whether from the distribution map or the average score, the distinction is not high. If we use two-dimensional evaluation method, assuming that 5 points and 4 points to the user when the choice of "like", playing 2 points and 1 points of users choose "Do not like", then the evaluation results are 120 people like, 46 people do not like, obviously more differentiated degrees.

Figure 9 The results of a user's evaluation of an APP in Google play

Therefore, when the number of participants in the evaluation is small, or when the product is on the line, two dimensional evaluation can be considered, on the one hand, it can reduce the difficulty of evaluation, encourage more users to participate in the evaluation, on the other hand, can more accurately reflect the evaluation results in the case of fewer participants

Evaluate the ease of operation

In the interface to show "like" and "do not like" two buttons, should be compared to present a group of Gray five-pointed stars and prompts the user to score to be intuitive. In particular, it is restricted by the interaction characteristics of mobile terminals to reduce the difficulty of user evaluation, and to improve the convenience of users to evaluate the operation, which is helpful to encourage users to evaluate. Like pea pods. This kind of mobile app downloads products that users use mainly on their phones, so it uses a two-dimensional evaluation method like "likes" and "dislikes" (as shown in Figure 5 above).

The design intention of the evaluation system

If the evaluation system is designed to lead users to download or purchase behavior, then we can consider avoiding the use of two-dimensional evaluation methods. Because the two-dimensional evaluation of the "step", "do not like" and other words will allow users to subconsciously create a psychological resistance, guide users not to download or purchase, and scoring evaluation of the "good" and "bad" the boundary is relatively vague, the user's psychological impact is not obvious, users can even five points in the "3 points" are counted as positive evaluation.

At the same time, multi-level and multi-dimensional scoring evaluation method can also provide more detailed and comprehensive and accurate statistics, such as Taobao for goods, services and logistics three points evaluation (as shown in Figure 10 below), can collect these three aspects of Taobao data, and statistical analysis, And then dig for greater value.

Fig. 10 evaluation results of the three aspects of Taobao for goods, services and logistics