Measuring Web Aesthetics and its Effects on Sales Performance

RESEARCH BACKGROUND

Initial impression influences the visitor’s subsequent attitudes and behavior
Good impression → visitor staying and performing transactions
Web aesthetics → positive initial impression
Seller reputation → sales performance and price premiums
Research Gap

​

​

​

​

​

​

​

​

​

​

​

​

RESEARCH GOALS

1. To identify and examine the concrete design factors that can affect web aesthetics. 

2. To study how web aesthetics can be operationalized in online marketplaces.

3. To investigate how the influence of web aesthetics on sales performance is contingent upon different levels of the seller’s reputation by using real-life data.

 

LITERATURE REVIEW

Aesthetics in various disciplines

​

 

 

 

 

 

 

 

​

​

Effects of Web Aesthetics

​

​

​

​

​

​

​

​

Cognitive Dissonance Theory and Confirmation Bias

​

​

​

​

​

​

​

​

​

RESEARCH MODEL AND HYPOTHESES

​

​

​

​

​

​

​

​

​

​

Hypotheses H1a, H1b:
High rep. → positive initial beliefs → look for positive attributes while ignoring negative attributes
H1a: For stores with high seller reputation, the impact of unity design on sales performance will be larger for stores with high level of unity than for stores with low level of unity.
Low rep. → negative initial beliefs → look for negative attributes while ignoring positive attributes
H1b: For stores with low seller reputation, the impact of unity design on sales performance will be larger for stores with low level of unity than for stores with high level of unity.

 

 

 

 

 

 

Hypotheses H2a, H2b:
According to research in experimental aesthetics (Berlyne 1970, Berlyne and Lawrence 1964, Hekkert and Von Wieringen 1990, Cox and Cox 2002), aesthetic preference is related to complexity level in an inverted-U shaped pattern.
H2a: For stores with high reputation, the impact of complexity design on sales performance will be larger for stores with moderate level of complexity than for stores with higher or lower level of complexity, with low-complexity stores’ sales performance moving toward the optimal level, while high-complexity stores’ moving away from the optimal level when complexity increases.
H2b: For stores with low reputation, the impact of complexity design on sales performance will be larger for stores with higher or lower level of complexity than for stores with moderate level of complexity, with low-complexity stores’ sales performance moving toward the optimal level, while high-complexity stores’ moving away from the optimal level when complexity increases.

​

​

​

​

​

​

​

Hypotheses H3a, H3b:
H3a: For stores with high seller reputation, the impact of intensity design on sales performance will be larger for stores with high level of intensity than for stores with low level of intensity.

H3b: For stores with low seller reputation, the impact of intensity design on sales performance will be larger for stores with low level of intensity than for stores with high level of intensity.
 

​

​

​

​

​

​

​

METHODOLOGY

Step 1: An in-depth interview and a focus group were conducted to better understand consumers’ behaviour in online marketplaces.

Step 2: Cart sorting were conducted to identify the most representative and manipulable elements of web aesthetic in online marketplace and to support the design of questionnaires for survey.

Step 3: A survey was conducted to measure consumers' perceptions of web aesthetics.

Step 4: Real-life transaction data obtained from Taobao.com (the largest online marketplace in China) were collected.

Step 5: Store design data from Taobao.com were extracted by using JavaScript and calculated by using the metrics developed from the previous steps.

Step 6: A series of controlled experiments were carried out to correlate the objective measures and the subjective measures and to test our hypotheses with potential confounding variables being controlled.

​

Operationalization of Web Aesthetics
Unity: 
U1- the completeness of the website structure
U2 - the consistency of fonts
U3 - the visual balance between pictures and texts
Complexity: 
C1- the number of texts on the viewable area 
C2 - the number of pictures on the viewable area 
C3 - the number of links on the viewable area
Intensity: 
I1- the number of animation components 
I2 - the brightness of background  
I3 - the saturation of background 
Data Collection – real-life transaction data from Taobao at two points in time (T1, T2) 
Dependent Variable – Sales Performance: the sum of revenues from T1 to T2 
Moderator – Seller Reputation: the sum of all the ratings of completed transactions.
Control Variables: Ratio of Positive Ratings (RPR), Pricing Level

 

IMPACTS

Theoretical contributions:
Quantifying the abstract concepts into concrete manipulable metrics
Associating web aesthetics with other related factors
Using real-life transaction data 
Practical contributions:
Providing a general framework for practitioners to gauge the aesthetic level