En Ru
+375 (29) 692-02-78 +375 (17) 219-48-27
Contact sales

Quantitative assessment of the quality of e-commerce web sites

18 May 2009 | admin | Tools


Presentation of a report for the conference Bynet 2.3 on the topic of  Quantitative assessment of the quality of e-commerce web sites.

Are you an owner of an online store or are you planning to launch one shortly?

If you are a usability expert, a developer or if you are somehow involved in interface evaluation and design - this is for you! 

I will tell you about the role of quantitative evaluation in site analysis, KPI method and about how a usability-expert should effectively use analytics. 

«A penny saved is a penny earned»

Introduction: Today this report will inform you on how to quanitatively evaluate success of a web site, with the focus on e-commerce web sites: online stores and platforms. There are several variants of evaluation:  

1) Usability testing

The idea of usability testing lies in users working  with your product (programm or web site), while professionals watch their behavior and identify occurring problems.   

2) Expert evaluation

Expert evaluation is an analytical identification of user interface problems, based on expert experience, comparative analysis of competing systems, collection and processing of user feedback. 

Expert evaluation includes expert site analysis in accordance with check-lists and  method of tasks analysis. The essence of check-list method lies in expert evaluation of conformity of  elements  to our quality interface requirements.  

Tasks analysis method implies that expert performs real user tasks at the same time answering the questions on how to speed user work, how to increase user satisfaction etc. 

3) Web-analytics

Web - analytics is quantitative evaluation of your project. Web - analytics helps us answer the question: what does not work on the site? 


1) Coverage of all site users  

2) Coverage of all site functional  

3) Time statistics, monthly 

4) Opportunity to track visits dynamics by yourself and track how many users reach their aim: i.e. order a product, purchase a product. Its obviously better to get a complex evaluation of your site: test users, get an expert opinion and what is even more important be able to evaluate how successful your site is and which innovations are required, as well as how efficient these innovations are. 

 Figures don't lie, they say.

1. KPI Method

 Key performance indicators are in other words key metrics of your site's success. The main idea of KPI is to take technical data, process it and present in a simple business format. The following characteristics of KPI:  

1) Indicators value is always prolonged in time 

2) Color indicators of rate change

3) Setting of thresholds and warnings 

4) Advice on indicators improvement 

Key indicators for e-commerce:

Order Conversion Rate Formula = number of orders / number of visits

Average Revenue per Visitor Formula = total site revenue / number of unique visitors

New Visitor Conversion Rate Formula = number of new visitors that placed an order / number of all visitors

Returning Visitor Conversion Rate Formula = number of returning visitors that placed an order / number of all visitors

Percent of revenue from visitors that place a first order Formula= total revenue from new vistors/ total revenue 

Average cost of visit Formula=marketing costs/number of visits

And many other metrics, around 100.

2.A detailed analysis of user behavior.   The number of online stores grows from year to year, range of goods offered is also becoming wider. Due to the fact that online shopping is still poorly developed in Belarus and a site usually contains only contact numbers and a contact form, it is really hard to track how many users actually make a purchase. Let's take migom.by for example. It's a typical internet platform that sells household appliances.  The site has registration, a catalogue and sellers' phone numbers. 

Typical users' tasks on a site of this kind: A. user searches for a product  B. user looks through search results (product catalogue)  C. user looks through a detailed description of the product  D. user places an order. Usability expert analyzes user's behavior and makes conclusions about site performance.  

Usability expert's tasks 

We need to find out:

1. how home page works, where user goes from it 

2. how often search for a product is successfull (how quickly a user finds the necessary product and orders it ) 

3. what actions are most popular on product catalogue page, product description page 

4. how many people proceed to order placement, how many cancel the order 

5. how users place an order, how many users proceed to the final stage  

6. how many users leave a review about a product, a delivery man 

7. how often users register on the site, return to the site 

Site owner's tasks.

We need to find out:

1. which goods from the catalogue are most profitable and popular  

2. Total Revenue from New Visitors

3. Total Revenue from Returning Visitors

4. Buyer Conversion Rate

5. Checkout Completion Rate 

6. Average Cost per Visit 

7. Average Visits Prior to Conversion 

8. and many other

By tracking KPI metrics we can have a flexible strategy and meet users halfway. It makes it possible to evaluate the estimated efficiency of such innovations. The presentation shows which data is used for analysis and which conclusions are made by usability expert.  

Conclusions for site owner.

Indicators from Jan 6, 2008 - Feb 5, 2008 to Dec 6, 2007 - Jan 5, 2008

Conversion level for orders 1,31% -> 1,57% This indicator ranges from  2% to 5% for stores. We can see that the indicator of migom.by is very low, inspite of the fact that many customers contact the seller directly without proceeding to order form. It is possible to increase it by 2-3 times to 3-5% (by redesigning interface), which will consequently increase sellers' (and migom.by ) revenue by  2-3 times.

"Power" of the Home page that makes users stay on the site and continue browsing  77,21%->76,64% We can see a certain growth in comparison with Christmas holidays. It might be a result of some changes on the Home page. However, the indicator remains low  - the Home page does not attract users to the site. Low indicator might be connected not only with having unattractive design, but also with untargeted traffic. We can notice a decrease after the end of Christmas holidays.   

Average revenue per unique visit 0,09 -> 0,11 We can see some after-holidays decrease. Order conversion rate 0,76%->0,83%. As you can see returning visitors place more orders. The proportion of new visitors to returning is 1,62->1,64. This indicator for stores usually ranges from  2.00 to 3.00 (i.e. number of new users exceeds the number of returning by 2-3). The indicator for content sites is less than 1.00.  migom.by does not have UGC (user generated content), that is why the indicator is more than 1 (few returning visitors). On site where users generate content this indicator will be 0.5 - 0.6 or less.

Average number of visits per user  1,38 -> 1,39 Main product group of migom.by is household appliances which are not cheap. You don't buy such goods at once, you need to thoroughly consider your choice. It means that before ordering a product on the site  user needs to compare several products and make his choice.  This indicator shows that visitors do not behave in that way, a user interested in buying a product will be returning to the site (making a choice is rather long, the more expensive the product, the longer it is). It also indicates low level of user loyalty to Migom.by - users are not motivated to come back - there are no communities or discussions they can participate in, there are no deals, no product or comparative reviews.   

Recommendations: In general we can say that the site has some problems, which are a result of "unattractive" and unappealing design, low usability level. It's worth to emphasize low loyalty level (an average user visits the site 1,38 times per month). It is a result of having no strong anchors on site that would motivate users to re-visit the site from time to time. It's recommended to consider possibilities for increasing UGC.   

To solve these problems we recommend: 

1. Redesign site interface 

2. Create aesthetically acceptable, neat and, consequently, persuasive design 

3. Analyze advertising campaigns of the site (choose the most desirable and make them a target) 


So, quantitative assessment of your site is one of the reliable indicators of your site's success. When handled by a usability professional it turns into a powerful tool of professional diagnostics, metrics and analysis of users' behavior. 

Liked this post?