Amazon Prototype
I chose Amazon.se as it was launched relatively recently to a very bad reception. As a former user in the UK, I knew the ‘niggles’ from previous experience. However, using it in another country has changed those annoyances into problems. After reading reviews from other users and reading how poorly the launch went, I decided that the Swedish version needed to be investigated.
​
At different stages there are multiple issues that could lead to either; bad reviews, drop-outs or low customer retention during and after the experience. I have also noticed that options and designs have shifted since I started to map this journey so there is no consistency across the platform.
​
My first assumption is most people know what they want to look for as it isn’t a casual browsing app. They have a motive for using it but not with a specific item in mind e.g. household items, fitness equipment etc.
​
As an iPhone user I have never used the app and always selected it use the website on my phone as the app does not appeal to my aesthetic. A study from Masaaki and Kashmura at the Hitachi design centre found that users perceived aesthetically pleasing design to be more useful. Whilst the design is cross platform, it still reads more android and does not necessary appeal to my apple user preference.
Proposed Solution
After testing my assumptions using a lo-fi Axure protoype with my test users we found some of the assumptions to be validated so worked on ways to improve those elements for the user journey.
​
The first solution for the language issue was quite simple, have a drop-down box that the user can select any language that they are most comfortable with. This would be a valuable addition allowing more users to have a better experience and be less likely to use a competitors website.
The issue regarding the auto fill delivery location underneath the search bar has been removed. I re-arranged the page to bring the login details underneath but eliminated the random address. There is still an issue of delivery times, if the user is not aware of how long the process will take. Once they have selected a product there is a new field for them to enter in their zip code for a more accurate result. The delivery times in the product description use the estimated delivery time for the item in Sweden.
After testing the lo-fi version with my users, I found that few of the potential customers used the cards to filter the results. So instead of the cards being an extra step, is to introduce them earlier when the user searches a specific term. Using the example of toys, my users gave me insights into what they would want to see as options. They suggested that they would type specific search criteria so an extra section for key words would be more beneficial. The users did comment that after scrolling through 3-4 pages of results, it would have caused them to drop out. So to bring the filtering process within the search criteria they have more control and eliminate another step in their journey.
The product pages provided too much text and the long scroll to reach the information they found more important was tedious. The user’s top priorities were reviews and delivery. Creating drop down boxes they could open and close the information, without having a laborious scroll, would have made the experience quicker and simpler. Also, there was too much information. This made it hard to differentiate so this page was rearranged to make it easier for the users to view.
The final issue was once the user had logged in; it automatically completed the order to which my users commented on being nervous when proceeding. To speed up the process and to create trust for the user, was to bring back the page allowing the user to select the card and the delivery address.