App Critique: Searching & Browsing Experience on Amazon
While doing this critique, we were sort of frustrated. We were too naive so that we did not expect that Amazon is changing its design on a daily basis, thus it is impossible to critique in a stable status. It made the critique process extremely hard. Therefore, what we wrote may not be consistent with what you are seeing in the app. We would like to show our opinions on what we saw at a certain moment (with images) and how the differences would influence the user experience. :)
About product searching experience
Search is probably the most frequently used function of an e-commerce platform. Amazon provides three main categories of search methods, through text, voice, and images. The weirdest thing happened when we tried to see how Amazon shows the search history from different ways of searching.
How would Amazon decide the way to present the search history?
While exploring the search function, We realized if an image is used as a hint for a search, the system would analyze the possible keywords (e.g. brand of the product) from the image as tags presented above the search results. If the tags were being selected, they would be displayed as search history together with the history generated by text and voice search. The way Amazon presents the search history is different from Taobao, a Chinese online shopping website, where the search history generated by images would be presented separately and there is a search history specifically for image search.
However, it’s where the weird part begins, when we try to screenshot all of these. The interaction of search changed. The icon of voice search became the same as Alexa! When we spoke,” Please show me some water.” Alexa answered, “showing the items matching water”! This is pretty good in our opinion, since it was showing the progress of the system when the user was waiting, compared to the old version, where the user had to wait for the system analyzing their voice quietly until the results were ready. However, when we checked the search history (below the search box when the cursor is in), it wasn’t displayed as it was before.
The strategy of showing the search results
While other e-commerce apps like JD.com, which is the biggest B2C platform in China, allows users to upload their own pictures as hints for search, users on Amazon, however, can only scan surroundings via camera as hints for search (Surprisingly, right before we are publishing this article it changed! Currently, it allows the user to upload pictures, but what we don’t understand is why Amazon allows users to upload pictures only when this specific context “Find the look”).
It is not only inefficient but also not that effective. Among the five objects we have tested, only one was found. We cannot really say it is because of the limited SKU or the immaturity of the “image scan and recognition technology”. Although the only one being found was really accurate, most of the time, it shows “Hmm.. we were not able to find a match” and tends to easily disappoint the users especially after waiting for almost ten seconds. Compared to what Alibaba did on Taobao, no matter what the user is scanning, the system would find something relevant. It may not be what the user is trying to search exactly, but still something with similar features, such as physical structure or color. The users may be able to discover something that also fulfills their needs in some way as well. The user experience is much better than nothing would be shown after waiting for so long. For a platform big like Amazon, we believe the technology isn’t a problem (maybe Amazon applies really strict conditions of a match), the search strategy they are using now might be intended to match their value and belief of their business.
Search hints
The text in the search box shows “Search Amazon”, which does not give users any hint about what could be the possible keywords. We think the optimum display of the text was probably researched thoroughly by Amazon. As a platform that has existed for more than 20 years, it has relatively experienced users. Instead of showing specific categories of products, strengthening the brand might be more important to Amazon. The wording doesn’t intend to nudge people searching for something that they have browsed either. Take Taobao as an opposite example, when users have searched for a keyword let’s say “pet”, the next time the user is on the platform, s/he might see “pet snakes, nonpoisonous” displayed in the search box. The way Amazon is using strives to help users focus on their purchase goal at the moment rather than being distracted while spending too much time on things they have accidentally checked once.
The search hints under the categories showing “Search in + CATEGORYNAME”, which kind of make sense in some way. Users would probably think they can only search for something under this category. However, they are still able to search for items from all the categories on the platform, which creates an inconsistency between what it tells users to do and what users can really do.
About product browsing experience
Product information occupies a very large area in the Amazon, which means the experience of browsing products is a critical factor for users to rate this app. However, after the app critique, we found that there are still aspects that are supposed to be improved.
Homepage
Layout
The layout of the product display of the Amazon homepage does not promote users to browse this page efficiently. In every section for product recommendation on the homepage, there are only four products displayed and most of them are the same kind of product. As a result, users need to keep scrolling to know what other kinds of products for recommendation will be shown on the homepage, which is a waste of the space and will slow down the users’ browsing speed.
Tag every section of recommended products
For every section of recommended products, Amazon gives them special tags that explain where the recommended products come from so that users will feel more secure because of knowing the data sources. However, when users get interested in some products and tap “see more & manage”, they can not find many similar products. In addition, there is not a refresh button so these recommendations change only when users re-enter Amazon.
Passionless deal of the day
“Deal of the day” is the first section the user notices when they enter Amazon. Combining with prices cheaper than usual and a countdown, this feature can stimulate people’s desire to consume. After users tapping “see all deals”, the “deals” page displays all the products on sale. Besides, the current prices are indicated in red and the countdown is green, and some products even show their remaining quantity. However, on the homepage, all the data is black or gray, which reduces a lot of the visual impact.
Product description page
Burdensome customer reviews and questions
On the top of this page, there are five stars to indicate the user’s rating of the product. And it is a very convenient function that when users tap the stars, the page will slide to the “customer reviews” area automatically. Unfortunately, not every user will notice that because the stars look unclickable. And users can’t see the percentage of people in each segment without tapping the customer reviews section. What’s more, if the reviews are not enough, users can’t choose a certain rating to see the comments, which might be annoying when users just want to see some low ratings.
Customer reviews and questions take up a lot of space on this page, and a large amount of dense information will reduce users’ motivation to read. Considering the rational use of space and appropriate word count, changing the customer reviews and questions to a foldable form can solve this problem. Thus, users can get a choice to read the reviews and problems in detail or not.
Description
Some products do not have any description or the language of description does not match the language that users chose for Amazon. This problem will make the user’s experience of selecting and browsing a product dissatisfied, because users even cannot get any information about what the product is and how to use it, not to mention deciding to buy it.
Written by Peiyue Jiang & Yuan Gong
