Amazon Shopping Cart

An observation and an idea for the Amazon shopping cart. 

One thing I noticed about the Amazon shopping cart is that their call to action for a user to remove an item from their cart is a link with the text: “delete”. I was expecting to “remove” but was surprised to see “delete”. I think this subtle word difference is done on purpose in order to discourage users from clicking on that link. As a user, I place a very negative connotation on the word “delete”. It implies that by clicking on such a link, you will cause a permanent change. The word “remove” is much more innocent and temporary. It’s much easier to bring something back when it has been “removed” than when it has been “deleted”. 


So what happens if I click on delete?


Interestingly enough, the confirmation text says that the item was “removed”. =)

Now, on to an idea for how to improve on the delete/remove functionality. As a user, I may have removed the item by mistake and want to immediately bring it back – an undo function. I’m surprised Amazon doesn’t have this. 

Here’s a concept mock of how it could look. By clicking on the undo link, the item is immediately brought back into the cart. 



Update to Amazon Sign Out

In a previous post, I shared what the Amazon Sign Out functionality (or lack thereof) looks like. Today I see a new experience (not sure if this is their next redesign or if I’m in an experiment) that does include an explicit sign out. However, it’s not immediately visible on the page as the user has to hover their mouse over the account menu in the top right in order to see the sub menu with the sign out command. 


Sign Out Pages

I did a mini-discovery of the “Sign Out” pages of some of the sites I often visit. This is the page the user sees after deciding to “Sign Out” or “Log Off”. The main purpose users perform this action is so that other subsequent users of the same computer do not have access to the original user’s personal and/or financial information.

First, I found it interesting that, from what I could tell, there are two naming terminologies to choose from for a site. Sign In / Sign Out and Log In / Log Out. I think I like Sign In / Sign Out more, but lets move on to the more important stuff..

For the site itself, the main purpose of the page can be to:

  • Give the user the ability to immediately enter the site again (Sign-In functionality on the Sign-Out page)
  • Take this opportunity to merchandise something to the user
  • Give the user a confirmation that they have indeed subsequently signed out

Let’s take a look at some examples.

WordPress, Facebook, and YouTube took the novel path of redirecting the user straight back to their homepages (this is why I haven’t provided any screenshots).

Chase communicates to the user that they have successfully logged off. I find it funny that they included the date too – not sure what the importance of this is for the user.


Gmail uses this page to merchandise the service (Gmail) to the user. Not sure why they would do this as the user is already a user – maybe to garner even more loyalty? Also, Gmail has a button that the user can press to go back to the homepage to sign in again. I don’t like this approach as I don’t think this page adds anything that the Gmail homepage doesn’t already have. In fact, it doesn’t even have a sign-out confirmation message. All it does is add more more unnecessary step for the user to sign-in again.


Twitter both confirms the sign out and takes a moment to encourage the user to use Twitter on a different platform i.e. Mobile. I wonder what their mobile sign-out page looks like. Maybe they encourage their users to try Twitter on a tablet or a “normal” computer?


Finally, I saved the most interesting one for last. As an Amazon user, it is NOT possible to sign-out! Well, actually, this isn’t entirely true. It’s just that there is no explicit sign-out link. The user only sees two links referring to not being the user that is currently signed in. As a user, I find this annoying and confusing. Why can’t I sign out? Obviously, they do this in order to keep people from signing out.


But the reality is, both of these links are in fact sign out links. By clicking on either of the “not so and so” links, you are taken to this sign-in page, and if you click back or go to the Amazon homepage, the original user is no longer signed in. Clever…but possibly annoying.


Error Messages

Because of my current job, I’ve had to work a lot with how best to handle errors. For some errors, the root cause can be attributed to the user i.e. (1) the user needs to provide a specific (and correctly formatted) piece of data for the flow to proceed or (2) the user did something “wrong” (I use the word in quotes because the customer is always right, and we need to be careful when labeling their actions as wrong). For other errors, the root cause can be attributed to the software itself i.e. something went terribly wrong in the code and the user cannot proceed in the typical manner the user is used to (aka fatal errors).

My philosophy on errors:

Give the user three important pieces of information (if you can):
1. What happened? Explain, as concisely as possible, what went wrong.
2. Why it happened? Explain, (again) as concisely as possible, the root cause of the problem – in language the user understands (don’t tell the user that there was a Null Pointer Exception!!).
3. What next? This is the most important step. Explain to the user what they need to do in order to resolve the issue.

For fun, here is an example of a fatal error (from Facebook):

Google Maps at the Edge of the Earth

Interesting thing happens when you explore a location in Google Maps that’s at the edge of the earth. For example, if you search for Fiji in Web Google Maps, you will see it appear just to the right of Australia. Now, if you want, you have the ability to click on the map and move it to the left or right as much as you want. You can keep moving to the right until you stumble right back on Fiji. It is as if you flew around the world and ended up in the same spot. 


However, if you try to do the same in the iPhone Google Maps App, you cannot, and you are figuratively stuck at the edge of the earth. It appears as though the iPhone version of Google Maps does not support the ability to move from left to right across the different hemispheres! 



Online Banking and Additional Security Questions

It is typical for an online banking site to ask you an extra security question or two if they don’t recognize the computer you are using to login. I took a look at two banks I use and how they treat this user experience differently. See below for the steps taken for each bank’s user experience. 

While users would (typically) go through three steps with each site, the two experiences are very different. In the case of Bank of America (bofa), their typical login – even without the extra security question – is split up into two steps. First the user provides the username, then bofa shows the user his/her site key, then the user enters the password. In this unrecognized user flow, the user enters their online id, is then asked an additional security question, shown their site key, and then he/she enters the password. So in all, the flow extended from two pages to three. 

In Star One, the user typically comes to the home page and logs in from there – so it’s just one page (and one page less than the typical bofa flow). In the unrecognized user flow, they login, are asked to answer an additional security question, and then are asked whether or not they are on a public or private computer. So their flow has been extended from one page to three pages. 

I think the best way to go for Star One would be to take a page out of bofa’s book and bring in the preferences option into the security page itself and leave it as a checkbox with help text or a help link popping out to an overlay (like bofa). This will reduce the user flow by one step and keep the product virtually the same. 

The other difference I noticed between the two flows is that bofa defaults into the user being on a private/safe computer whereas Star One defaults the user into being on a public computer. This is a simple trade-off of: what’s more likely vs. extra security for the buyer. I’d be interested in seeing analytics on what users typically click on for this preference. 

Bank of America

Step 1: (partial) Login with username


Step 2: Answer an additional security question
Optional step: Remember this computer 


Optional step: Help Overlay when user clicks on “?” by “Remember this computer”


Step 3: (complete) Login with password


Star One

Step 1: Login with username and password


Step 2: Answer an additional security question


Step 3: Preferences layer


Google Finance and Google +1

According to Google, the Google +1 product is a “button to give something your public stamp of approval.” 

I noticed this button in the individual stock pages of Google Finance. In addition to seeing this button, the user also sees the number of people who have clicked on this button – effectively seeing the strength of the public’s endorsement of a page – and specifically for Google Finance, we see the popularity of a stock. 

Intrigued by this concept, I queried several different stocks and then I noticed something interesting – the stock with the largest +1 number that I could find was…wait for it…none other than GOOG themselves! For me this was humorous as, in this case, the fact that GOOG had the largest number wasn’t a testament to the +1 feature working as designed – i.e. finding the most popular stock – but rather a function of the employees of the company who implemented that specific feature being more likely to (a) visit Google Finance and (b) click on the +1 product.

I know Google Finance and Google +1 are two separate Google products and companies often times like to use one produce to market the other, but this use case had a funny result. 

Some screenshots for reference:

AAPL: 1.1K
GOOG: 1.7K
MSFT: 87