connecting*the*dots

I’m really liking where the Amazon UI is heading.

Not to say that their interface has been devoid of good design decisions over the past 9 years. I mean, they were the first e-commerce company to truly leverage collaborative filtering, essentially taking advantage of The Long Tail of online consumerism, long before the terminology was officially anointed within the context of new media. Just getting those features into the interface — placing more eyes on more products — drove the business model and the promise of the New Economy circa 1998.

But crafting the balance between user search, participatory and discovery scenarios within the interface was definitely an afterthought. As a matter of fact, at one point in time, it seemed that Amazon had little concern for any interaction design considerations within the interface at all, satisfied to add features within a scroll-driven, hierarchical construct to no end.

Well, that has seemingly changed. Check out a new test version of the product page:

Am I the only one standing up and applauding? The interface now completely supports the common needs of all users at a point-of-purchase.

1. Clearly display the product price & savings
2. Show the product itself
3. Provide the ability to purchase it or add it to a Wish List
4. Then provide the ability to move into search, participatory or discovery scenarios

Amazon reduced the cognitive friction of the previous interface
by removing the bookend columns of contextual navigation and by moving only the highest priority scenario features (reading reviews and looking inside/searching the book) under the primary sales window. The affordance for the DHTML window on the product image needs to be increased, as it is currently one of purely learned behavior, but the simplicity of the new interface does wonders in focusing my attention.

Take a look at the old interface in comparrison:

For someone that has reached the conclusion of their search or discovery scenario, the shopping experience is now set up for the user to determine price feasibility, drop the product into their cart and either head for checkout or continue to discover.

For someone who’s in the midst of a discovery process, the interface hasn’t changed for the worse, as an anchor link still drops the user down to user reviews. It’s now just better placed. The interface continues to present other collaborative filtering information below the fold. With these smart design decisions, Amazon now finds itself in a position to add elegant customization, because elegant design parameters have been established.

The Amazon jungle is evolving, and in very smart fashion. In my book, that’s pretty cool.

What do you instantly think of when you read the word "fence?"

Unless you’re my friend Fleur (she’s a fan of the romantic sports), you probably lean towards the world of "things that stick in the ground and enclose other things" and you would be correct in thinking such, as fences do just that; they enclose, divide and protect areas and organisms for various purposes. Now, if we can agree to such a broad (yet particular) definition, we should be able to agree that fences define parameters of use, for if one can’t get past the fence, one can’t engage with anything outside (or inside) of its parameters.

So why am I fixated on the definition of a fence? Well, a good deal of tactical user experience design within authenticated environments is centered around the concept of defining parameters; as experiences that support specific goals rely on the thoughtful crafting of usage parameters. If that’s a vague concept, think about it in terms of designing site navigation behavior:

• Does the top/down navigation live in a consistent pane in the interface?
• If it’s located at the top of the screen, how does it behave?
• Does it expand onRollover or onClick?
• Does it display horizontally or vertically?
• Does it use rollovers for greater affordance?
• How does color and behavior reinforce the brand?
• Where is bottom/up exploration located?

Before we get too granular, let’s step back outside for a moment and think about how we might design the use of a fence in the real.

The placement of a fence’s corner posts could be determined based purely on property lines, or relative requirements based on the needs of the owner. The height of the fence, and its types of rails or wires, could be designed based on security requirements, or based purely on style. One thing is for sure; when these attributes are specified, usage parameters are in place and an experience is established. These specific choices define our cognitive perception of a fence that keeps sheep grazing within a field; a fence that protects property in the midst of urban renewal; a fence that surrounds a house at the end of a cul-de-sac.

Design is an iterative process, a constant remodeling based upon the objectives and desires of a business and user, and the perspective, and ultimately, the epiphany of the designer. But moving from good design to great design requires an effort of reduction to reach an elegant solution, where less is more and the complex takes form in simplistic presentation.

So the sheep are now grazing behind an elegant, rustic, utilitarian, wood fence. You’ve designed the perfect experience to meet the usage parameters of your client. Congratulations. But don’t pat yourself on the back for long. Your farmer client is now telling you that his sheep are going to be racing for cash on the property and that 25 other farmers are going to be using the property for similar, yet different purposes… and each are going to be equal owners, with equal say in decisions.

You now have 25 clients to please.

Parameters now seem passe, as each owner has slightly different needs and they’re starting to bicker. Your approach to good design (reduction and elegance), as applied to the original usage parameters, must now be redefined as well. So can you satisfy each of you client’s requirements with one static, structured, definition of a fence? Probably not, but that doesn’t mean design parameters go out the window.

An even more difficult challenge now lies on the horizon: How do you broker the creation of apropos usage parameters moving forward, through the creation of both useful and usable customization options? You need to meet their needs, but not by making "fences" too complicated or by coming in too close to their limited operating budget. How does one iterate the elegant reduction of the addition of customization?

Hmm… how indeed.

I noticed something the other day about my online shopping pattern: I don’t browse and I hardly ever buy on a whim.

After contemplating such startling, introspective findings for awhile (uhm… finished), I started to wonder whether this could be the case with most people that shop online? That got me to thinking more, so I mosied on over to Amazon.com to perform an off the cuff user interview/usability study of myself and my tendencies, comparing them to the holistic user experience of the site.

I admit it wasn’t very scientific, but what I discovered was interesting.

• When I shop at Amazon, 99 times of 100 (I multiplied this session by 99) go directly to the search field and enter a product attribute.
• After landing on a results page, I choose a product description page.
• 4 out of 10 times (ballpark figures folks), I’ll review a few user comment reviews on the product as a the final check before deciding to make a purchase or not.
• If it’s a go, then WHAM! I’m in the cart experience and out the virtual door.

What I just described is the Amazon shopping experience from 1996-97. It’s a utilitarian approach and it fits me when it comes to shopping, especially in the real world (but that’s a whole other conversation.)

Online shopping, for me, is about keeping your recommendations to yourself, don’t clutter the page with collaborative filtered data such as ‘purchase circles’ or ‘listmania,’ and just let me find what I’m looking for and get me out of the store… fast.

My assumption is that this is how most people use Amazon. I could be wrong. I’ve been known to be… often. But if that were the case, even if the numbers were only 30% that followed my shopping pattern, wouldn’t the overall user experience be a bit too much? Wouldn’t Amazon be relying too much on quirky collaborative filtering techniques to become the next generation shopping experience?

I mean, first there’s a welcome page, then the Sean store, then a trailing history of what I’ve viewed, then recommendations built dynamically based on those views and presented throughout the UI. On top of all of this there is untargeted and targeted marketing messaging (hmm, sometimes in the form of recommendations?) presented to me at each turn…

Who’s handling the UXdesign here? How many marketing MBA’s are in a room with a cornered IA or DBA hatching plots to create cross-pollination of product silos? Come on, do I really care that people who bought my book also wear clean underwear from Target? Is this the smartest Amazon can be?

Amazon deserves credit for recently removing some of the superfluous features of the site from the main experience and placing them into the well hidden ‘explore’ category on a secret page somewhere in their navigation scheme. But the powerful information that this site could provide the customer should be generated at the customer’s whim.

Please, Mr. Bezos, we’re living in the realm of customization. Allow us to create our own product page. Let us choose if we want these extra features and how we want to view them. I understand there’s a fine line between satisfying the needs of your retail partners and pushing product, but when in doubt, bite the bullet and go with the us, the people, the customers.

Amazon is now six years old, the ancient mother of e-commerce sites. Now would be the time to design an experience for individual shoppers. Create a completely personalized shopping experience through customizable interaction design. It’s the next step in online shopping, and you know how the saying goes in the retail world: the customer is always right.