Found on Johnny Holland

The local Washington DC Interaction Design Association (IxDA) chapter recently conducted an IxDA ‘09/IA Summit ‘09 “redux” mini-conference, allowing members of the DC user experience community to hear abridged versions of select presentations from each conference. Jared Spool’s presentation focused on Amazon.com’s smart design decisions and how these decisions have made Amazon a successful business. In his presentation, Jared stressed that just copying Amazon’s design isn’t necessarily a good idea and doing so may have unintended consequences.

When working with clients we are often asked to “make it work like Google”, or “we want to do it like Amazon does”. These requests are understandable, because both Google and Amazon are really good at what they do. Google allows us to find information we are searching for quickly and easily, and Amazon makes buying stuff so easy that instead of going out to the store and buying something, we will wait the couple of days it takes to receive an item in the mail. It makes sense that a client would like to copy the success of either Google or Amazon, but just copying the features and interfaces of other sites is not necessarily a recipe for success. 

Amazon.com sells a lot of stuff. They sell a lot of stuff because they provide an amazing user experience; they have designed a system that facilitates users buying stuff quickly and easily. I buy a good number of items from Amazon.com every year because they have great prices, I can find what I want easily, and I am given a lot of information to make a decision on whether or not to buy a certain item. If fact, there are people who go to Amazon.com first just for the reviews even if they think they might buy the item somewhere else. There is no doubt that Amazon has created an incredibly useful recommendation system. 

Yet, just because this system is effective for Amazon, doesn’t mean you should copy it for your site. In fact this might be an incredibly bad idea. For example, Target.com uses leases Amazon’s platform for their own e-commerce site. Target’s recommendation system works in nearly the exact same fashion as Amazon.com. Yet, if you search for an alarm clock on Target.com, you will get a ton of alarm clocks that are rated very low (or not rated at all). If I were in the market for an alarm clock, I would be wondering why Target sells so many crappy ones. 

In his presentation, Spool examines why this system works so effectively for Amazon but tends to fail for Target. Amazon and Target.com both sold a few million copies of the popular book “Harry Potter and the Deathly Hallows”. Yet, if you look at this book’s Amazon page, there are over 3,000 reviews for the book, where the Target.com book page only has 9 reviews. It is obvious that Amazon users are much more likely to post review than the typical Target user. In fact, if you look at the quality of Amazon reviews vs. Target reviews, the Amazon reviews are also much more useful. Amazon reviews are often well written, detailed, and provide a lot of useful pro and con information about the product. Target reviews are typically shorter and contain less useful information than their Amazon counterparts. The fact that Target gets less review for each item also makes the reviews less useful. People are more likely to write a poor review of a product if they have a bad experience than write good reviews when they have a good experience. Thus, Target.com has a lot of alarm clock that are rated very low as not a lot of people are revved up to write a review of that new alarm clock they just bought. Amazon, on the other hand, typically gets a very large amount of reviews for its products. Since products on Amazon tend to get a large number of reviews, there is more likely to be high quality reviews of the a product on Amazon. A product with only 9 reviews on Target is not going to have the same reliability of information for a user to decide to buy or not buy a product.

So, the next time a client asks “Lets do what Amazon is doing”, remember that successful designs are always successful in a specific context. Copying another company’s successful design will not always bring the same success to your company. As user experience designers, it is our job to design the right system for the correct context. It is likely that your context is very different from Google’s or Amazon’s, so copying what they do is not likely to be successful for your situation. As always, informed design that takes into consideration the unique situation of the company’s business, users and technology will always win out over copying successful designs of others.

I have read a lot of books about user experience. I think the first UE book I read was “Don’t Make Me Think” by Steve Krug. Since then I have read many books about information architecture, interaction design, user research, design deliverables, etc. These readings have helped to shape my knowledge and opinions about what UE is, and how it should be practiced.

I have a list of books that I have been recommending to people who are interested in becoming a user experience practitioner. The list below is a good starting place for a UE newbie who needs a place to begin exploring the vast amount of information that exists about user experience design.

This book was first published in 1988, and its contents are still very relevant today. After reading this book, I knew that user experience design was what I wanted to do. Don’s books are easy reads, insightful, forward thinking, and entertaining. Read all of his books, but read this one first.

Great overview of what interaction design is, and how it is practiced.

You’ll like Mike Kuniavsky’s broad selection of practical user research methods–presented clearly and usably. It demonstrates how to discover what is in users’ heads, and suggests how we might balance those considerations with business objectives.

Why software design is broken. Alan Cooper (former software engineer) explains why and what to do about it.

While “The Inmates are Running the Asylum” book tells you what is wrong with software development, this book tells you what to do about it in detail.

This is the only book I know of that focuses completely on UE deliverables.

A complete overview of what information architecture is and how it should be practiced.

Currently I am reading “Sketching User Experiences” by Bill Buxton. I am half way through it, and I am thinking I should add this one to my UE newbie list.

Jared Spool talks about why experience design is so important

| View | Upload your own

In Alan Cooper’s “The Inmates are Running the Asylum” and “About Face 2.0“, he talks about how the user’s goals are the most important thing to consider when making design decisions. He also mentions the need to consider the person who is buying the product as well. If you can’t sell the product to the customer, then designing for the end user becomes a moot point.

In the situation that I am currently in (see previous post), we have a problem where instructors are not using the course modules that are being provided for them. In this case, they are like the ‘customer’ that Cooper refers too. If we cannot sell the module to the instructor, then fulfilling the user’s goals, again, becomes a moot point.

During the research phase of this redesign, I will be very interested to see why these instructors are ‘working around’ these course modules. What is it about the current product that frustrates them? What goals does the instructor have that are not being met?

It will be interesting to see how the importance of the instructor’s goals compares with the importance attributed to the customer in Cooper’s texts. Will the instructor be treated as the ‘customer’ in the research and modeling process, or will he become another primary persona that needs to be considered in the design process?

A response from someone I was having a conversation with at lunch. I was trying to describe to my fellow conventioners what I did. I told them I was an interaction designer, and proceeded to explain the methodology of user centered design. The responses? “Oh, we just usually skip over that part.”

Classic.

So, I am at the CFUnited conference here in DC. Sitting in the ‘Flex Coding Kitchen’ presentation, I realized that I am truly hanging out with 1000 of the brightest, most talented inmates in the asylum. It is interesting to hang out and hear them talk, and I am glad I am able to get a better insight as to what makes them tick.

I just got done listening to a guy from Microsoft who was talking to us about “User Experience”. He proceeded to describe some elements of what user experience design is. They included: it works well, it looks good, etc.

I know that we have to consider the audience: mostly Coldfusion developers. But, his ‘elements’ of the user experience were not explained well (let alone NAMED very well), and he did not fully connect how they relate to the product he was featuring: Atlas.

Atlas is Microsoft’s Ajax framework. It is designed to primarily work with .NET technologies, but also interfaces to other platforms as well. I am sure there are probably more bells and whistles when using Microsoft platforms, which is certainly why they are giving it away for free. But, exactly how does Atlas support and improve the user experience?

This is where the danger of using technology blindly can occur. Ajax is fun for developers to use, because it allows them to make user interfaces do things that they could not do before. This is also why interaction designers like it as well. Yet, without implementing user centered design techniques, using Ajax has the potential of creating an equally bad user experience as plain old HTML. The main issue that comes up is breaking the user’s mental model of how web pages normally work by implementing asycrounous data calls to the server and displaying them in the user interface– before the user hits the ‘Submit’ button. So, letting the inmates run loose with the design of these new types of interactions is dangerous.

Now, some of the questions of usable Ajax interaction have been solved since I was made aware of them at User Experience Week 2005. For example, if you have an e-commerce site that allows you to customize a product, and you are using Ajax to dynamically update the product view to reflect these customizations, Atlas creates a simple way to incorporate those attributes into URL variables so you can send a friend a direct link to a page that represents the same state the page is currently in.

Ajax itself is no different than html was in the early ’90s. In the wrong hands, it can cause major frustration for users. Interaction designers applying user centered design techniques can embrace these tools and use them to create great user experiences.

I just hope we can convince the inmates, and get them to trust the designers to create usably pleasant interactions.

This is possibly one of the hardest things for me to explain to the people I work for. WHY a design process that includes the usability engineering is so IMPORTANT. Reading the forward to Alan Coopers 2004 edition of “The Inmates are Running the Asylum” has given me some greater insight into why the old ways of creating software/web applications are still prevalent today.

Cooper talks about how it is the ‘old economy’ thinking that equates programmers in the same way that it classifies variable costs. In business, there are two strategies for increasing profits: reduce costs, or increase revenues. In the old economy of manufacturing physical goods, reducing variable costs became the best way to increase profits. If I can build widget A with cheaper materials and less labor, then I can produce it at a lower cost per widget. Building and distributing software/web applications is a completely different animal than creating a physical good. Increasing revenues through higher sales is the preferred way to increase profits because the cost of distributing software after it is design is, for all practical purposes, zero. Sure there is support to provide, packaging to design, salesman to hire, etc. But, the actual good itself, the code, does no incur further costs as sales increase. If you make widget A out of steal, widget A will always incur a cost for the amount of steel needed to create it. Code does not incur cost for each time it is reproduced. Therefore, it is desirable to SELL MORE to increase revenue, rather than reduce the variable costs.

This is an important distinction to make. Since old economy thinking strives to reduce variable costs, it tends to group development resources into these variable costs. Since you cannot create software/web applications without programmers, they are often left to design the interfaces and interactions as well. This is where the danger of old economic thinking comes in. Would a building project, needing to cut costs, cut out the architect and leave the building design to the people swinging the hammers and laying the pipes for the plumbing? This would be recipe for disaster. In fact, I would be surprised if a building could be built this way at all. Yet, software IS often built in this way. Leaving out the Information Architect, Interaction Designer, or what ever you want to call it, is like building a building with out the Architect– a recipe for disaster in anybody’s book.

I have only read the forward for this book so far, but I am excited to read the rest so that I can arm myself with the knowledge the can change the minds of the business leaders in my organization.