The Insighter

October 2006

View all articles on our new site

Neo Insight's e-newsletter on Usability topics and techniques.
We invite you to subscribe to our monthly e-newsletter.

Our next workshops

November 30, 2006 Usability challenges of new Web technologies – One day Workshop.
Save $100 if you register by November 17th.
December 7, 2006
Designing usable Web-based applications – One day Workshop.
Save $100 if you register by November 24th.

Upcoming events

November 14, 2006 World Usability Day – Come see us at the RA Centre. Participate in the e-Government Showcase in the morning or the Usability Showcase and Trade Show in the afternoon. View the program. There is still opportunities for exhibiting and/or sponsoring.
November 22, 2006 CapCHI Meeting – "The Canadian Copy-Fight: Copyright, Culture, and the Internet" with Michael Geist from University of Ottawa at the Ottawa Public Library.
April 28 - May 3, 2007 CHI 2007: Reach Beyond – Annual Conference of the Special Interest Group in Computer-Human Interaction, San Jose, CA

Special event - AJAX and Usability Presentation

Mike Atyeo, President of Neo Insight presented on AJAX and usability to a large crowd at a recent CapCHI meeting in Ottawa. Mike's talk covered both the benefits and pitfalls of using AJAX technologies, with numerous examples of various usability issues associated with some of the current crop of AJAX applications. After the talk, Mike sat down with Jeff Parks and created a Podcast which captures some of the highlights of the evenings presentation. You can download the slides (8MB) from the CapCHI web site.

Rich Internet Applications (RIAs) - Where are mapping applications headed?

As the web evolves from a data dissemination tool to providing transactional services and complex applications, usability evaluation and design is becoming more and more important. The simple hypertext link model is giving way to increasingly rich interaction models which include sliders, drag-and-drop, zoom and pan, etc. RIAs let users perform complex tasks without re-loading pages, and offer users direct manipulation, better visualization of data, and more. They make a web page into a mini application. Thus, to be successful, RIAs increasingly need to draw upon the established user experience design principles and guidelines previously only associated with desktop applications.

Map technologies are working, now the challenge
is in getting the overall user experience right

Most of the challenges in developing RIAs are design related, not technological. Moreover, the design decisions made with respect to the user experience often provide the critical competitive advantage, especially given the similarity of feature sets in many of today's web-based offerings. Whoever masters the overall user experience will have the advantage!

Comparing mapping applications like Google Maps and Yahoo's Map Beta, we see many similarities. But we also see they are still working out a number of critical interactions affecting the user experience, in such basic things as users entering data, searching, or zooming.

Comparison of layout between Google and Yahoo maps

Both show similar maps with a field to enter in a new location. Yahoo decided to show hotels by default. With Google you have to type in your request "hotels in anaheim, ca". The side panel on Yahoo can be hidden but then you can't enter a new area without reopening the panel. Google shows their maps full width and people can keep entering different locations. With Yahoo, you can immediately swap to the upper two input areas to get directions. Using Google, you have to navigate to the "Get Directions" page. Which user experience is better depends on the task.

Other important differences show up in the usability of the zoom feature. On Google, there is no indication of what you might see at each of the different zoom levels. With Yahoo, a partially transparent overlay appears, when the cursor is near the control mechanism, showing what zoom level to expect. It also is oriented to match the user model of zooming down to the details. The Google model is the opposite. You zoom up to the highest level of detail but the cues are not as salient.

Example of good and poor cues for zoom level

Although both applications provide roughly equivalent details regarding hotels in the area, only Google displays the critical information (address and phone number) in a panel on the left without requiring clicking through to each hotel separately. It would be nice if they supported people sorting the hotels based on price, distance from the location entered, rating, etc.

Google's combination of map and hotel data in a left hand panel

Each application can support additional overlays of information for more specific needs. For example, the Yahoo application allows people to overlay a live traffic report summary, showing travel conditions and the location of accidents or traffic problems.

Yahoo's feature for mapping real time traffic conditions

Exploiting the power of combining mapping information with information databases, ASFusion has created an interesting real estate search application. To experience the Home Locator application you will require Flash Player 9.

Sophisticated map interactions can happen
without re-loading the web page

Instead of navigating from page to page, all the interactions occur on the same page. First you select the geographic area from the map on the left, then you filter based on price range, number of bedrooms and bathrooms, the square footage, and amenities. As you filter, the middle column is dynamically updated with any homes that match the criteria. Clicking on a home of interest centres the map in the top right around the property location and displays the property information or house pictures in the bottom right. People can collect their favourites by dragging and dropping them into the blue favourites area at the bottom.

Screenshot of prototype Home Locator service

Web maps are are evolving into
rich interaction environments

Although this is just a prototype, it provides a good feel for the direction that web-based applications are heading. They are moving away from a page-based metaphor. These kind of RIAs use an interaction environment well suited to user's needs to get an overview, zoom and filter, and see details before finalizing a transaction. Obvious functionality missing includes the ability to view and print favourites, do side-by-side comparisons, filter based on neighbourhood, etc.

New tools are emerging to permit the design of powerful, interactive web applications. We've talked about AJAX and usability in our July 2006 newsletter and Mike Atyeo recently gave a Neo Insight presentation on Ajax and Usability which also includes an associated podcast. The Home Locator application was created using Adobe Flex 2 which is based on Flash. Flex 2 enables rapid development of RIAs for almost any platform and, due to the almost ubiquitous browser support of the Flash Player, it provides a very attractive development environment.

Microsoft Vista will be accompanied by other promising development tools, namely Windows Presentation Foundation Everywhere (WPF/E). Microsoft is fine-tuning its graphical user interface subsystem that will be part of the soon to be released Windows Vista operating system. WPF/E is a subset of this subsystem based on eXtensible Application Markup Language (XAML) and Javascript. It includes vector-based drawing, transparency, hardware acceleration, etc. The most interesting part is that it purports to support a wide range of platforms, including the Mac and Linux.

As this trend towards more interactivity and rich media continues, it will become more and more important to focus on the usability of these map-based applications. The trends are making it more efficient for users to interact with map-based web sites, and providing more direct ways for users to manipulate interaction mechanisms. The new interaction environments offer an alternative to traditional page-based metaphors like navigation, browsing, and searching. Now the challenge is in getting the user experience right as map-based applications still vary widely. Ultimately, the only way to know if your application works well for users or is better than your competitors is through testing.

Both of our upcoming workshops Usability challenges of new Web technologies and Designing usable Web-based applications address these trends and the challenges of designing for usability in these environments.

We'd love to hear your feedback.

Usability catastrophe of the month
by Gord Hopkins

In this usability catastrophe we focus on the BestBuy web site and the trials and tribulations of trying to check Store Stock Availability.

The scenario started with needing to buy a new battery for a cordless phone. I found the battery quite quickly at, under a fairly logical path of: Home > Telephones > Telephone Accessories > Cordless Phone Batteries. I even found a link to Store Stock Availability under Store Services at the bottom of the page. I was getting excited as I thought I was near my goal of finding out if the local store had the desired item in stock. However, this is where things started to go very wrong.

Screen shot showing link to Store Stock Availability

When I clicked on the link, I was expecting I would have to identify which store to check, either by entering my postal code (to find stores close by) or by picking from a list of all Canadian stores. Instead I was taken to a Help Page that talked about a Change Stores button and the Check Store Stock Availability button (see text below). It also told me that they they update the system every 2 hours. Great! But, why tell me about the feature for checking store stock availability rather than just letting me select the store and check? Worst yet, when I returned to the product page, the buttons they referred to on the Help page were nowhere to be found, nor were the procedural steps they mentioned.

Screenshot of help file for selecting preferred stores

They talked about Preferred Stores. Maybe I have to set those up first. I clicked on Shopping Help and was pleased to see a link labelled How do I select my Preferred Stores? The problem is that it took me back to the same Help page as previously.

What if I click on Store Locator? Maybe that's how I pick my preferred stores. Again, I thought I was getting close. It allowed me to pick one of the stores near me. But, alas, all I got was address, phone number, and store hours. There was no way to select this as "my Preferred Store". Well, at least now I had the product number and could call someone, if I wanted to wait through the interminable voice response system.

After a few more failed attempts I did resort to the phone. Later I found out that you can check store stock availability online, but only if you create an account and login to the web site first. Hmmm, it would have been nice to have told me that somewhere. Instead all indications were that this functionality was there for the asking, without needing to create an account or login. Moreover, if I'm planning on doing business with the physical store, why would I have to login anyway?

Why is this a catastrophe? The amount people spend as a result of visiting a retailer's web site is more than just the amount spent on the site itself. The experience impacts subsequent in-store purchases as well. Unfortunately, the poor retailer might never know they have a catastrophe. In fact, they might believe it was a good experience when their traffic analyst boasts about the numerous page views on their web site, or when their advertising department brags about dozens of hits on their ad server. But it was NOT a good customer experience!

"Last year consumers spent $93 billion online and more than $137 billion offline as a result of Internet-influenced purchases (The Dierenger Research Group, 2003). Additionally, in a 2003 study of successful cross-channel retailers, Neilsen/NetRatings found that a good Web experience could increase the likelihood of a store visit by up to 27 times. This indirect Web effect ultimately accounts for more than 40 percent of total sales for many traditional retailers." (Source:

We'd love to hear about and publish your usability catastrophes.

Which type of usability investigation is right for you?

There are a number of methods for investigating the usability of a product or service. Determining which should be used is not always easy. The following table provides some guidance as to when you might consider one of these three common methods.

Focus Groups
Heuristic Evaluation
Usability Testing
Typical method
Moderated group discussion with 6 - 9 participants.
Usability experts inspect the user interface.
Individual or paired testing of 6 - 12 participants.
What do you get?
Users' attitudes, goals, desires, and reactions to mockups or prototypes.
Identification of issues based on experience and established principles.
Observation of actual user behaviour and reactions.
What don't you get?
Problems users actually have working with the product or service.
Issues relating to specific types of users and contexts.
Issues related to specific contexts and long-term learning and use.
When or why should you use it?
From early concept to early prototype, to get reactions to proposed product value and functionality.
From early prototype to final product, to quickly identify usability issues and design options.
From early prototypes to final products, to measure performance against targets.
How will you know you are ready?
When you have a design hypotheses that needs testing.
When you need expert opinion prior to testing or launching a new design.
When you have a well developed concept of the user experience.

Focus Groups - Our experience with Focus Groups is that they are great for getting at attitudes and beliefs about up-front design concepts. However, they are generally very poor for predicting actual behaviour. In many cases, what people say they do, or would do, in a particular situation bears little correlation with their actual behaviour. We use focus groups to discover the overall context of the users' goals, tasks, and their environment, and we create concrete stimuli to evoke a response and focus our discussions.

Heuristic Evaluation - This type of expert evaluation can be very cost-effective in identifying major usability issues and getting them addressed before involving actual end users. It also helps to focus subsequent testing on items important to test with real users. Critical to the success of an expert review is the expertise of the evaluators. Only they can step outside their shoes and bring their past observations of hundreds of users to bear on the analysis. Non-expert evaluators judge more based on their own personal reactions.

Usability Testing - This can be carried out in a controlled environment. However, more and more we are using portability provided by Morae software to permit testing of applications, products and services in context. This permits more accurate evaluation of the impact of common workflows, multi-tasking, interruptions, etc. Critical to the success of usability testing is the expertise of the evaluator in designing the study, facilitating the test without guiding the user, and knowing when to probe for the rationale behind someone's behaviour.

Preview of TechSmith's Morae 2.0 features

TechSmith will soon be releasing the next major release of its very successful usability testing software - Morae 2.0. The new release will add three major improvements:

  1. Automatic measurement and graphing of standard usability metrics - Standard metrics for effectiveness (error or success rates), efficiency (time on task), and user satisfaction (System Usability Scale) will be calculated with a single click. Custom metrics can also be defined based on timing, mouse clicks, mouse movement, markers logged, etc. All metrics can be viewed as tables or graphs and can even be included in the highlight videos.

    Screenshot showing graphing of time on task
  2. Built-in user satisfaction survey - User satisfaction survey capabilities will use the default System Usability Scale (SUS) but also permit you to create your own satisfaction survey and customize the graph of results. The survey will automatically be delivered to the participant at the end of the session.

    Screenshot showing graph of user statisfaction survey results
  3. Enhanced logging capabilities - A new observer's log display will show all tasks and markers logged during the test, making it easier to capture detailed, accurate user data in a glance. You will be able to predefine tasks and log them during live testing or locate and edit previously logged markers or notes at any time during the test.

    Table showing advanced logging capabilities

Contact us if you'd like a Morae demonstration for your team.

Read about the current version of Morae (Version 1.3)

Save $100 on our November and December workshops!
Early registrants save $100. There are even more savings for group bookings. Come join us.

Usability challenges of new Web technologies – Save $100 if you register by November 17th.

Designing usable Web-based applications – Save $100 if register by November 24th.

Quote of the month

"In most projects, the first system built is barely useable. The only question is whether to plan in advance to build a throwaway, or to promise to deliver the throwaway to customers."

Frederick Brooks in The Mythical Man-Month

If you have any comments on The Insighter, or ideas on usability topics you'd like to hear about, send us an email with your comments.

We invite everyone to subscribe to the Insighter, our monthly e-newsletter.

If you wish to unsubscribe, just send us an unsubscribe email.


  Home   About Us   Services   Case Studies   Training   Teamworks