It appears that overconfidence in Moody’s was one of Desjardins biggest mistake in this matter. The irony here is that Moody’s rate negatively institutions (Desjardins) that are in difficulty because they relied on Moody’s positive rating of certain financial product. It sound like Moody’s rates Moody’s’ negatively.

1. Experts cannot assess terminology choice and evaluate task compatibility. Terminology and task compatibility account for the majority of usability problems. To my knowledge, the only effective approach to evaluate terminology and task compatibility is through usability testing and field study.
2. The importance of heuristic (guidelines) depends of the context of use. For example, is the user going their once in a lifetime? Or is he a power user going there all the time on a mandatory basis.

As I always do, I complied telling the client I will do heuristic evaluation with the help of informal usability testing instead of pure heuristic evaluation. That way, I don’t loose time explaining clients why they are wrong. I asked a few colleagues or friends to act as subjects for the informal usability testing who could be representative of the audience.

Following the evaluation, I found the design of the portal was poor but I was not surprised, I have often seen bad design throughout my consulting practice. As usual, most usability problems were associated with confusing wording, task compatibility issues and funky use of interaction techniques. Up to now nothing unusual. I went on presenting my findings. But, while presenting, the client constantly interrupted me saying this or that; we cannot change it because it is a decision from the upper management. In some cases the minister himself got involved in the design choices.

Only later, after the formal usability testing where all users failed on all tasks, I got the client to admit the design was poor. At that point they had no choice but accept my recommendations. This situation is painful because it places the management in a bad position. They were involved in poor design decisions.

More recently, a client in the airline industry told me he does not want to hear about usability because everybody has a different opinion. They often end up asking the CEO to decide. Again, I will have a very challenging consulting project because it might show the CEO made the wrong decision. Of course, I will try to find ways around it but nevertheless, it is not easy.

For many organizations, design (user interface design, user requirement, and functional specifications) is a painstaking process involving fierce battle between stakeholders. When they do not agree, they escalate to upper management for decision.

Escalating design decision that otherwise should be taken by the design team is a sure recipe for politics. If decisions are wrong, it will be uncovered sooner or later. Then, you will be in serious trouble.

How to prevent this?
If you’re an executive, never get yourself involved in detailed design aspect. Keep yourself at the goal level and let your team design solutions. If your team tries to get you involved in design deadlock, it is because they have not done their homework or they lack expertise. Your project manager might say there is a deadlock; he needs your help to move forward. He will try to reassure you by saying he will fix the usability issues later. Don’t believe that, it will never happen and you will get stuck with a poor design for years. Stress it to your team you don’t care if the screens are blue, white or yellow as long as users were able to achieve their goals rapidly, without hesitation and errors. Get an external usability company to evaluate the design of your internal team so you will get an objective assessment of the design. Don’t worry about the susceptibility of your people, there is an easy way to manage that. If you don’t show leadership, you risk to be stuck with poor design for years if not forever.

Never forget, usability equals productivity. The first reason why we use technology is to improve productivity, so good design is extremely important.

If you are member of a project team, base your design decision on facts instead of opinion. How? Learn usability best practices, build a prototype and test your design with users with usability testing. If you do not have this expertise, don’t hesitate, get and external consultant in usability.

This graph shows the evolution of George W. Bush’s approval rating over time. His approval rating rose to 68% in March 2003, at the beginning of the Iraq war, and declined steadily after that. It is now at 26%. Not only did the war not proceed the way he had told people but there is a growing number of Americans questioning the idea of going to Iraq in the first place.

How to lead?
First: be right
No matter how good you are at executing and communicating, and how much charisma and resolution you have, if you are wrong, you will fail. Churchill was resolved but he was right. In the hiking story above, the leader was sure of himself but wrong. He overestimated how lost we were.
To be right, you need facts.
In real life, we do not know with certainty the actual situation. In the context of uncertainty (most real life situations), the more certain you are the more likely you will be wrong. Facts are even more important in uncertain situations. Why? What you know and what don’t know are facts. ”I am lost in the forest” is a fact. If the leader in the hiking story would have admitted the true level of uncertainty about how lost we were, he could have proposed a series of tests. For example, he could have propose a short walk in a different direction to gather more facts about the current location. A negative result would then have been seen as new information on where not to go instead.
Second: Communicate facts (reason for change)
It is very difficult to convince people to change something if they are not convinced about what the problems are and their urgency, see John Kotter. Two economists might disagree even if they use the same economic model. By separating fact from opinion, we might understand the input or assumption and understand why those economists disagree. Whenever you make a decision where people must follow your lead, facts are the essence of getting people agreeing on the current situation and the course of action.
In business transformation (change Management), facts are more than data “Not everything that counts can be counted, and not everything that can be counted counts.” (Sign hanging in Albert Einstein’s office at Princeton), facts are like an observable situation that can be described and shared.

Once facts are gathered and separated from opinion, they can be shared during working sessions. Those sessions permit the group to agree on problems. Consequently, it will be much easier to get the group to agree on a course of action once they have agreed on the problems. During those working sessions, the focus is first on the current situation, not the course of action. Ensure everybody that they are respected and never discard people expressing bad news (don’t shoot the messenger). Often they might be your best source of facts. To ensure the free flow of information, remove red tape, formal presentation and favor a more informal organization with a high degree of respect, see ‘‘The Death to Bureaucracy” from Jack Welch.
Third: gather more facts
Often, you need to collect more facts. Creativity and imagination are required to gather facts. For example, in the above story, a series of short walks in different directions is an approach to gather more facts on the current location. Gathering facts is boiled down to: observation, experiment, calculation and simulation. Fact gathering is not a consensus driven process. You are collecting evidence. Never rely on opinion. Humans are biased and opinions rarely correlate with facts.
In business transformation initiative, the best sources of fact are field studies. The field is rich and wild; people develop very sophisticated tactics in real work environment. Knowledge lies in the field, not in meeting rooms. The worst source of facts is the traditional hierarchy : line manager, director and executive. Facts gathered through this path will surely be altered, trimmed down or hidden.
Fourth: prevent bias
Facts may be distorted. When relying on human sources, multiple independent sources are required. Be careful with surveys or focus groups, they do not gather facts but opinions. They provide the tip of an iceberg. Avoid group meetings. Groups are only good to solve problems, to brainstorm or share information.
Objectivity does not exist; you have to cope with that.
No matter how good the information is, if you are biased, or already convinced of the answer, you are in danger. A bias could, for example, lead one to accept or deny the truth of a claim, discard evidence or look for evidence that confirms a preconceived idea (confirmation bias). In the hiking story, the guy seemed so sure of himself that we trusted him even if the direction he gave did not ring a bell to us.
To reduce bias, observe yourself to see if you feel emotional. You need to detach yourself. It is well known that a lawyer should never defend himself or a medical doctor should never diagnose himself. Ask different persons in a one on one session to interpret facts. Don’t give them your thoughts. Observe the difference in their interpretation. Try to think as fresh as you can. Gather all facts prior to a decision. In order to avoid bias, good investigators will collect evidence and then conclude. After all this work, let the decision emerge naturally. If you still do not feel not comfortable, your instinct will tells you there is something wrong, listen to it. It means there are still uncertainties that you have to work with, or simply wait.
The four steps described here relate to making the right decision and communicating that decision (vision) to a group of people. Once you have the right direction and a group of people agreeing with you on the course of action, executing change is a piece of cake.
Leading people in the right direction requires hard work, time, experience and wisdom. You can get there faster by practicing the right thing. For more on the subject, I suggest reading John Kotter on this subject and “Engineering psychology” by Christopher Wickens.

2. Avoid having wrong design goals. For example minimizing the number of clicks. To ensure having the proper design goals, know the users, their goals and the context of use. If users are going to your site once in a while on a non-mandatory basis (online banking, news, shopping…), you have to treat them as a novice. They have to understand instantly how to use the site. In this context, it is preferable having a few more clicks and pages with less information. On the other hand, for power user, speed is more important. In this context, less actions, pages and clicks are more important.

3. Do few things and do them well. For example for an e-commerce site, ease of finding the product, having the right information on the product and trust are the key success factors. For a search engine, ease of entering queries, speed and pertinence are the key success factors. Don’t spread your energy on other features unless you master your key success factors better than any others.

4. Design for user limited attention. On the web, many users have little patience and do many things at the same time (they might go on your site while at work). Within 10 seconds, they will abandon if they don’t find what they are looking for. Since the reading speed is about 200 words per minute, you can provide a maximum of 35 words to read before a user abandons. So minimize the quantity of text. Contrary to TV where the full screen is in use, on a computer, only a small portion of the screen is viewed simultaneously (roughly 6 to 7 cm diameter). This is due to the useful field of view, which corresponds to the subtended angle of the fovea vision. Information must be grouped within this diameter. That’s why banners are so ineffective. For most page layout you can afford about 4 to 5 groups. Avoid long lists of items, users read only the first few items at the top of the list. Organize according to the natural eye path. Remember users start from the top left corner, go to the center and if they don’t find what they are looking for, then they try to go to the left expecting a menu. For this reason, avoid menus on the right. The bottom right is the worst real estate of a page.

This sketch show the typical eye pattern on a web page.

5. Avoid relying on branding. Users do not remember seeing your ads, brochure and other corporate information. Keep your brand simple and name your entire product line in relation to a user goal. If you already have a brand for a product, provide a prompt matching the user goal. This also help SEO (Search Engine Optimization). For the same reason, you don’t need to be that consistent between corporate brochure and web site.

6. Users don’t read, they scan. Users will scan text and when they find something that is likely to meet their goals, they click on it. You better to expect that users won’t read your text. A good approach is to provide just a list of hyperlinks with a large font. You can also provide a small description with a maximum of 15 words with smaller lighter font under the hyperlink. This technique is ideal for a newsletter. Users will scan the list, if they are interested, they will click on it and read the description under it.

7. Ensure users understand your text and graphics. At least 50 % of the usability problems are related to the wrong choice terminology. Users get easily confused between concepts such as events and news, schedule and time table…. Minimize the number of concepts. Don’t worry, users mix apples and oranges. Test and test again your site with real users to find the right terminology.

8. Don’t reinvent the wheel. Use basic web controls. You don’t need to create complex navigation structure. A simple menu and the browser’s back and forward button are Ok for most of the cases. Once users have a successful behavior, they continue to us it for other goals. For example, if they find something through the menu, they will try this method again for another goal. Limit your design to familiar controls. Use straight hyperlinks to select a page. You will make more money with blue underlined hyperlinks because the users’ reaction time is about 30 ms less.

9. Use readable font. If users find the information and can’t read it, it’s zero. Why using small fonts? If your fonts are too small, only younger users will be able to read it. With the right size, everybody will be able to read it. Fonts must have a size equal to 1/200th the distance of reading for proper readability. On a computer screen, fonts must be at least 3 mm. For example MS Sans Serif or Verdana at a minimum of 11 points are Ok. Font adjustment controls don’t work because most users that would need them never use them. Those having serious reading problems already have special adjustments on their browser.

10. Be permissive to errors. Users make errors. The best way to manage errors is to allow them. For example, be permissive to various date formats. On the other hand, if users make errors, ensure proper detection, recovery and simple messages. Messages must have two items: the status and the corrective action.

For more information , you should check the Top 10 Web Design Mistakes of 2003 from Jacob Nielsen

“Whitman says: “Despite its controversial reputation, nuclear is efficient and reliable.” Yet since 1950 there have been 20 nuclear accidents. One was major (on Mar. 28, 1979, at the Three Mile Island nuclear reactor) and one catastrophic (on Apr. 26, 1986, at the Chernobyl nuclear power plant). How can we say a major nuclear accident will not happen again?

Here’s the reality: Accidents occur because humans make errors. Those errors can be hidden at the design, maintenance, management, operation, and even the regulation level. In addition, there are conditions that increase the likelihood of human error such as fatigue, stress, attention and memory overload. The risk of another catastrophic nuclear accident exists. While the risk is low, the consequences could be disastrous.

You have to ask yourself: On what do you want to bet your future?”

Sounds familiar, you’re not alone. Almost all users of the classical stove design, as showed in figure 1, activate the wrong burner at one time or another.

Why? It is in the human nature to associate objects that are side by side. As showed in the figure 1, the controls of the back burners are closer to the front burners. We cannot tell witch control activates what. This layout is highly error prone. To ensure we activate the right burners, we need to pay attention and look at the instructions. See also Don Norman’s Design of Everyday Things.

Fig.1 Classical Stove Design

Fig 2. New Stove Layout

With the layout presented in figure 2, there is no confusion. Association of control with the burner is now easy because the layout of control is compatible with the layout of the burner.

This example demonstrates design can play an important role in preventing human errors. A design that respects human’s nature and limitations will reduce the likelihood of errors.

Human error in managing human error

After having done hundreds of interviews with managers, we observed when discussing employee’s errors, most managers:

1. Blame employees for lack of attention or for lack of training (practice).
2. Overlook the possibility of preventing errors by a redesign of workstation or processes.

In our highly advanced technological society, we are still at the Stone Ages in term of human errors management

The stove example is the tip of the iceberg. Human errors are everywhere; from day to day errors such as trying to pull a door that should have been pushed to high impact errors such as: being attentive on the wrong signals and crashing a plane killing hundreds of people, having the wrong information and making bad investments, forecasting the wrong economics figures, misjudging intelligence report and brining a nation to war, making the wrong medical diagnosis, underestimating a budget for a project and being insouciant of consequences of our action on the environment.

Most of those errors, boils down to a combination of human errors that are at the cognitive level. They are errors of detection, attention, planning, estimation calculation, judgment, decision, comprehension, discrimination and execution.

Although aircrafts are still crashing, the safety record of the aviation industry is impressive when compared to other industries. At the end of World’s War 2, the aviation industry realized, following the pioneer work of Alphonse Chapanis and Paul M. Fitts, that human errors were caused by subtle inherent human cognitive limitations. The Aviation industry stopped blaming pilots or maintenance personnel for human errors decades ago. They tried to understand the inner cause of human errors.

This demonstrates it is possible to move away from the stone ages in a few decades. Managers need to stop limiting explanation of human errors to the lack of attention, lack of training or poor personnel selection and expand their analysis of to the real nature of human errors.