This caused a lot of controversy as it places the usability of the users interaction before security considerations and challenges an established convention.  However, as he points out, many people are accessing sites in situations where they are not over looked and making it more difficult to enter passwords may causes users to choose simpleless secure passwords. This is particularly true with mobile devices where users often enter shorter passwords to make it easier as they have a numeric keypad.  He also argues that obscuring the password does little to protect the password anyway since if someone wants to work out your password they could always look at the keyboard.  However, if the password is on screen it is certainly easier to see from a distance if for example you are working in a open plan office.

His objection may be correct, many people will be entering a password in a situation where they are not overlooked and making it difficult for these users just because some people are does not necessarily make sense.  However, failing to mask password characters may have wider implications.  As the site does not mask the password it may create the perception that the user does not need to safe guard this information resulting in more careless behaviour.

Jakob is aware displaying the password may not be ideal in all situations and suggests that a control could be provided allowing users to select to hide their password.  This approach is currently used within windows network settings and some WiFi software.  In a recent article on list apart Lyle Mullican explores this approach in more detail.  However, theapproach places the responsibility for managing whether the password is displayed with the user and adds complexity.  It is also a modal control (the user selects either to display asterisks or actual characters) which can cause usability issues.  For instance, the user may start typing the password without realizing it is being displayed revealing it to those around them.

IE 8 password entry dialogue allows the user to select whether the password is shown.

Chris Coyier in his article Better Password Inputs, iPhone Style suggests doing something similar to what is done on the ipod touch/ iphone interface where only the last letter is displayed on screen. This is fine for a mobile device where the user can take steps to ensure no one is looking at the screen while they input the details but may be an issue when displayed on a monitor.  Users may also fail to notice mistakes when they press another key immediately after they mistyped.

An alternative approach which addresses many of the issues is to hide the password by default but provide a button that when held down reveals the password.  Although the user doesn’t receive immediate feedback they have the option to check their password before submitting and the user could hold the button down while typing if required.  This approach highlights the importance of keeping the password secret and only shows it when the user expressly indicates it is safe to do so.  It also removes the chance of users accidentally revealing their password.  This is not necessarily an ideal solution in every case and there will be instances where it is best to display the password in full by default.  However, assuming thew user is entering a password in a private office or passing all responsibility for safe guarding secrecy to the user are not ideal whatever the usability issues.

The reason for these images becoming more and more difficult to decipher is that the programs used to place spam become more sophisticated in order to overcome the obstacles placed in there way.  Gmail, yahoo and hotmail have all had their CAPTCHA images broken by spammers and although it is fairly easy to present a image in a different way it does mean they have to be fairly obscure.

Gmail, yahoo and hotmail’s CAPTCHA broken by spammers

This method is fairly widely used on the internet but places the emphasis on the user to prove they are not a computer program.  There are alternatives to this approach for instance akismet looks at content submitted to a web site and assesses it against various criteria to determine whether it is likely to be spam.  Flagged content can then be reviewed by the administrator.  Similarly anti-spam programs can be used to filter out spam sent from contact forms rather than making it more difficult for customers to contact you.

Honey pots are another method that can be used on forms to help prevent submissions from malicious programs.  This involves placing extra fields within a form that will be invisible to your users.  As the programs do not view pages in the same way as users they tend to complete these fields meaning you can reject these submissions. Ned Batchelder’s article Stopping spambots with hashes and honeypots covers the use of honey pots to prevent spam in more detail.  It also indicates how to make things even more difficult for automated spam programs by randomising field names and using the submitters IP address to prevent automatic submissions from groups of machines.

These methods can do a lot to help minimise misuse of your site by malicious programs.  However, to some extent it will depend on the value of your site to spammers.  Gmail, yahoo and hotmail have seen their CAPTCHA images broken as it is worth the effort in order to get large numbers of programs automatically registering email accounts and sending spam.  Honey pots on there own will be easier to overcome if there is good reason to do so, although Ned Batchelder’s full method is likely to be more difficult.  You can reduce the value of overcoming your defences by doing things such as using no follow links in comments on your site so they will not improve search engine rating.

The main thing is not to use CAPTCHA indiscriminately on forms.  There may be some situations where it is useful particularly if the returns for beating your defences are significant.  However, in the majority of cases you are creating a barrier for the people using your site.

Most Returned Products Work Fine, Study Says

User generated content can have several advantages, it involves users in the site, means content is frequently changing and can encourage users to promote the site to others.  However, it also has a down side in that it can provide the opportunity for those dissatisfied with your brand, site or actions to speak out.

Chevy discovered the negatives of user generated content when they tried to get people to create their own adverts for the Tahoe using video clips and music they had created. People used the clips to bash Chevy and the fuel economy of there vehicles. Although not as extreme, users could be seen using the mash up functionality to criticize the site this week.

A user has used the cartoon above to post the message “Who cares about mash ups? All we want is a simple fast web site to read Dilbert! This new site sucks. Bring back the old web site!!”.

It isn’t uncommon for site redesigns to have some negative responses even if improvements are for the better. Regular users get use to where content is and the design and are put off by change. This is one of the reasons why it is important to manage change, informing users what’s happening, introducing ways to feedback and when possible making incremental changes.

This comment also highlights an important point, just because there is extra web 2.0 features does not mean users are willing to put up with poor performance on key functionality. Why it was decided to display the cartoons on the home page within a flash area is incomprehencible when an image would do much the same job more efficiently. The use of the ‘beta’ will also do little to appease users when they had a perfectly good site before.

The search seems to be a bit hit and miss currently. Sometimes when I excluded words I thought had nothing to do with my search sites I thought highly relevant disappeared. However, the functionality is interesting and allows the user to easy search by excluding irrelivant content filtering results.

One of the advantages of usecases is that they detail the exceptions to the normal flow.  this is something that sometimes is missed when creating wireframes and results in the developer having to decide what to do as they come across issues.

The only issue I can see with this suggestion is that some times use cases can use multiple screens.  In these instances I would suggest integrating the use case with a flow diagram and linking from the diagram to the appropriate pages.

The page allows the user to edit options regarding harddrive size, amount of memory etc. The options are initially set to a default configuration so if the user does not want to make changes can simply add the product to their basket. Users that want more memory or a bigger hard drive can simply select a different option using a radio button which clearly communicates the impact on price. What the ‘most popular configorations’ control does is allow the user to switch between the basic setup and the one most frequently selected by users. This changes multiple options in a single click providing a short cut, as well as suggesting a configuration that has been suitable for others.

It is fairly easy to see how this technique could be used on other sites where the user is required to configure products or packages. For example, when selecting the channels I want in my digital TV package I could be given links that select all the options applicable to someone that likes comedy or crime dramas.

A common misconception with user generated content is that if you add commenting, reviews, etc. to a site suddenly everyone will be sharing their opinions day and night. In reality only a small proportion of users regularly write review or create original content.

Even on digg there is a relatively small group of users that regularly submit stories. However, the simplicity of the voting system encourages a wider audience to get involved. Voting for a story requires a single click once a story has enough votes it is placed on the front page. This is not to say that the mechanism behind the voting process is simple. It uses a highly complex algorithm to try and prevent users manipulating the site. However, from the user’s point of view all they need to do is click ‘digg it’ if they like the story.

The site: digg

Too frequently within the on-line environment the target audience for campaigns and promotional sites is ignored.

Often on-line promotions are little more than a translation of something developed for another medium. In the worst cases a campaign is transferred directly. This ignores much of the research and consideration that went into the original campaign and transfers the limitations of one media to another. It ignores the strengths of the Internet and highlights its limitations.

Customers interact with television or print in a different manner to the Internet. Television allows little interaction beyond selecting which channel to watch. The Internet allows users to provide information and respond to messages to gain more appropriate solutions and meet goals. It provides customers with a chance to deal with a company early in the relationship. Ignoring this results in solutions that fail to fully capitalise on the medium.

Interaction provides an opportunity to reinforce brand pillars, but brings with it risks that are often ignored. If it is poorly thought out, inconsistent, misleading or error prone it will affect users’ perception of the brand. A site wishing to promote a car as a pleasurable, comfortable experience will fail if it does not exhibit these qualities.

The reason greater consideration is not given to the target audience may be reluctance to spend money on a medium that was hyped then failed to live up to expectations. However, the on-line population continues to grow and offers opportunities not present in other mediums.

Promotion on the Internet often fails to learn from marketing in other mediums. Market research of customers needs, attitudes and behaviour has been used for years. There are still examples of companies buying banner or promotional space without considering the audience and whether there may be a more effective alternative. Market research provides a good basis, however, it usually does not go far enough, ignoring the interactive nature of the medium.

Internet promotion needs to take into account context, when it is encountered and what the user is doing. The use of splash screens is a key example. The user has already made the decision to visit a site yet before they can achieve anything they must view an animation promoting the company. This is the equivalent of your local green grocer refusing to serve you until they have told you they are the biggest and most profitable grocers in the area.

The use of sponsored items on search engines is a good example of how promotion can be effectively integrated with the user’s goals. The sponsors relate directly to the topic for which the user has searched producing benefits for both parties. This is not the only way this can be achieved. Rich media banners can provide immediate value to the customer and even traditional banners can support users’ goals if presented in a timely manner.

Any promotional offering on the web needs to consider the full life cycle of the users interaction. Very often drive to site and how best to maintain interaction with the brand are considered independently. Many promotional sites seek short-term solutions, relying on a single interaction to promote the benefits of a brand. They seek to generate awareness as quickly as possible rather than initiating a relationship.

If the Internet is to truly meet its full potential as a marketing tool it needs to be approached as a unique space but draw on previous learning within marketing, human factors and software development. It needs to take account of the target audience their goals, abilities and the tasks they wish to perform in context. In short a customer centric approach.