July / August 2003 Feature Article

Usability Testing: Recipe for Success

It is 21h45 and I am tired. First of all, it took three attempts to get connected to the Internet, with "Error 678" (probably some computer code for: "normal people are sleeping at this hour, please try again tomorrow") popping up every time.

The first two job sites I found forced me to register before I could browse for available positions. Yeah right, why should I give them my e-mail address and name? So they can harass me?

I am finally on a job site where I can Fill in an online application. Half way through, I need to paste my cv in a white area. After I have done it I see the text is all scrambled. Un fortunately I can see half the text! The other half is out of sight, and there are no scrolling bars to view it! Well, let them battle with it! I don't have the energy now. After about twenty minutes (on my telephone account!), I'm done. The only thing left now is to submit the application. In the bottom right-hand corner is a half green button, I can't see the name and no tool tip appears when the mouse is placed over it (AGAIN, no scrollbars!). In good faith, I click on the mystery button and wait...

"ERROR". "Telephone number in wrong format" And it throws me back on a BLANK application form!! Do they honestly think I'm going to start all over again? It is 22h30 and I'd rather stay unemployed.

With the last bit of energy left in me, I ill send them an email, giving them a piece of my mind (the one I'm busy losing). Just click on CONTACT US ... and ... what the ...?

"Does your query involve:

  • Services offered on the portal.
  • Advertising or newsletter.
  • The content on the portal."

This sounds like something from a Star Trek or Stargate-SG1 episode. What on earth is a portal? Well beam me to bed, Scotty, I'm done with the WWW (What Went Wrong) for tonight.

This is not fiction. Do you know how your users experience using your product? Tom Erickson of the Apple Computer Interface Group said: "Anyone who knows enough to design a product, knows too much to understand how users will react to it". Are your users so thrilled by the experience of using your product that they keep coming back for more? If the answers to these questions are YES, you are most probably doing usability testing before a release. If your answers are NO, there is a good chance you don't know what usability testing is.

Let me expalin:

What is Usability?

According to the ISO 9241-11 standards, the definition for usability is: "the extent to which a product can be used by specified users to achieve specified goals with EFFECTIVENESS, EFFICIENCY and SATISFACTION, in a specified CONTEXT of use". In simple terms: "the ease of use".

Why is the usability of your system imp[ortant?

A usable website attracts more users, more traffic means more business for you. As we all know, more business means more money (the reason you have a business in the first place).

A usable application means your staff get more done, with the minimum errors, in less time. Time is money (the reason you have a business in the first place).

What is Usability Testing?

It is a type of non-functional testing. It evaluates how easy a system is to learn and use. Now I've probably lost half the IT audience. Functionality is the system, who cares about the rest? The answer is: your users do. Jef Raskin, the creator of Macintosh said: "As far as the customer is concerned, the interface is the product".

There are mainly two ways of doing usability testing: with testers or with the involvement of the users themselves. Let's look at both in more detail:

  1. Usability done by testers

    In this testing method we use general user characteristics; attributes common to most people because they are human. Examples of such characteristics are: how much we can remember at once, the way we see the world, norms of society, etc.

    One of the most popular usability testing methods is to use Jakob Nielsen's ten usability heuristics. Examples of some of the heuristics the tester will look at:

    • Speak the user's language (for example: prevent jargon, like "portal").
    • Minimise user memory load.
    • Constant feedback to the user.
    • Prevent errors as far as possible.
    • Sufficient HELP in the system.

  2. Usability testing done by users

    3. In this testing method we use specific user characteristics; attributes unique to a specific type of user. Examples of such characteristics are: level of skill, limitations (e.g. disabilities), context of use, etc. User observation is the most commonly used technique. This enables the tester to measure the user's level of performance and opinions of the interface. Users can either be observed in their work environment or in a usability lab.

What is the Cost Benefit of Usability Testing

Usability testing does not have to break the bank. Data collected from 863 design projects show that usability costs were between 18% and 13% of the total project budget. Based on these findings, Jakob Nielsen (a usability guru) concludes that current best practices call for devoting an average of 10% of a projects budget to usability. The cost of usability doesn't increase linearly with project size. Many usability activities cost about the same, or proportionally less with bigger projects. A project that is ten times bigger than Project X, for example, typically requires only four times more usability spending than that for Project X.

Usability testing offers many benefits to all stakeholders:

  • The Business

    Lower design/development cost and faster development timelines than projects with no usability testing. Due to the fact that products look better and are easier, faster and more effective to use, there is an increase in sales. Customers prefer products that are easier to use, which will certainly increase your competitive edge.

    • Online customers spend most of the time on sites with high usability (Nielsen, 1998). A majority of customers are loyal only to a few brands (1-3) and buy repeatedly on these sites (A.T. Kearny, 2000). Satisfied customers lead to increased sales for both e-commerce shopping sites as well as for software products. Satisfied customers stay longer on sites and consume more.

    • 82% of users attempted to purchase but gave up as a result of poor design and usability (A.T. Kearny, 2001).

  • Project Management

    Risk management is improved as it lets you test designs and decide which should proceed, long before coding begins. Planning is simplified; allowing you to calculate required build times and eliminate the need for rework due to incorrect design and miscommunication. Detailed reports, ratings, recordings and video confirm confirm the validity of the design at the earliest stages of the project, which provides evidence of success early on.

    • 80% of all software lifecycle costs occur during the maintenance phase (Mayhew & Bias, 1994).
    • Most maintenance costs are associated with "unmet or unforeseen" user requirements and other usability problems (Pressman, 1992).

  • Design Teams

    You will have more accurate designs and proof that the designs work, as actual users validate them long before they are built. Even before you start coding, you know how users might use the interface. This also results in less stress at the user acceptance testing phase, since designs are tested before they are built.

    • The cost of making a design change is least during the user-centered design phase. The same change would cost ten times as much during the development phase and one hundred times as much after product release (Mayhew & Bias, 1994).
    • American Airlines reduced development costs by 60%-90% by making corrections in the design phase (Mayhew & Bias, 1994).

  • Support Teams

    The cost of training material is reduced because there is less to document and less material needed. The training time is reduced, which means more hours of productive work for the support teams and the users.

    • End user training for a usability-engineered internal system was one hour compared to a full week of training for a similar system that had no usability work. As a result of usability improvements at AT&T, the company saved $2,500,000 in training expenses (Mayhew & Bias, 1994).

  • User Groups

    Task time and errors are reduced, increasing user performance and productivity. The staff turnover may be reduced due to higher staff satisfaction and motivation.

    • A major computer company spent $68,000 on usability work to improve a system that was used by over 100,000 people. The resulting productivity improvement saved the company $6,800,000 which is a cost benefit ratio of $1:$100 (Mayhew & Bias, 1994).

What effect does bad usability have on a business?

  • Almost 50% of users do not come back to a web site if they found it hard to locate relevant information (Forrester research, 1999).
  • 43% of purchase attempts failed (Creative Good, 2000) and 35% failed in another similar study (User Interface Engineering, 2000).
  • Research acroos all sites showed that visitors could not find what they were looking for as often as 60% of the time (Forrester and Juniper, 2000).
  • The results for e-commerce sites are devastating, and lost sales due to bad design and poor usability represented a $3.8 billion loss in sales for the year 2000 (A.T. Kearny, 2001).

I think these facts are more than enough to convince anyone that functionality alone does not make a system. If you are not yet convinced, why not visit the http://www.usabilityorg.net and http://www.useit.com websites and read up some more?

Conclusion

Roget's Interactive Thesaurus, says that another word for USER is CONSUMER or CUSTOMER. Now, maybe this is where we've been missing the point. By realising that our users are, in fact, our CUSTOMERS (regardless of whether we are developing in-house systems or systems for clients), we should take a whole new approach to developing. Just imagine yourself walking into a grocery store. As a customer, you expect:

  • The displays on the shelves to be clean, neat and uncluttered.
  • Similar products, logically grouped together.
  • A variety of products to choose from.
  • Wide aisles to prevent you from banging into other customers or into the shelves.
  • Staff to be friendly and helpful.
  • Quick and efficient service at the tills.

Shouldn't your "customers" feel the same way when they experience your product?


Antonet Bekker

 
<< May / Jun 2003
September / October 2003 >>