After spending a day performing about 500 Google searches of various businesses for a recent project, we made the observation that many organisations are not optimizing the metadata tags in their websites for their users; namely neglecting to use metadata to communicate the nature and content of the website and to encourage people to click on their link.

Example of Metadata

(Note: don’t know what metadata is? It’s essentially ‘data about data’, and is entered into the head tags of your HTML content. It is made up of a number of tags, including “description”, “keywords” and “language”. Search engines essentially read metadata to ‘understand’ what a page is about).

Our observation about metadata was inspired by the search results of a large business, who had not edited the descriptive metadata tags in their website after it was created with a template, leaving the ‘lorem ipsum’ filler text to display in their search results. Metadata is the foundation of all information retrieval processes and its importance in good user experience should not be taken lightly. It is possible that certain metadata tags have been both neglected or abused over the years, due to aggressive black-hat SEO tactics, or theories about SEO and the relevance of the metadata tags. Despite the rumours and the SEO myths, metadata is still a critical step in correctly identifying and defining the content and structure of a website.

Descriptive metadata entered in a website source code is important because it is the first port of call for users to gain an understanding of a website (this is the metadata entered in the meta “description” tag of your website code). It is the text that is displayed in search results and is generally all a user will refer to when they determine if they want to click on that link. If the metadata gives the users false expectations of the content of a website, than user dissatisfaction can occur. Users do not want to have to click around to find information they expected to see elsewhere, or be taken to a dead end. Descriptive metadata needs to be a brief, accurate and keyword rich description of the site’s content. No more than 250 characters will be displayed in search engine results, so keep it short and to the point so users don’t miss out on vital content.

Metadata within a website should also be consistent; a unified metadata structure also indicates a clean and intuitive taxonomy within your website. Using a metadata vocabulary such as Dublin Core can ensure your website’s metadata is thorough and aids in content retrieval and management. When evaluating the information architecture of your website, look beyond your navigational structure and ensure your metadata is also intuitive, relevant and matches your target audience’s expectations of content and classification. Good information architecture structure encompasses all elements of the page, including those the users don’t necessarily see.

I recently purchased a new mobile phone from a major Australian Telco, at the same time I moved from a pre-paid service to a contract, all with the same telco. Unfortunately, all did not go smoothly and this led to numerous interactions with customer support over a one-week period. Throughout the process I experienced a number of frustrations, all of which were exacerbated by having to engage with an Interactive Voice Response (IVR) menu each time I called customer support.

Don’t get me wrong, I think that IVR is a great tool, and has many brilliant applications. However, what made this Telco’s implementation of IVR frustrating was that rather than providing the caller with a number of options to select from, the system initially asks the caller to explain the reason for their call. This assumes that the caller already knows the options that are available to them and the words/phrases that they will need to use in order to access the correct menu option.

In my opinion, this is like telling me that I need a key to unlock the door in front of me and rather than providing me with a set of keys that I can compare to the keyhole I am given some materials and told to go ahead and make a key a key that fits the lock. What makes it worse is that I HAVE to go through this particular door, there are no alternatives. Best of luck.

Here is an example of my first exchange with “Lara”:

Telco: “Hi, I’m Lara, welcome to [Telco name]. I’m here to help you with your call today. All you have to do is talk to me. So using just a few words, tell me the reason for your call.”

Me: “I have ported the number from my old pre-paid sim card to my new contract sim card. I am now receiving calls to my old number on my new handset but my handset indicates that the phone number of my sim card is different. I suspect that this may be why I can’t access or setup voicemail.  I would like to get this fixed. ”

Telco: “Sorry, I missed that. I’ll understand you better if you use just a few words.”

Me:  frustrated pause…”issues with my new sim card and voicemail”.

The system then latches onto some words in my statement that relate to their menu structure and offers me what it sees as the most relevant option. I have no idea if this is the right option as I don’t know what all of the options are and I won’t know until I speak to someone, quite often after a prolonged period on hold. As a result, I frequently found myself in the wrong department with no idea how to get to the right department. The very friendly call centre staff often transferred me to another department, but due to the Telco’s use of VOIP, the call repeatedly dropped out before I got to speak to anyone and it was back to Lara to start all over again.

All-in-all, it leads to an extremely poor user experience which tarnished my perception of the brand and has led me to actively recommend family and friends avoid this Telco. My experience serves to highlight how menus provide invaluable orientation for users. I suspect that my frustration would have been greatly reduced if Lara had started off by providing me with a list of options to choose from. Perhaps even more importantly for this telco, I would not have clogged up their customer support line with my queries for a week if I had been provided with these options up front. You have to wonder how much money they are spending on servicing customer support calls which could be avoided by simply offering customers a list of options to choose from at the start of the customer service call.

Organizations are always under tremendous pressure to keep ahead of the competition. As a result, they explore novel ways to shorten their product development lifecycle. Most often it is the various testing phases which bear the brunt of the cuts in time and resources.

Due to budget and time constraints, usability testing is often left out of the process altogether. Some organizations mistakenly associate usability testing only with the end product, which is actually contrary to the principle of user centric design process. Introducing usability testing (particularly the rapid iterative form, as explained below) at the early stages of product development can in fact save organizations from facing unpleasant and time consuming surprises later.

When your organization has an aggressive launch schedule ahead, rapid iterative usability testing can support your launch quite effectively if conducted during the appropriate phases of product development. As the name implies, it is a rapid/quick process and is done in iterations.  The end result will typically be as if you conduct two rounds of usability testing before the product even hits the market – all without sacrificing time and slowing your production schedule.

The success of the method rides on few critical aspects;

1. Don’t save the best for last, start usability testing early. A working prototype could be an excellent candidate for testing. Utilize paper prototype in case all the sections of prototype are not functional.
2. Besides the testing team, the product team’s involvement in this exercise is critical. Product management and development teams should also make themselves available to view the sessions. This will ensure that the team understands the severity of usability issues and also be able to quickly decide on major amendments to the product.
3. Sprints should be carefully planned to accommodate the time required to make changes on the prototype.
4. A rapid testing process does not mean the number of testing participants should be reduced – test with too few participants, and you risk injecting bias into the process.
5. Setting the expectations of the participants is crucial especially when prototypes are used. This will ensure that participants do not get too focused on the shortcomings of the prototypes.

At the end of the session, the testing team develops an issues list to highlight the problems faced by the users. The issues list should be discussed with the product development and management team through a group discussion to ensure that changes are prioritized prior to solutions being explored.

Breadcrumb trails, those links you see at the top of the page that often start with text such as, “You are here:”, provide a number of benefits to users and website owners alike:

• They act as a signpost to where the user is within a site. This is particularly beneficial to users that arrive on your site via a search engine. The breadcrumb trail allows these users to identify what section of your site that they are in, as well as providing a quick preview of the type of content the site holds.
• Breadcrumb trails also allow users to move back to any page in the trail with a single click, rather than multiple clicks of the browser back button.
• They also provide some SEO benefits and Google is now including breadcrumb trails in their search results.

The name “breadcrumb trail” suggests a path that the user has taken to reach their current location (think Hansel and Gretel) and this is certainly true when considering a very linear user experience. However, that linear experience is no longer truly representative of the way in which many users interact with websites, particularly content rich sites where users have a high level of engagement.

Consider the following example, a user arrives at the home page of a news website, they click on the primary navigation option Entertainment, then a sub-heading Movies and then an article. Up to this point the breadcrumb serves the users well in reflecting the journey.

It is quite likely that in order to increase the amount of time that the user spends on the site, in our example above, users are encouraged to click on some form of related content (e.g. blog, video, image slideshow). Typically, this content is located in a different section of the website which results in a resetting of the breadcrumb trail. This suggests that despite its name, the breadcrumb trail is actually intended to demonstrate the location of content within a site’s IA, rather than how the user has arrived at a particular page (insert image).

What does all of this mean for the implementation of the breadcrumb trail, and recommendations we make as user experience professionals? Although our experience suggests that only a minority of users will use breadcrumb trails we still feel that they have a role to play on a website; however the definition and application of that role is a lot more ambiguous than it used to be. What do you think?

Of late, there have been discussions around the office, and with some of our clients, as to why in one-on-one user testing we get users to begin each task from the homepage. Following on from the previous blog post ‘Is your homepage, your homepage?’ there are arguments that not all users arrive at the homepage, so getting them to begin from there is not representative of a real world scenario, particularly since Google and other search engines encourage users to arrive deeper into the site.

That being said, even though the majority of your users might be landing deeper in your site, it is not to say that particular content should not be easily accessible from the homepage. Often when users are browsing a site they will use the homepage as a base to access different areas of your site, continually coming back and using the homepage as a reference or starting point. So even though users may have initially arrived deeper in your site, there is a good chance that during their visit they will click through to your homepage in order to locate other content.

Generally speaking, we feel that tasks presented during user testing are largely independent from each other and we encourage users to approach each task as if they had just arrived at the site. While it may be appropriate to base the first task of a testing session on a page other than the homepage, in most cases, we recommend that subsequent tasks begin from the homepage. The homepage provides a logical reset point from which users can begin their next task and provides an opportunity to test the effectiveness of the homepage and site navigation.

In testing recently, we came across a user behaviour that was challenging, not so much in itself but more so for its impact on the category and how it needed to be taken into consideration during website development.

Taste.com.au Recipe

One of our Clients was in development of a website that targeted aspiring cooks. In a nutshell, they wanted to confirm that their planned website was in line with user needs, and also that they were on the right track with the specific functionality that had been developed.  We used focus groups and moderated one-on-one user testing on the project.

During the focus groups, the importance of search became very clear. As mentioned above, this in itself was hardly ground breaking. The intriguing bit however was how much search engines (and by default, Google) had become the de facto homepage for this particular category. User behaviour was to go to Google and search some aspect of cooking, and to proceed to the websites recommended in the natural search results. The use of Google was so much a part of the process that the users were nearly universally unable to name any of the websites which were dominant (by traffic counts) in the category. The users knew the sites only by the pages that they had linked to via Google, which tended to be deep within the individual sites.

The behaviour was so prevalent that it became clear that our Client would need to rethink their homepage design and information architecture. Specifically, they had designed the site so that some of the core features were only clearly visible from the homepage, and were less easily accessed via the individual pages of the site. Understanding the user behaviour in this category led us to recommend to the client that they rethink their content pages, and consider how each one was effectively the ‘homepage’ of their site for most users. They needed to reconsider how to dial up the site’s branding, and how they could introduce some of the other features of the site (which had only previously been available on the site’s homepage) to the users without getting in the way of the core purpose for the users’ visit.