56! Wonderful number. 56 is also the number of links available on the gov.mu homepage without touching anything. These links make nearly 60-70 percent of the gross content of the home page. On a first view, it’s all about links and from a usability point of view, a visitor does not even spend the time reading all the available links on this page.

One of the great answers that some project managers are fond of is: “If they’re on this site, they’re looking for something, let them do the work now.” Here there is more to it. The visitor has to battle through the content to find data, if it ever finds it.

Decluttering the navigation is not easy task on this site. To be able to do this, we’ll keep in mind the main target for the site: the Mauritian population. Thus, the elements (as they are right now) must be organised in groups targetting the most used sections of the website. Having no access to the analytics tool of the site we cannot really know which sections are the more important. We can however make a guess on these.

Process.

How would this process be followed in a real life project?

• Use an analytics tool and make a report on a whole year.
• Targetting the main audience, extract information about the most viewed pages.
• Determine what are the completed goals for these pages (downloading a file, filling a contact form, filing a report or a demand…)
• Check if there are any conversions or determine how to follow them in the tracking.
• Use these pages as a base for navigation, at least section building.
• Determine section groups and target landing pages as main section pages

One navigation menu or more?

Looking into the navigation system actually used on the site as well as the sub-portals, the site’s overall navigation is separated into 5 parts dispersed around the page. Below is a screenshot of all the other navigational elements.

A good experimentation would be to group all the main navigation in one place and keep one and only one for less-important links such as disclaimers and “about” pages. On the main navigation, experimenting with mega menus might be the best alternative to the clutterred navigation problem. This would add direct links to specific pages in one place and improve user experience.

The mega menu would have to be above the fold and the second “lesser” navigation in the page footer.

Another interesting experimentation to hold on navigation on this website would be to display a block of information by topic to help finding the information instead of using the drop down menus currently in use.

How to link?

A very important problem crops up when analysing the links. All the bottom page ones are javascript links. Being a government website, the most important factor in the use of the website is easy navigation. Using javascript generated links is one of the worse errors to make in web implementation for 2 main reasons:

• Non visual browsers will have problems following them as they are non-javascript readers. An “accessibility” link written in javascript shows that accessibility is actually NOT the main objective of a site that needs to have the largest audience possibilities.
• Search engines will not be able to crawl the pages. SEO (from a business-oriented point of view) might not be a core necessity for such a site but it is important because SEOing it will allow it to have direct links available in search engine results pages, these being what drives people to the information they are looking for.

It is therefore important to check all the links and set them in href (HTML). Furthermore, adding title attributes to them would improve the SEO and facilitate navigation. Anchor texts with keywords will also help navigation as well as SEO in one go.

Conclusion

Navigation is a core element of user experience and experimenting with better content delivery on a site like a Government’s website would be beneficial to all. Improving and decluttering this will have a positive impact on browsing. Experimenting with different solutions to find the best one would not be throwing money out of the window.

We’re in for some fun on this new experiment. Before digging too deep, let’s see how are the current state of affairs. First things first. The major target for the site being the Mauritian population. On its “About E-Government” page, the site states:

The vision of the Government is to provide an effective and efficient delivery of services, on a 24/7 basis, to citizens as well as to the business community. In this respect, the Government has invested in the necessary infrastructure, namely, the Government Online Centre and the Government Web Portal as a gateway to provide Government services online.

This calls for some thought. Delivering quality information first comes through delivering it easily. How are these services delivered then?

The loading time question.

The government website is a public service, hence needs to be effective and practical but the first thing that many users see of this website is the time it takes to load. So let’s dive into this.

Swift page loading is a must today. According to usability guru Jakob Nielsen, web users are ruthless, they have very little patience and page loading time is a huge drawback. (Source)

“People want sites to get to the point, they have very little patience,” he said.

[...]

Web users were also getting very frustrated with all the extras, such as widgets and applications, being added to sites to make them more friendly.

Such extras are only serving to make pages take longer to load, said Dr Nielsen.

Consequently, people get bored with the loading time of websites and end up not using the tools built for their own use. We should also take into consideration one important element, the evolution of bandwidth. With more and more bandwidth available, less time is needed to load a page. In the late 90s, a decent load time was tested and set at 7 seconds for a page on a 56 kb/s dialup. Today, this is measured in milliseconds and under 4 seconds for a page.

Load time testing.

So we did a small research on the website’s loading time. The government’s website has been tested on Pingdom Tools. On a T1-ADSL network, the site takes between 13-14 seconds to load. The full test is available here.

To get a clearer view, some Mauritian web users have been asked to test the site from different locations on the island. Here are the results:

Around 15 seconds on Orange ADSL 512 Kbps @ this moment(1.28 PM) on Chrome! Fond du sac North!
Source : on Twitter

Takes approx. 18sec to load www.gov.mu via Firefox/ Emtel Wimax
Source : on Twitter

My.T 512k. Actual speed is around 90kbps | Firefox 3.5.7 – 94s | Opera 10.10 – 102s | Location: Triolet
Source : on Twitter

The good, the bad and the ugly

The good thing is that, depending on the bandwidth, the site meets only nearly twice the late 1990′s recommendations, it might have been really bad. Who are we trying to fool here ?

The bad thing is that, on the actual standards of bandwidth consumption, the site is 4 or 5 times slower than what is usually seen around the web nowadays.

The really ugly thing about all this is the way local users have not been taken into consideration as well as availability. How? Some points discussed online.

Many Mauritians don’t have ADSL/My. T & Emtel. I expect it would take more time(over 4 mins) when being used on dial-up.
Source : on Twitter

A gross estimate is that there are around 400k Internet users in Mauritius… & around 60-70% of them are dial-up/Nomad users.
Source : on Twitter

As an aftermath, more than 50% of Mauritians need to wait over a minute to load the first page. Now that’s what cannot really be called service.

Sources

This loading problem on the website might have different sources:

• Server capabilities. The server can have difficulties delivering pages. Modern browsers also accept Gzipped pages, thing that the server does not actually do.
• Code. Messy code is present everywhere and each page is a great example of what should not be done.
• Seperation of content and layout. All are tightly mixed making it difficult to read for any browser.
• Excessive use of unnecessary javascript.
• Javascript not loaded in external files.
• Technology used. It seems that the portal runs on a java or ASP cms and that no page is optimised in terms of code. (Can someone confirm in the comments please?) Standard fast PHP and MySQL might be an interesting try.

SEO Friendly?

Ok, lets admit that some persons care about SEO for such a portal, why would it matter? Simply because people need to find the information. It comes to the quality of the information delivered. For the time being, no real user-centered information is delivered. Just have a look at how the site’s whopping near 60 000 pages look like when indexed in Google.

An example of SEO necessity can be a person having to re-issue his/her id card. Typing the term “mauritian identity card” in Google gives this:

Another fail maybe, but why take this into consideration? Google Webmaster Tools has been updated to include page performance tools. One new criteria in site crawling is now SPEED. This is linked to the load time problem. Another thing is that, content remains king and is a better king if easily found through search engines.

An idea?

An idea just popped like this some days ago. It was a simple, somewhat crazy but cool idea which might have some great impact, at least some minor incentive, on the Mauritian Web Design field. This crazy thing is currently called the “2010 project”. A better name might be found later on. Now, the cool thing about this idea is that this project will have a collective aspect. Yes COLLECTIVE in the sense that the Web Design Bureau’s (small handful of) readers will be associated to the project.

Why?

Why associate the Mauritian readership? Because the project concerns them, the project’s subject is some kind of representation of Mauritians all over the world. Yes you are right! You’re getting the idea of this project and what it will consist of. And you will participate!

What?

As said earlier, your representation on the web: the Mauritian government thing (yes, you can’t call that a website). The idea is that, as time goes by, more and more information is being added to this site. It is currently impossible to know how many pages there are in that site, how it is really organised and what SEO/SEM strategies are applied to it. I tried a spider simulator on it and ended with an #epic #fail (hello Twitter world) with a lot of 404 errors and a whopping 69 links in only one page while getting you lost in there.

So here is the project. It’s not said that it’ll be a killer success but it at least is an attempt to provide and alternative. The project is therefore a redux of the site.

How?

I have no idea yet but the concerned persons are Mauritians first. So we need to address the Mauritian public, well the site needs to address the Mauritian public. This is why major decisions on the design will be set as polls open to the readers. All ideas will be welcome and active participation will get interesting. No code will be generated, it is only a graphic design redux of the site just to imagine how it would have been if the right questions had been answered when necessary.

Let’s do this!

Ok, this project can be a huge flop but it has the advantage of budding (at least). The idea is not to take the place of anybody nor steal the job off anybody either. Just the fun of trying a real project management challenge on an equally challenging subject along with the web friends.