Web Governance is more than checking broken links …
Web governance has been developed to ensure that an organisation’s online presence can fulfil its objectives. This role is usually interpreted as maintaining website ‘quality’ by meeting internal style standards, ensuring brand consistency, eliminating page errors and ensuring high levels of accessibility. These elements are important, but they are means to an end, not ends in themselves.
Higher education websites exist to meet their stakeholders’ strategic objectives and they do this when delivering the best visitor experience. We generally measure service delivery success via customer satisfaction (the visitor experience) and the same holds true for websites. By measuring customer satisfaction we can understand which elements of a website we might need to change to improve or maintain performance.
Recognising the key visitor experience factors in turn helps shape the organisational structure needed to achieve the desired outcomes. By starting with the outcomes and working backward we can identify who needs to have accountability, what roles contribute to success and where authority for decision-making should reside. As Active Standards states: “Having a well-designed digital governance framework minimizes the number of tactical debates regarding the nature and management of an organization’s digital presence by making clear who on your digital team has decision-making authority”
Adopting an outcomes-based approach focuses on the why of web governance, lessens the chance of being caught up in the what, and broadens the appreciation of the website maintenance elements that need to be addressed.
Less is More …
One visitor experience element that web governance often overlooks is page-loading speed. Higher education website visitors appreciate alacrity and they increasingly appreciate sites that are speedy on mobile devices. Page loading speed is influenced by both content and technical implementation factors. This makes it a good use case for developing an appropriate web governance structure, as optimisation is achieved by information technology (ITS) and content creator collaboration.
To develop a sense of how well current web governance structures might be working we examined the loading speeds for the home pages of Canada’s U15 universities. For good measure we also included U15.ca’s home page as well. In fact, we tested seventeen home pages as one U15 member has both French and English home pages.
We’ve anonymised the results. However, you are welcome to contact us for details or you can simply re-run the tests, as detailed.
Page Loading Speed
We used WebPagetest’s website performance software to emulate a visitor’s experience connecting to each site’s home page via broadband and a 3G wireless connections.
To normalise the results we simulated visiting each site three times and used the average values. We followed this process twice: once for broadband and then for 3G. Our simulation measured the experience of a user in Phoenix, Arizona – not equidistant to all sites, but not too far off being so.
The same tests conducted on or adjacent to a campus will undoubtedly show faster page loading times. As major universities attract global audiences, optimising for a more typical response will yield benefits for all visitors.
Chart 1: Broadband Home Page Loading Time in Seconds.
These results summarise the page loading times for a visitor using a broadband connection.
The average page loading time is 8.2 seconds, however, pages typically start to render at the five-second mark.
Chart 2: 3G Wireless Home Page Loading Time in Seconds.
This simulates the typical response for a visitor accessing the home page via a 3G wireless connection.
The average page loading time is 16.6 seconds. Pages typically start to render in 8.5 seconds.
How Can The Visitor Experience Be Improved?
On the reasonable assumption that website visitors are more satisfied with sites that load rapidly than those that do not, are there changes that could be made to improve performance?
The simple answer is yes – some changes are technical others are content related.
Web Server Configuration
A material impact on page loading speed can be obtained from turning on so-called server compression: often called gzip or deflate. Roughly forty per cent of the websites in our test group were not using server compression at the time of testing: implementing it will automatically shrink compressible files and reduce the amount of data being served to the visitor.
Chart 3: Server Compression Implementation.
Using the web server’s inherent gzip file compression functionality will reduce the total data transfer burden and browsers understand how to process files that are transferred in this way.
Image File Compression
An equally effective approach to reducing the data to be transferred each time a page is displayed in a browser is to ensure that all image files are compressed to the smallest size compatible with the required image quality. Of the seventeen sites examined in our review, only two had fully compressed all the home page images. By using image compression, the remaining sites could reduce the total data burden on loading the home page by an average of 568KB (0.5MB).
We note that site ‘7’ had the slowest loading speed over both broadband and 3G connections: compressing images would reduce the size of the home page by 3.3MB.
Some experimentation is required to ensure that image quality is not lost on compression, but free tools such as JPEG-Optimizer and TinyPNG can be used to make image file compression a routine step in adding new site content.
Chart 4: Image Compression.
The chart shows the total amount of data that can be saved by compressing the home page images.
Image compression should be a standard component of the process for adding new site content.
Content Distribution Networks
WebPagetest is exacting in determining whether a site makes effective use of CDNs, as it only considers this to be the case where 80% of the static resources are being served from a CDN.
On that basis only one of the sites we examined uses a CDN effectively. We note that this site went live in the last month following a substantial re-design.
And, Finally Broken Links
We didn’t say broken links don’t matter, we observed that the visitor experience is more than making sure that links aren’t broken.
Just to be clear, even checking for broken links is only part of the story. All the resources (images, stylesheets, icons, authors, etc.) on a page are there for a reason, so a checking process should examine if all resources can be reached not just other website pages.
We submitted the seventeen home pages to our higher education website diagnostic service’s link reviewer to find nine broken links, five links to other website pages and four links to image files.
Web governance is a methodology for managing websites. It is not an end in itself and following a set of governance rules will not ensure that sites achieve the outcomes they exist to meet. But, by paying close attention to a site’s visitor experience it is possible to focus on the why of website maintenance and the appropriate organisational structure can then be put in place.
eQAfy is a subscription risk assessment and quality assurance service that helps complex, content-rich websites be error-free, deliver their intended user experience and meet their marketing and communications objectives.