Benchmarking Websites Workshop

28th June 2002

Feedback / Discussions

 

Aims of the session - Identified at the start of the workshop:

 

  • Ammunition – show colleagues how we are performing
  • Accessibility – Tools and Skills to prepare for the SEND Act September 2002
  • Establish what needs to be done for the way forward over the summer.
  • The knowledge to challenge marketing and others  - regarding design / functionality.
  • Tools to use like: Bobby / Netmechanic
  • Make friends share ideas and best practices.

 

 

 

Some very positive comments received  - and for those who missed the afternoon session here is a quick summary:

 

Summary of the Afternoon:

 

13.00 – Hands on session continued – exercises and tasks completed.

13.30 – Report Back  - here we looked at the each of the exercises and recorded the highest and lowest scores and discussed how different tools gave different results also how these results can be interpreted.

 

Report Back / Discussions – sharing best practices:

 

Below are the results of our group discussions this is not intended as an exercise to show who is best  - but to show how diverse and different our websites are. The exercises and the results help to identify what actions need to be considered when designing and building websites – also how regular auditing can in fact improve how users access your website:

 

Size of your home page:

 

Biggest Page Size

Smallest Page Size

56k Download Time – longest

56k download Time  - Shortest

218 kbs

15.5 kbs

32 seconds

4.19 seconds

 

 

 

 

 

Comments: We found that different tools gave different results  - this is due to the ‘exclusions’ on each individual tool. For example: exclusions for web BOTS on certain directories like the images directory can affect the results. Jakob Neilson  - has also stated that the time taken for web pages to load can be used as a useful indicator as to whether users will stay or give up.

 

 

 

 

The Look of your Home Page:

 

Home Page

Group Result - Comments

Traditional Menu Structure

The majority of us opted for the traditional menu structure for our home pages. An approach that works!

Changeable Page, with News Feeds

None of the sites actually used dynamic features like news feeds on their home pages. This adds a refreshing - constantly changing feel to the website – worth looking into RSS news feeds ?

Personalised Pages

Again none of us personalised our websites. Dynamic pages can add functionality to a site and make it more interactive and engaging – also user friendly.

Splash Screens

Some sites use splash screens. They don’t really add any functionality to the site and can sometimes be viewed as another obstacle to the information you are getting too. If used, try and give navigational choices as soon as possible and try not to use too many images as they serve little or now purpose + slows the download time over a modem.

Spawns in a New Window

Add no real functionality to the site – and many frown on this type of practice. Brian picked out Derby Uni as a culprit: http://www.derby.ac.uk/ (try the links on the right-side).  But having said that, copyright issues of linking to external sites can be an issue. Sometimes it can be beneficial to let users know they have linked to an external site by spawning a new window.

Pages requiring specialist browser functionality (e.g. plugins, Java support etc.)

3 of the websites had used Flash, which poses accessibility issues  - providing an alternative text format is an option but this can prove difficult to maintain. – Unless you investigate tools like Betsie - http://www.bbc.co.uk/education/betsie/about.html

Alternatively loose the flash. Plug-ins can force users to change things or download other products; again, the chances are the user will not bother to come back if they can’t use the site.

 

 

 

Your Organisation’s 404 Error Page:

 

Comments: Unfortunately none of the sites tested had customised their own 404 pages – this is something each of us could do quickly and easily. It needs to convey meaningful information to the user  - fit in with your house style  - provide navigational aids or even a search field.

Here is a link to Brian Kelly’s 404 article:

http://www.ariadne.ac.uk/issue20/404/

Accessibility and Validation:

 

Having used all the various tools like:

 

 

Comments: All the sites but one had priority 1 status errors using Bobby testing tool – however the general consensus was WAVE was easier to use and understand. The reality is we all fall short of the W3C compliance standards / guidelines: http://www.w3.org/TR/WAI-WEBCONTENT/ -(WARNING - this is a bit intense).

Accessibility and the information on this issue is overwhelming but no clear cut checklist list exists  - here is a link to JISC resources and information on the subject (also see the TechDis website):

http://www.jisc.ac.uk/general/faq.html#access

and a link to Webaim.org  - this covers ‘Section 508 of the Rehabilitation Act’– a US feel to the website and offers a checklist of the US requirements for the above act – I am sure our recommendations can’t be very different. Also some other useful accessibility info:

http://www.webaim.org/standards/508/checklist

 

 

Link Alarm was also another useful tool – broken links can give a bad impression.

 

Comments from the RSC

Accessibility is a concern and not many were aware of the event held in April about Accessibility and Inclusive Learning here at the RSC. Many wanted a checklist of the things they must do or comply with. The reality is no checklist of this type exists and each website needs to address its own unique problems.

 

Statistics and Web Popularity:

 

Comments: One college had 381 links to it from a given search engine and another had 34 links – as with all these results the figures are subjective and not sure if links from their own sites were excluded? – Links to the homepage or deep site links? – Figures like this can be used as ammunition but can be misleading – use with care! Search Engine listings can take sometime newly created sites will not be listed immediately  - sometimes up to 6 weeks.

 

Comparing Your Results With WebWatch Surveys

 

Survey Question:

Results and Comments

No.s of Servers

Mainly one web server used. However, as the influence of the web grows and the introduction of VLE’s the situation could soon change.

Search Engine

3 of the sites we looked at did not use a search engine on their websites. Sometimes users can be put off having to trawl through a site to find information  - a search engine can make things a lot simpler. This becomes an important issue when a website begins to grow. There are many free web resources that can be used to quickly and easily set-up a search engine for your website – ATOMZ – is free and comes with a Dreamweaver extension: http://center.atomz.com/dreamweaver/  - very easy to set-up. Other services like Google also provide similar facilities. Different tools very in the way that they index files and the way they treat dynamic pages.

Size of Home Page

The larger the page  - the longer it takes to download. Its important to keep file sizes down  - especially when images are used on a page – remember ALT tags!

What’s Related

The results from the tools used varied and this is largely due to the parameters used by individual tools. We found that two different tools gave different page sizes for the same websites. If these tools are used and the results interpreted correctly by comparing all the different tests and using some common sense  - you have all the ammunition you need. This type of benchmarking needs to be done on a regular basis as  - websites are constantly updated and rapidly changing. New languages and developments in the field mean you need to keep your skills refreshed. Stress testing the website and checking the security of the server – as this is your door the outside world and from time to time you will get unwanted guests knocking the door.

Web Server Software

IIS  - most common, UNIX/LYNX – RED HAT, SQL Server

Links to Site

Search engine listings are important and tagging your pages properly can improve your rankings depending on the descriptive words used.

404 page (what’s missing)

None of the pages we looked at addressed this issue. Navigational aids and possibly a search field on the 404 page was recommended.

 

Comments: None of the Colleges had actually paid a commercial company to do any benchmarking on their websites.

 

Conclusions / Comments from the RSC

 

Above all  - I think benchmarking of this type should be done on a regular basis at least one a year. Results from these types of tools can be subjective but they can be used as indicators  - ‘how long it takes on a 56k modem’ – or – what rankings do I have with different search engines – and why? – Customising your 404 page. We all fell short of the W3C guidelines on accessibility – a number of issues were raised:

 

  • Does a checklist exist of things to do for September 2002?
  • Clarity regarding content?
  • RSC – involvement?

 

How can the RSC assist with this issue? – Some suggestions raised were:

  • Having RSC consultants doing a benchmarking / audit workshop of regional websites – (I am investigating the possibility of this type of workshop.)
  • User testing with an experienced screen reader software user – or member of the RNIB to test individual sites and make recommendations – (again investigating.)
  • Doing an Annual Benchmarking websites workshop – doing this as a collective means we can share ideas and it gives you time out to do the testing  - again more up to date ammunition!

 

The comments received after the event were very positive also any comments on Accessibility are welcome on the online forum:

http://forum.jisc.loucoll.ac.uk

 

Regarding email lists – I think the last thing any of us need is another email list cluttering inbox’s - but there is an existing JISC email list servicing HE and FE Web people nationwide.

Browse the archives at:

http://www.jiscmail.ac.uk/lists/WEB-SUPPORT-SITESERVER.html

 

If you would like to subscribe to the list (some in the region use it already and I have found it invaluable) – fill in your details on this form:

http://www.jiscmail.ac.uk/cgi-bin/wa.exe?SUBED1=web-support-siteserver&A=1