Cabinet Office
The web is a very different medium than the printed page. Data on the web is not structured like a printed book, as links enable users to go from one source of data to another in a completely transparent way.
Readers must be able to orientate themselves immediately, know that the data can be trusted and find it easy to understand. Most importantly it must be demonstrable that the document is up-to-date.
For this reason it is important to ensure that the document has been constructed correctly with the user in mind.
Use each checklist to ensure that your web pages comply with these guidelines
3.2.1 Checklist and summary: Core guidance
Checklist -
Summary -
Because web users can surf using other pages or search facilities as a first step, it cannot be guaranteed that they have come to your document through the front door.
Therefore each page in your website must have a self-contained identity and be capable of being seen as the first page. It should contain context orientating data for the user, such as:
consistent navigation to the rest of your site.
3.2.2 Structuring documents for the Web
By using file downloads, plug-ins, graphics and hyperlinks, there is no real limitation on what can be accomplished.
A site looked at as a whole may make complete sense and have a form of logical construction, but this picture can look somewhat different when a single page of that site is looked at out of context. This is the situation that many web users have to contend with on a daily basis.
There are many different ways of getting to a particular piece of information, and going through a website’s homepage to get to it is perhaps one of the less likely. Another site may have linked to this information, a newspaper may have given a particular page’s URL or a user may have located the information using a search facility.
Any one of these situations may mean that a user is introduced to your site without seeing your introduction. They will be unfamiliar with your navigation system and may not necessarily know who the owner of the information actually is.
The design of each page within a website must therefore be consistent, usable, and immediately identifiable with the information’s owner. Ultimately this means that each separate page of information should be seen as an island, existing on its own but identifiably part of a whole.
Each of these ingredients will affect the experience the user has while in your care.
Important: Do not use splash screens with automatic client-side redirection as an introduction to your site or to a section of your site. Some browsers still do not support this HTML feature and will therefore not automatically redirect. This just adds another click to a user’s journey through your site.
3.2.3 File naming
The file name component of a URL will be displayed in the address bar in the user’s web browser and will be exposed in the status bar at the bottom of the browser window when the user moves their pointing device onto a link in the document. For this reason it is better to make file names indicative of the document’s name or purpose. This will also help with the general housekeeping of the website file structure, as file names should be fairly obvious. For example, section three of the corporate business plan for 2001 is far more meaningful if called ‘businessplan01-03.htm’ rather than ‘350165.htm’.
Where file names include dates these should conform to the ISO and W3C standard, that is they should be in yyyy-mm-dd format. This ensures that file names have meaning in the long term and lists of such files will sort in a sensible order.
A number of conventions should be considered when naming your web files, most of which may well be determined by your hosting service:
3.2.4 File sizes
Large files are time-consuming and expensive to load, particularly for clients with slow modems. Large homepages have the added disadvantage that the user cannot choose a more economical means of accessing the site.
Different types of web page require different file size restrictions, thus:
In all cases, if the file size exceeds our recommended maximum, the user should be warned. Good internal navigation within a larger publication is also required (this is covered in section 4.2.4).
File size can be an issue because of the differing levels of formatting data contained within them. Depending upon the text-to-graphic ratio of a document, either a PDF or RTF file may be considerably smaller or larger than the originating document. Plain-text will always save with the smallest file size as it contains the bare minimum of document structure and no images at all.
Whenever these options are offered they should be listed at the top of the HTML document, directly under the document summary. The file size for each should be shown next to each file to inform the user of potential download time.
If a proprietary format such as PDF or Microsoft Word is used, a link to the reader software download site, using a standard form of words, should be included next to the document.
Subsequent sections of a long document should have a link to these downloadable formats at the very top, as many users may not have been introduced to the document from its homepage.
3.2.5 Sequential structure
A long document, for example an organisation’s business plan, may be 80 pages long in its printed format. This document would not transfer satisfactorily to the web as one file - so it is better to break it up into manageable sections, which may well translate into a number of separate HTML files reflecting natural breaks in the content.
As stated earlier in this section, a user visiting your website may not have been introduced to a document from its home page. A link from another site, such as a search engine or another department’s website, may have brought them to page 5 of a 10-page document. It must therefore be made obvious from the very start, where they are and how they can access other information within the document.
This can be accomplished by adding section numbering to the top of the page, to illustrate each element of the document in a linear sequence.
Consider providing a single file downloadable version of large documents, suitable for reading off-line.
3.2.6 Document identification
Any document or section of a document prepared for print will invariably be larger than what can be physically displayed within a web browser’s window at any one time. Most users will have monitors with a screen size of 800 pixels by 600 pixels, although many will have smaller screens.
Documents written in English are read left to right, top to bottom. A document displayed on a browser will always have the information contained in the top left-hand corner on screen when the page is first delivered to the user. It is very important to make best use of this space.
A browser has a scrolling function that enables the user to move up and down a page. At worst, the need to scroll through a document may discourage its use by users with a motor disability who find it difficult to use a mouse.
Each page should include a selection of internal links to different parts of the document that has been loaded. This will enable users to access important information as soon as they receive the page.
Each of the sections contained within the document should include a Back to top button and internal hyperlink so that a user can quickly return to the top of the displayed page.
3.2.7 Information to be included in every document
A user should always be informed that each publication contained within the organisation’s website is relevant and up-to-date.
This could be made obvious by the use of a document footer within each of the files, whether they be in HTML, PDF, RTF, XLS, CSV or text format.
Your approach could include the following information:
3.2.8 Version control of web documents
Making information available on your website is a publishing process and organisations have established procedures for content approval prior to posting. Because of the nature of web publishing some information, for example, press notices, can be produced in a very short time and changes or updates to existing web pages can sometimes be undertaken in a matter of minutes. It is therefore important that users are aware how current the information is that you are publishing.
Every web page should clearly display the date it was first published on your website. And this date should subsequently be updated if you amend the information on a page.
With information that is frequently updated web managers should consider maintaining a version control record.
The simplest record would be to track the published version as an HTML comment note - the HTML mark up which is ignored by browsers but which can be seen by viewing the source code. For example:
<!-- version control record of page /annex01.htm -->
<!-- first published 01/02/2001 -->
<!-- updated 21/02/2001, 06/03/2001, 09/08/2001 -->
The latest date recorded here should correspond with that displayed via the browser.