Content management may seem like a simple concept, but there are
hidden complications. Danny Bradbury reports
Many might argue that the Internet is the biggest step in the
dissemination of content since the invention of the Gutenburg
printing press. The truth is that the Internet is to content what
DTP was to the church newsletter. Just as vicars went crazy
designing ridiculous layouts for misspelled text and badly written
articles, so companies put out content on their Internet sites
without much thought for its quality.
A recent survey produced by NOP and sponsored by content
management software vendor Mediasurface, for example, interviewed
104 companies to find out whether they thought there was any
duplication of information across their Web sites. While 41%
replied positively, a high percentage of companies believed that
their Internet sites were important marketing tools, with 52%
placing the greatest importance on the quality of content on the
site.
Meanwhile, 47% of the respondents to the survey revealed that
they had no content management or personalisation software in
place, and of those that didn't, 75% said that they would like to
publish content to the site directly.
Content management tools can be used to ensure that text is
created and updated in a structured way, to check Web sites for bad
links, and to ensure that all of your graphics, audio and video
clips are stored properly for later retrieval. One of their biggest
goals, other than ensuring the structure and quality of text on a
site, is to ensure that content can be exchanged between different
parties, found easily by visitors to a Web site, and also made as
relevant as possible to those visitors. Consequently, many content
management systems also deal with content personalisation. In spite
of the proliferation of multimedia content on the Web, text is
still the most popular form, and it is also one of the most
difficult to manage properly, just as in the publishing industry,
text can be misspelled, badly constructed, or just plain
inaccurate.
Problems can emerge when text is created by a multitude of
people from different departments, all of whom will have different
writing styles and levels of literacy. Building editing processes
traditionally associated with the publishing sector can help
alleviate this problem, says Nick Gregory, vice president of
marketing at content management software firm Mediasurface.
"Digital ink never dries. It has to be constantly rotated. So you
have to think about a lot of authors for content," he explains.
Many of the better content management systems will make this
process non-technical, building workflow processes into the system
so that business managers can enter information and then have it
sent to an editor who collates and checks the information, for
example. Such systems can then be used to assign rights to specific
people. You may not want to let an author change content once it
has been posted on the site, for instance.
Another step in the content management chain that many people
ignore is localisation. English may be one of the more popular
languages in the world, but there are plenty of potential customers
speaking other tongues, and any good strategy for managing content
on a global medium will enable you to reach them. Gregory works
with translation companies that take content from one repository
and translate the information for clients. Although machine
translation (the automatic translation of text using computer
software) is faster than manual translation, it can produce literal
translations, producing the sort of text that you see in manuals
for video recorders translated directly from Japanese, for example.
Manual evaluation of translated content is vital, he insists.
Of course, not all of the information that you produce on your
Web site or intranet may be your own. Syndication is big business,
and companies such as Newsedge, a firm that has been aggregating
content for 12 years, will sell you it.
"It's important inside a big corporation or bank to reduce the
amount of information to the most relevant stories available on a
particular subject," says Jon McNerney, senior international vice
president at Newsedge. The company takes roughly 100,000 news
articles a day and puts them into a repository, where it can filter
duplicates and then deliver customised news to corporate customers
based on their predefined profiles.
But as the amount of content grows, finding what you need on a
Web site can be difficult. One way around it is to design your site
to be intuitive. In May last year, the Word Wide Web Consortium
(W3C) released the initial version of its Web Content Accessibility Guidelines
recommendation. This specification created a clear set of
expectations for the design of Web interfaces that would make the
navigation of information simpler for users. Concepts in the
specification include the use of style sheets to separate the look
and feel of the content from the content itself, so that designers
can define how content will appear on the page by changing a few
central parameters, without having to make any changes to the
actual content.
Search engines are another good way to keep your data
accessible. Sometimes, when there is a huge amount of content on
the site, end-users can find it daunting.
David Heath, European business manager for AltaVista Business
Solutions, explains that his company is involved in selling the
AltaVista search engine for corporate use. This gives people the
ability to extract information from Internet and intranet sites,
and also includes a developer kit so that companies can create
interfaces from the search engine into their own Web
applications.
This also raises the issue of dynamic content. Much information
on Web sites is generated from databases, meaning that it isn't
stored on static HTML pages within an organisation. It can be
difficult to manipulate this content if it is only displayed in
response to a query. As many people are now accessing such data
through the use of Active Server Page and Java Server Page scripts,
manipulation commands can be written using these scripting
languages.
Heath explains that the AltaVista search engine can also access
this data. It can be set up to perform regular queries on data, say
every 15 minutes, so that when an end-user performs a content
search, up-to-date database content is also included in the
results.
Finally, the eXtensible Markup Language [XML] is a useful means
of handling content. One thing that the language allows you to do
is to define what the data in a document means, rather than simply
how it should be displayed. Using languages written in XML, then,
it is possible for developers to produce content that can be
syndicated and searched on more easily, because a search engine can
know what relevance the data has by reading the tags associated
with it. If, for example, you are searching for news articles
marked up using an XML-based language focusing on the news
industry, end-users could specify that they want to search for news
articles with the tags "politics" and "US". This will ensure that
when they search for "George Bush" they don't get back articles
about bush fires or bushels of wheat for example.
XML can also be used to ensure that your content is displayed in
different ways depending on what sort of medium it is viewed on,
such as a PC browser, PDA or Wap phone, and browsers can also be
used to interpret instructions within XML code to display content
in a particular way. An example is drawing a chart based on a set
of numbers stored in an XML-related format.
Managing content isn't simply a case of ensuring that it's spelt
correctly and then slapping it up onto a Web page, then. Rather, it
entails careful attention to localisation, accuracy, and
searchability, along with some innovation in the area of content
display, and the handling of dynamic information.
If you can invest some cash in the right tools and skills for
the job, you will help to make your Web site a frequent point of
return for visitors, hopefully increasing your sales into the
bargain.
Content Management Players
- Allaire produces Spectra, a content management
system with workflow and process automation capabilities.
- Mediasurface announced version 3.5 of its
content management product, which includes a Java API so that it
can be hooked into other applications.
- Chrystal Software sells Eclipse, a content
management system that provides layout and design capabilities,
along with personalisation.
- Arbortext sells Epic, a content management
system that lets companies produce and manage personalised
content, translating files from XML into formats including PDF,
WML (for WAP phones) and Open Electronic Book (OEB) format for
ebooks.
Top Tips for Content Management
- Separate your content from your presentation data using style
sheets
- Use XML to describe your content in more detail
- Set up a workflow process so that key personnel can edit
content
- Capitalise on your content strategy by creating an information
portal focusing on your chosen subject, thereby building your
site's reputation
- Follow the W3C's content accessibility guidelines
- Include some form of translation in your content strategy to
extend your business to foreign language speakers
- Make sure that you include dynamic content in your strategy -
it's arguably your most valuable content asset
- Make your content searchable so that users can find things more
easily