Many large websites have reached a critical juncture. Whether intranet or public, they have grown very fast with little control or planning.

Progressive organisations are saying that enough is enough, and are turning the chaos into order. They are publishing quality content and dumping the rest.

In March, I spoke to about 800 people, in 11 locations in the US and Australia, about information architecture and website management.

These were people involved in the management or running of large sites, and the same type of problems kept coming up again and again, whether in California or Canberra.

One of the most frequent issues was the question of how to deal with sites that have gone out of control. Many now have masses of content, much of it out-of-date, poorly written and badly organised. Some managers weren't even sure how many sites they had, let alone how many pages.

Another big question was how they should go about the creation of the basic structures and processes that would enable managers to professionally manage their sites.

Some were facing more fundamental issues: how to justify the existence of the website; and how to show that it was delivering at least some value to the organisation.

Other issues included quality control, motivating others to publish consistently and regularly, and classification of content.

Yet others wanted to create a consistent navigation that helped people find key content quickly. Others wanted better metadata or improved search.

While in Australia I was told the story of the IBM intranet. IBM, seemingly, had 7,000 intranets. In a survey charting where people looked for information, the intranets rated poorly.

So IBM decided to have one intranet that was well organised and had quality content. Now, it is rated as a source of valuable information.

Many organisations simply don't understand content: how much it costs to publish; or how to understand the value it can create. I came across a consultant who was dealing with a large software company.

This company had found that 99 per cent of the visitors to its public website were reading the same one per cent of its content; 25 per cent of the content had never been read.

Imagine the sheer waste of time and resources involved in this. Think of the sheer futility of creating vast quantities of content that nobody has any interest in reading.

If a car manufacturer kept making cars that nobody wanted to buy, then sooner or later they'd figure that out and stop making those cars. Yet organisations are busy, every day, creating content that will never be read.

Content should drive actions and to do this it needs to be read and understood. But many organisations simply don't get that connection.

Saying that you have a 500,000 page intranet is not something you should be proud of. You would probably have a much better intranet if you had 50,000 pages of well organised, high-quality content.

The problems that large websites face will not be solved by software. Very often, software makes the problems worse.

Everywhere I go, I come across the great publishing challenge: getting the right content to the right person at the right time and at the right cost.

When it comes to content management, the software industry is peddling giant machines, the type of equipment you'd use in a coal mine. It cost-effectively shifts vast quantities of content. The problem is that this causes, rather than solves, information overload.

Intranets and public sites have become unmanageable because content is being treated like it was coal. It's not. Content is never about volume; 500 words are invariably better than 5,000.

Slowly, we are seeing a recognition of this basic fact. I talked to one manager who was responsible for a series of sites with nearly one million pages.

He was in the process of bringing it together under one unified structure. There was a launch date for the new structure.

Those sites that had not published their content under the new structure by that date were being switched off.

During my workshops, I proposed that at least once a year all content on a website should be read to make sure it is still up-to-date. Some people questioned whether that was possible; their sites simply had too much content.

But if you don't have time to check whether your content is up-to-date, what does that say? Is your content of so little value that it doesn't deserve basic quality control measures? What does that say to readers?

The problems organisations face today with content are not particularly complex. They have little to do with software and lots to do with people and processes.

Start off by figuring out who it is that really needs your content. Figure out exactly what content they need, and forget about all that like-to-have stuff. Then write this content well and publish it in an organised manner.

Gerry McGovern is an author and consultant. He has recently published Content Critical and The Web Content Style Guide (FT Prentice Hall).