Methodology for choosing new content

What process do you use to choose which content to add to a site?

[itworked]

I have an established site, 15 years old, with several pages of long tail keywords at position 1.0 according to Webmaster tools.

We haven't added much new content in the past year but want to start publishing one well researched article per week.

At a high level does it work best to publish articles that are related to the highest performing (traffic or revenue) existing content? Or is it better to go broader and publish on topics completely new topics still within the scope of the website but quite different than the existing subjects in hope of creating new, high performing subject areas?

I anticipate part of the answer being "Well what do your visitors want to read?" The answer is, I know what they are reading, but "they" is ever changing since the majority are not repeat visitors.

If I piggy back off of high performing existing content then I can link from and to those articles which seems like a healthy strategy. Introducing new topics would make for fewer inner linking opportunities initially.

[mack]

I like to use the analogy of a fishing net. If you remain extremely targeted to your current top performing pages you are aiming at a very small spot on the ocean.

As you broaden your scope your net widens and you are able to catch more fish (visitors).

It is however, importaint to retain a topic. Don't drift to far away from your key subjects or you run the risk of diluting your content.

Mack.

[engine]

The process to choose might also include topical items, such as breaking news, or market trends. Both those two tend to allow scope for the creation of new material.

But, for sure, as mack says, keep on topic.