hyperlinking is an art. it is more than adding some html to a page. it is more than connecting two documents. it is more than saying "click here". because a hyperlink directs the reader to something other than what they intended to read, it is an opportunity to gain a reader's trust by enhancing their enjoyment of your content. but with this opportunity comes risk. adding a link to a web document creates a contract. break the contract and the reader trusts you less. the art, then, is two-fold: 1) attracting your reader, and 2) pleasing a reader with serendipitous content. a recent new york times article broke this contract with me, and i want to use it as an example to illustrate my point.
the article of my ire describes three new robert moses exhibitions. in the third paragraph, links for the institutions hosting the exhibitions are listed. the institutions (Columbia University, the Museum of the City of New York, and the Queens Museum of Art) are not likely to be known to all their readers, so this is a natural place to create hyperlinks. of particular benefit to the user would be links to the institutions' homepages, where one might read further about the show, and more importantly get the physical addresses of the institutions. unfortunately, this is not what the times does. instead the times links to their research page (see below, view actual page) for the institutions1. 
even worse, the vast majority of the links at this address are archived and require a subscription to "times select" to view them. the hyperlink contract is broken. the links are revealed for what they are: not serendipitous information for the reader, but an advertisement for the times select product. the times has missed an opportunity and now alienated a certain portion of their readers, making them less likely to trust them in the future. for the times, this can mean, fewer links followed, fewer readers in general, and worst of all, fewer "select" subscribers, which is presumably their goal.
by placing ads for their product in their content, the times is following an ever increasing trend. product placement can work, but to do so, it needs to be done well. for instance, last night i watched bladerunner. ads for coca-cola, atari, pan am, tdk and others are prominently displayed throughout the movie. yet the ads are weaved seamlessly into the futuristic urban landscape. they fit. moreover, i remember them this morning. compare that to what the times did or to movies like the departed, where the action practically stops so the viewer can note the dell computers and motorola phones. instead of thinking how cool these products were, i remember thinking it seemed highly unlikely that texts could be successfully sent from the train. the ads failed in their overtness. i knew a text message was being sent, the brand of phone was of little importance. moreover, was i really supposed to believe that all these people have the same phone?
i empathize with advertisers. getting someone to pay attention to you is increasingly difficult. yet, while fooling someone might work once, it won't work over the long term.
1. one might argue that the times does use the anchor's title attribute to explain what they link to, yet this is implemented by modern browsers as a "tool tip" which takes a couple seconds to display after the user hovers over the link. in all likelihood, the reader will have clicked the link before they see the tip.