A most fundamental principle of interaction design is user control. I prefer performer control, as the emphasis is on the task or activity to be completed, rather than on the tool being used. But I digress.
Performer control means that there is an obvious and available action (something to click, swipe, pinch, or otherwise interact with) that leads directly to an intended outcome. For example, if you are not interested in reading this article, you can make it disappear by clicking a familiar red dot or x that can be found in its usual place.
The opposite of performer control is sometimes referred to as forcing. A simple example is the modal dialogue, which is a child window that insists upon your input before you can do anything else. In the very best of applications, forcing (also called a forcing function) is applied to obviate risk or harm, such as a financial investment system blocking transactions that could lead to runaway selling and a devastating stock market crash. The existence of common human errors, such as capture errors and loss-of-activation errors, are good reasons to impose forcing or, at least, redirecting action in a computer interface, to mitigate such natural error. In less compelling cases, modal dialogues are used to prevent wasteful or illogical actions. For example, MS Word’s Insert Table dialogue becomes modal if you enter 0 rows and 0 columns.
Inappropriate applications of forcing often result from poorly conceived back-end engineering (e.g., forcing to accommodate an ill-conceived data model), from a desire to simplify the programmer’s code (while inconveniencing the performer), and to unnecessarily accommodate very low probability (or impossible) edge cases.
The latest assaults on the performer through forcing today are influenced by sales and marketing activities. I first experienced this while writing for an online music publication and finding a modal 30-second video dialogue that was a gateway to my latest article. The reader could not get to my article without viewing the entire video (hence a forcing function). Many of my readers complained about this, some of whom did not read the article because of the inconvenience. When I conveyed these findings to the publisher, he said that the advertisers insisted on the modality. “The publication is free, so if they want to read it, they need to see the complete ads,” was the rationale for the strong-arm tactic. Further, the online table of contents for the issue did not include active links. Readers had to navigate through 55 pages, linearly, so as not to miss an ad, to get to my article. You can image what good became of all of that before necessary changes were made to save the publication. Human beings, after all, will ultimately vote with their feet.
Lately, we see a compromise in forcing that allows you to dismiss an ad after 3-5 seconds of play. This is perhaps a bit less annoying than complete modality, but annoying nonetheless. There are still many ads that do not allow dismissal at all, prevalent in free apps and games. The logic (?) being that if you want it for free, you have to suffer through the ads. Another is the identity collector dialogue, which appears in the following context. You encounter a link to an article that interests you. You click on the link, but the article your arrive at is greyed out in the background, while a modal dialogue box appears asking you (forcing you) to enter your email address (and perhaps more). You quickly notice that there is no obvious way to dismiss the dialogue. In some cases, clicking anywhere off the dialogue will dismiss it; in other cases, you simply cannot dismiss it until you provide your contact information. Horrible. Your only alternative is to dismiss the entire article’s tab or the entire browser, resulting in loss of the original context. Further (and most devastating), such forcing dialogues proliferate worst-practices and subvert standards.
The real tragedy of such practices is that they are too often attributed to the user experience (UX), usability, or human factors engineering activities of application development when, in fact, they stand in stark contrast to accepted principles of human factors engineering. Forcing functions, in this context, are designed to benefit the advertiser, and do so at the expense of human preference. Too often it is what a misapplied A/B test suggests as that which will yield a higher margin in revenue or mindshare that is implemented, rather than that of a legitimate application of UX engineering. Or perhaps there are no human factors principles in the mix at all, and the forcing function is no more than an ill-conceived “gotcha.” Either way, the practice is rude and offensive.
My suggestion? Vote with your feet. You will not only be voting for far less annoyance in your life, but helping to correct a very bad trend that is an affront to human dignity.