Grids are the best thing to ever happen to graphic design. They form a rational basis for organizing information. They support the harmonious distribution of elements and visual weight. We don’t design things with grids because it’s easy, or because everyone else is doing it. It’s certainly not a trend. We design things with grids because they enable the human brain to efficiently and enjoyably process information.
The text of this article is left-aligned, and there’s a reason for that. Left-aligned text gives the eye a consistent line to return to. Instead of searching for the beginning of each new line, a kind of muscle memory helps us avoid wasted cognitive effort. This is a perfect way to understand how great the grid is.
Grids are like left-aligned text for your entire layout. They establish a rational and consistent method of scanning the various bits of information that make up a whole design. They help establish hierarchy and rhythm. They support good proportion. Grids are everything good and right with design.
There could, of course, be a good alternative to the grid. Grids work because they support the human brain, and intense study and effort could reveal other systems that support it just as well—or even better. “Perhaps I’ll run out and invent something better than the grid for my next client project, to help spice things up,” however, is not a good plan. It’s a bad plan in the same way it’s a bad plan to design your own cryptography, or design your own aircraft engine: The chances your contribution won’t fail are are slim, and the consequences are often catastrophic.
If you try to think up something novel for the sake of novelty, without considering why the standard exists in the first place, you’re likely to come up with something terrifically bad. Take this example, in which the hover state for a link is strikethrough text:
It’s different, right? You usually don’t see that. Well, there’s a reason. Strikethrough isn’t just a cool look—it has meaning. It means that something is no longer valid. Strikethrough would be appropriate in the event of an error, if the link no longer went anywhere, or there were some other problem. In this case, our novel solution has implied exactly the opposite of the truth. It’s worse than nothing at all. This uncommon solution is uncommon precisely because it’s comprehensively wrong.
If you do something novel because it’s better for an articulable reason, you are designing. If you reject standards out of hand because they’re not interesting enough, you’re simply bowling for bad solutions.
Use a grid.