Half the columns makes it twice as nice.
I expect Facebook to implement the experimental layout, or a variation on it, systemwide.
The reason is that Facebook's current, multi-column stream is amateur-hour design that subtly irritates and frustrates mainstream users.
In fact, serving up a design that features undifferentiated multi-column stream layouts is the easiest way for social content stream companies to limit the broad appeal of their product.
Facebook's one-column experiment
When you click your name in the upper left corner of the current Facebook interface, you're taken to your Facebook Timeline.
The Timeline has two columns of equal width and "weight," or visual emphasis, separated by a blue line.
A mostly algorithm-selected list of reverse-chronological user posts begins top left. On the right are "modules," such as "Activity," "Friends," "Photos," "Likes," "Places" and apps.
When the modules run out, the Timeline fills in the space with more Timeline posts.
As you continue to scroll down the page, you see two equal columns side-by-side all the way back in time to your birth, theoretically.
From a design perspective, the current Facebook Timeline is a train wreck.
The new experimental design makes a lot more sense. It gives more emphasis to the left column by making it wider -- it spans about 60% the width of the page, leaving about 40% to the "modules." And when the "modules" end, the space is left blank.
This subtle change doesn't entirely "fix" the Timeline's busy structure, but it improves it a lot.
If I was grading the design as a teacher, I'd give the experimental layout a solid B, and the current design a D-. Users will enjoy Facebook more once the company rolls out the redesign.
Where non-linear streams come from
The decision by Facebook and other companies (below) to both ignore centuries of content design best practices and the entire history of online content streams exposes the hubris of Silicon Valley, which favors change -- any change -- over the status quo.
It also reveals a decision-making process by a certain type of Silicon Valley company. I've worked with many valley startups and I've occasionally encountered software developers who reason that because coding is "harder" than anything else in the company, programmers are "smarter" than their co-workers. Because they believe they're smarter, they reason that they are better able to solve design, marketing and other problems better than specialists in those areas.
In some engineering-driven Silicon Valley companies, you end up with an internal culture where software engineers overrule designers on design, marketers on marketing and so on.
And that's where multi-column social content streams come from. They come from a company political culture where engineers do the design, or overrule the designers, especially engineers who believe every bit of empty space should be filled with content, and that using design to emphasize specific visual points of entry and provide hierarchies of content is arbitrary and unnecessary.
What's so great about linear social content streams?
Social and content apps and sites live or die by the unmeasurable quality of their user interfaces. Most users are unable to tell you why they like, or don't like, use or don't use a specific user interface. You can ask them, but they can't tell you.
All social applications that have been successful over the long term have presented incoming messages in a linear format, with a single stream ranked by reverse-chronological order or "best" first or some other method: Email, instant messaging, texting, and others always provide perfect linearity. For social content, this is the only system that provides broad-based, international mainstream users with peace of mind and a sense of contextual awareness.
You don't want to fill every square inch of a content interface of any kind with content for the same reason you don't want to fill every square inch of your home with "stuff." The most beautiful and pleasant homes tend to have lots of unused open space.
And you don't want multiple streams of incoming content for the same reason you don't want five mailboxes and six telephones -- it's annoying and distracting.
Who's doing it wrong
In the past few years, a plethora of aggressive content startups have shipped otherwise compelling apps, except for their sloppy disregard for a human-centered interface. These apps present users with two, three or more parallel and visually equal streams of content.
One example of horrible, multi-stream design is a news aggregator app called Pulse. The app works somewhat like Flipboard and has been widely praised by pundits. However, its three- or four-stream interface (depending on whether the tablet is in landscape or portrait orientation) will prove the kiss of death in the mainstream market as the company tries to move beyond the early adopter bubble and into the larger world.
Pinterest gets an F. By default, Pinterest presents the users with a minimum of four columns of undifferentiated emphasis. As you expand your browser or shrink the size of text on the screen, Pinterest scales up the number of columns to a huge number. Pinterest has been successful among early users. But as their user-base grows, they'll need to fix their interface to embrace linearity, or they'll wither.
Instapaper gets an A on their browser and phone apps, which are perfectly linear in reverse-chronological order. But their tablet app presents users with two equal columns of content. I give their tablet app a D. Why they chose to go two columns on the tablet is a mystery.
You'll note that these examples of bad, multi-stream design are all sites or apps that you could argue are very successful, thus invalidating my premise. But the same could be said for Facebook, and that company is apparently fixing its multi-stream problem in order to keep expanding into the larger mainstream user base.
Multi-stream social content feeds tend to be appealing to technical people, early adopters and tech pundits, who drive the early success of some of these apps and create a false sense of confidence for the companies who make them.
But the larger market -- stressed out, type-A personality "skimming" executives, older people less comfortable with screen clutter, non-technical users, teenagers, people reading content in a second language -- who collectively make up the "mainstream" user base will choose a linear interface over a non-linear one.
When the dust settles on the ongoing war over eyeballs in the social content space, the winners will all have linear, single-stream content interfaces.
Who's doing it right
Google+ in a browser also gets an A. The company offers a Twitter-like stream of perfect linearity. In fact, they even got a lot of flak from the nerdy techie segment of their audience for not filling up all the white space on the interface. It's interesting to note that a failed precursor to Google+, called Google Wave, suffered from a two-stream layout, which was awful and contributed to its failure. The phone app for Google+ is also very linear. (The tablet app, not so much.)
The Fancy gets a B+. They have a single, linear stream, for the most part, but every few items they'll slot two items in side-by-side. It's not a visual disaster, but it does represent the primacy of being "different" over user appeal.
One stream to rule them all
We live in the age of the social content stream. Users are gravitating to social content streams because they bring order to chaos, and give people a sense of context -- usually time context.
The thing is, an interface with more than one visually equal stream of content isn't a stream at all. It's just more visual chaos.
Users will ultimately gravitate to the sites, services and apps that can give them a single stream that presents social content in a single, orderly stream of updates. Twitter is already there. Google+ is there. And Facebook is getting there.
All those other sites will either figure out that linear rules, or they'll fade away into oblivion.