Given the audience of this magazine, it hardly needs to be argued how much better, more beautiful, easier, and more useful desktop Linux has become in recent years. It's not like there's nothing left on the "to-do" list for a better and friendlier Linux, but there's also no denying how many items have been ticked off the list – and how many areas where Linux is already well ahead of the competition.

Even a few years ago, a Linux user needed to have enough technical skill to occasionally convert a file, download a codec, or update a driver. All these issues have almost totally gone away. The major Linux desktops are every bit as easy to use as macOS, and they are way more intuitive than Windows 10.

So has the Linux market share improved along with these significant improvements to the Linux user experience? It should vex us all that it hasn't. No matter how much we want Linux to break through to the mainstream market, it seems eternally stuck in the geek space.

Consider the case of the Valve Steam gaming platform. The Proton compatibility layer has brought thousands of new games to Linux [1], and the graphics drivers keep getting better, to a point where some native Windows titles now run faster on Linux. This night-and-day difference in Linux game support has lifted Linux's share of Steam users from under one percent to … under one percent.

So What's Missing?

Nonusers, by definition, are disconnected from the software. As I've become more engaged with this topic, I've become ever more fascinated to probe how far this disconnection goes. A big corporation that has a question about how their product is perceived by potential users typically commissions a marketing study, which could easily cost tens of thousands, or even hundreds of thousands, of dollars. But Linux isn't one big corporation.

I decided to do my own fact-finding mission. First I read through all the published data on what Mac and Windows users believe and know about desktop Linux. This was a very quick reading assignment; there is no published data.

So I spoke to as many nonusers as I could about Linux. I maintained this hobby as I traveled for work and pleasure, opening conversations with Irish backpackers, German school teachers, Albanian guesthouse owners, Israeli airport security, Thai scientists, and beyond. The most common take on Linux was "that's something to do with computers, right?"

Even many IT professionals do not think of Linux as a desktop OS: "Of course it's fast; the only thing on it that might slow it down is a text editor." Tragically, these types of outdated and simplistic responses even came from some admins who deal with Linux servers every work day.

What nonusers see of desktop Linux is contextual: social cues, media visibility, market position, perceptions, reputation, and so on. So no wonder the immense improvements in the software haven't driven adoption; to nonusers, these changes are invisible. The context they can see still says that Linux is a system for geeks.

Engaging Intuition

The Linux ecosystem has become quite good at speaking to the conscious, rational mind. If you want half-hour videos delivering depth and detail, they exist and are often excellent. But slow, careful thinking is cognitively expensive, so humans use it sparingly. We mostly default to much faster intuitions [2] that feel effortless and automatic.

Intuitive thinking makes rapid judgments from visible information and then immediately moves on. It's not tuned for objective truth and perfect choices but to navigate a busy and uncertain world with quick decisions that reliably avoid disaster.

When speaking to this fast style of thinking, the context is at least as much a part of the communication as what's said. So when the audience uses this surface-level stuff to decide that desktop Linux isn't for them, the code has no chance to change their mind.

Social Proof

By now, you can probably make a lot more sense of why Linux has done so much better across servers and embedded systems. It's not just that it meets these needs very well, though clearly that's a big part of it. It also engages quite excellently with the slow, careful thinking of the professionals paid to agonize over these decisions. To win with intuitive reasoning, we don't just need more marketing for Linux – we need a different kind of marketing. It needs to be alert to the biases and heuristics that characterize busy, distracted minds running on instinct.

The most important of these might be "social proof". This is a term coined by academic psychologist Robert Cialdini [3] to describe how people feel influenced to copy what they see others doing. We're more influenced when we feel uncertain, and we're more likely to be influenced by people we perceive as similar to ourselves [4].

It seems rational to suppose that, barring any technical impediments, the fact that professionals and power users choose Linux would drive its adoption with less technical users. Clearly, the pros know computers best, right?

A social proof perspective, on the other hand, argues that users will choose an operating system that they see others at a similar technical level using. If a consumer looks around and sees that users at similar technical level are only choosing Windows and macOS, it doesn't matter how many times you tell them about the stability and security benefits of Linux. As long as Linux is only visible as an OS used by the technically brilliant, the intuitive mind screams that it's too difficult for anyone else. If you understand this as a social proof problem, desktop Linux's adoption challenges look simpler yet more diabolical – it's unpopular for being unpopular.