Apple long ago abandoned its original “Magical and Revolutionary” tagline for the iPad, probably out of some embarrassment at how the word ‘magical’ made so many of us groan. But the more I use, build and learn about touch-based software, the more I think magic is really a key component of this stuff, even if it’s not exclusive to the iPad.

I thought about this recently when a co-worker introduced me to Moves, an iPhone app that tracks the number of steps you take, with the aim of getting you to be more physically active from day to day. Once downloaded, you use Moves by doing… well, almost nothing. The app does everything for you, recording and parsing out your steps by mapping where you’ve traveled over the course of the day, how far and how fast, all with no user intervention required. All you have to do is the walking part, and the app quite literally does the rest, generating a complete, metered itinerary for all the walking and (most of) the places you visited in a given day.

Less Typing Equals More Output

This automation is the headline feature of Moves, and with good reason. It’s no small thing to completely absolve the user of all data entry obligations, to passively collect what other apps would demand users actively enter (and to do so without draining battery life at inordinate levels). This feat directly addresses one of the central pain points of most mobile software: typing is a chore.

No one wants to type more on a multi-touch phone or tablet if they don’t have to, so when they see an app demonstrate that typing can be eliminated entirely, it’s an eye-opening moment, for sure. But is it magic? Almost. To simplify is huge, but what matters just as much is the end result, what the user gets out of the simplification. If the simplified process produces satisfactory results, great. But it’s magic when the software generates a disproportionately meaningful output from that minimized input.

Moves really couldn’t ask for a lower level of effort to use it. And yet the output is several orders of magnitude greater than that input: it’s automatic; it’s real time; it’s accurate; it’s recognizable; it’s useful; it’s beautifully presented and it just works. Even better, everyone I know can’t stop talking about how cool it is.

This inverse relationship between active user input and automated output is wonderfully consistent with how real people use mobile software. Unlike desktops, mobile devices are more often than not complements to other, real world activities, where ‘computing’ is not the main activity. Phones and tablets are used in situ, and so their software cannot afford to demand high levels of input effort. The trick here — the magic — is to ask for less from the user while at the same time raising, increasing, augmenting, or biggerizing the output so that it exceeds expectations, so that it᾿s out of scale to the reduced input.

Many of my favorite mobile apps do this: Instagram requires only that you press the shutter button in order to produce a beautiful photo; Shazam requires only that you hold up your phone in order to identify any song being played near you; Dark Sky requires only that you open it up to know all that you need to know about going outside in the next hour, etc. Magic is what Apple is going after with Siri (just tell your phone what you want) and what Google is going for with Google Now (do whatever you were going to do, and we’ll help you do what’s next). It’s even the same idea behind stalled innovations like NFC (just wave your phone to transact) and QR codes (just point your phone at a blotchy thing for more information). Magic is hard.

The Value of New

In fact, magic is not even as simple as I’ve described it here so far — the output needs to be amazing but it also needs to be original. The history of software is rife with examples of magic that later faded; for those accountants to whom Dan Bricklin and Bob Frankston first demonstrated VisiCalc in 1979, spreadsheets were magic because they reduced the input effort and scaled up the output disproportionately. Enter numbers in a few cells and perform any kind of arithmetic on them, instantly. That was pretty amazing then, but over the years it has become decidedly pedestrian.

My formula for magic comes with a corollary, then: it demands a certain amount of freshness. The software needs to produce a result that is both out of scale and of a type that is new, unexpected, or unfamiliar — the designer has to be the first (or among the first) to perform a given trick for the audience. And of course, the more often you see a trick, the less impressive it gets, to the point where pulling a rabbit out of a top hat becomes mundane.

Moves could have presented all of the data it collected in a humdrum table view, leaving you to make sense of it on your own. Instead it generates a novel visualization of your day’s walking, replete with dimensional graphics, drop shadows, incidental iconography and arguably superfluous gradients. It’s more than just your data, collected; it’s your day designed. There’s nothing quite like it out there that I’ve seen, but I’m sure that will change soon.

As much as I’ve tried to express this idea as a kind of formula, it goes without saying that conjuring up magic is rarely easy, and hardly predictable. Only a handful of apps out of the hundreds of thousands in the market have managed to capture this idea, at least for my money. But I think understanding this relationship between reduced user input and out-of-scale, novel outputs is a useful guide for the apps that are still to come. Grasping the concept makes the challenge easier, though only mildly so. By its very nature this will always remain a difficult thing to pull off. Magic is not a science, after all.

Thanks to Felix Salmon, Semil Shah and David Jacobs for reading early drafts of this post via the awesome new Editorially app.