The Mobile Web
As you might imagine given my Mozilla background, I’ve been thinking about this a lot over the past couple of years. I spend most of my time thinking about mobile & tablet, and so of course, how to build — both apps and user bases — comes up a lot.
There are a ton of aspects to the question, and some issues you could debate for hours. So I think what I’ll do here is to put down a few framing thoughts, then come back to some of the more interesting topics in more detail later.
At a high level, here’s how I see the situation: for maybe the decade between 2000 & 2010 (give or take), most of technology & innovation was happening on the Web. Anyone with a little bit of knowledge of how to get Apache running and an ability to cut & paste code from viewing the source of another web site could get started, start messing around with things, put up something useful. Not anyone to ask permission from. Nobody to check with to see if they objected to the content, or the competition. This was a pretty unusual state of affairs — and maybe unprecedented in the wave of creativity it unleashed — generativity, to use Jonathan Zittrain’s term.
It was a robust & interesting technological development because it wasn’t centered on any particular company or person; rather it came from everywhere around the world, and got stronger naturally as more and more people came online and were connected. In very many ways, the applications that we created on the web weren’t as good as the applications that had come before — the fit & finish wasn’t there. You couldn’t do a lot of the things with Google Docs, for example, that you could do with MS Office. You couldn’t do as much with Oddpost as you could with Outlook or Eudora. And the web was pretty useless on an airplane.
On the other hand: the utility of being able to do things in a networked way, with other people, connected all the time, and being able to mash up stuff from one place into another — well, that was transformative. And it turned out that it was so useful that these toy apps that didn’t work as well, didn’t have the same fit & finish, didn’t have nearly the functionality — well, they just matched the fundamentally networked way that we all want to work a lot better than native, non-networked, non-web apps did.
Then came the iPhone. Or, more accurately, a year after the iPhone came the iPhone SDK and the ability to write native apps. And Android’s, too. Neither is very webby in nature. You can’t really mash them up or search across apps. You have to get developer credentials, you can only get them to users through an app store controlled by Apple & Google (& increasingly Amazon). But you can make wonderful, amazing, immersive, intimiate experiences with them. You get apps that work all the time, online & offline, that know a ton about where you are and what you’re doing and who you’re with.
There’s been a big debate about whether you should build those apps in HTML or native code, but in my view for most companies that’s mostly an implementation decision. It’s very clear to me at this point that most apps right now should be built with native code — Objective C & Java — but have shareable artifacts that are web-based. In other words, native for engagement, web for distribution.
What’s less clear to me is how long this state of affairs will persist. Philosophically, I struggle with whether the open, no-permission-needed web is inevitable. Or rather, was simply unique. Many of my close friends assume that the openness of the web will obviously & inevitably win over the long haul.
And, for whatever it’s worth, that’s what I said, many times, as the CEO of Mozilla. The narrative that I used was this: technology always starts closed — in the garages of inventors. But eventually it starts to open up as the layers become clearer, as more people and organizations learn how to build interesting parts. And the open phase is significantly bigger and more creative and more interesting than the proprietary phase that came before.
But then I read Tim Wu’s excellent book, The Master Switch. What Wu argued is that technology tends to start closed, then open up, then close down again around proprietary networks of distribution. And he gave example after example of that — movies, telephone, radio, television. Importantly, not the web. But I think his overall point is right on the money.
What’s been confusing (to me at least!) is that with mobile, we seem to be in this third, closed phase around networks of distribution already. Things are pretty locked in around Apple, Google and Amazon as distribution brokers. But it doesn’t feel to me like we had a really robust open period — it feels to me like we’re in the mature distribution-oriented phase of the technology wave way before we’ve really figured out what it means, at a human level, to have tiny communicating computers with us all the time.
At a more prosaic, more practical level, I think the specific way that this is manifesting is that Apple and Google are doubling down, and doubling down again, on native apps. They’ve both built browsers for exposing the web to users, but they’re somewhat half-hearted reflections of browsers on the desktop, but smaller and wonkier. The structural issue is that neither Apple or Google is really incented to make a great, new, completely innovative browser experience that can deal with identity, that can be a great offline experience, that can use sensors — that can reinvent. They’re not incented because of this arms (and dollars) race around the native app ecosystem they’ve each built.
So there’s real tension.
In my view, with mobile browsers the way they are, there’s not yet an opportunity to build really big, robust mobile web properties that are amazing. You’ll tend to get the reflection of the desktop web — most in the in-app browsers that you find in apps like Twitter, etc.
Anyway, that’s my very high level reaction to Fred’s initial post. That I think most of the smartest people I know want the web to be inevitable. We want the creativity and generativity we had with the open web, but also want the intimacy and user immediacy of native apps.
We want it to be ineveitable, but right now, it’s not obvious to me who’s working on making it inevitable.
There’s opportunity for sure, but we have to change things to get there.