Jeff Bezos has been a trendsetter on a lot of subjects. One of the most recent? The idea that mobile users want their apps and websites to be fast.

Facing a complaint from a reader about the sluggishness of the Washington Post‘s mobile app, he told developers to get on top of it.

“We looked at the problem and I told Jeff I thought we could improve the load time to maybe two seconds. He wrote back and said, ‘It needs to be milliseconds,'” Post Chief Information Officer Shailesh Prakash told the Wall Street Journal last year. “He has become our ultimate beta tester.”

The Post has become famous in media circles for taking the speed issue seriously, developing its website, apps, and other platforms so that they load almost instantaneously. Take a random Post article and load it on your desktop, and you might be stunned how quickly the text and images show up on your page. 

But that’s nothing compared to its progressive web app, a mobile offering it launched as a test in May. It’s an impressive piece of work that loads pages so quickly that you might blink more slowly than an average piece of content shows up. (Poynter says the load time for a page in the app is 80 milliseconds, an absurd number.)

The Post‘s focus on speed has set a precedent for everyone else, one that most assuredly will eventually spill down to the smallest of sites.

And the hammer is already starting to be put down: In recent months, Google has put additional focus on its Accelerated Mobile Pages platform (AMP), and Facebook (which offers its own Instant Articles solution) last week ruffled the feathers of its advertisers by revealing they would limit the reach of ads that link to sluggish sites.

“Our goal is to give people the best ad experience possible on mobile,” a Facebook spokeswoman told the Wall Street Journal. “By considering website performance and a person’s network connection, we can improve that experience and help drive the outcomes advertisers are looking for.”

This increased need for speed comes at a time when advertising is increasingly getting left off the menu by savvy users who are frustrated by slow load times. But these solutions for speeding up sites come with challenges of their own.

hands on with Google’s AMP

Over Labor Day weekend, because my idea of fun during a three-day weekend is hiding in a window full of code, I spent a bunch of time playing with the new AMP templates offered by my out-of-work content management system of choice, Ghost. It took a little while for Ghost to support AMP, and having played with their implementation, I understand why.

The average Ghost site on its own is pretty fast, but the AMP pages, even on the desktop, loaded zippily, with little lag. Relatively large images like GIFs loaded instantaneously, and I was able to get advanced features like custom fonts to load without too much trouble.

(I didn’t get to WaPo‘s 80-millisecond ideal, but I’m pretty sure I got it to load a pretty rich page in less than a second.)

But I had to give up some things to build that AMP site. I couldn’t rely on my desired framework, Zurb Foundation, to build pages. I fortunately had a relatively recent project from which I could pull look-and-feel code, but it was most definitely a shift. (Lead developers of the popular competing framework Bootstrap have said they won’t support AMP, but welcome others to try.)

Developers at associations who already have limited resources are really going to have to step up their game if they want their sites to load faster.

As far as the actual code goes, AMP supports most CSS styles, but with enough exceptions that you might find yourself checking your notes frequently. A much larger shift for many web developers is that you literally cannot use any JavaScript on the site, barring a handful of specific use cases (Twitter embeds, analytics support, video embed support) that would be so egregious to leave out as to discourage uptake.

And the validation process—I frequently found myself on this page—highlights just how fragile building a fast-loading site can be.

It was an interesting experience, but let’s just say I’m in no rush to implement Instant Article support.

A Progressive Strategy

All in all, this kind of work strikes me as something that will benefit users, but developers at associations who already have limited resources are really going to have to step up their game if they want their sites to load faster.

In the case of AMP, online publishers won’t be able to ignore the technology in the long run, but web-app developers will probably pass on it. They have a whole ‘nother ball of wax to worry about: progressive web apps, like the one the Post is playing around with.

These apps focus on speed and offer up a native feel so slick that you won’t miss the confines of the app.

Google has its fingers in that phenomenon as well. The company’s Alex Russell wrote on his blog that this approach, akin to “websites that took all the right vitamins,” can offer a superior user experience.

“Users don’t have to make a heavyweight choice upfront and don’t implicitly sign up for something dangerous just by clicking on a link,” Russell wrote last year. “Sites that want to send you notifications or be on your home screen have to earn that right over time as you use them more and more. They progressively become ‘apps’.”

A good example of the potential of this approach in action is the Flipboard mobile site. Clearly, the company has made its bread and butter on apps, launching in 2010 as an iPad-only platform. But the mobile site is so good, so clean, and so slick that you don’t even miss the app experience—because there’s no delay when you’re going from page to page.

(I don’t know if you’re thinking what I’m thinking, but here’s what I’m thinking: This is almost good enough to replace an app, and just in time, too, as associations were caught off guard by Apple’s announcement last week that it would drop apps that aren’t regularly updated from the App Store.)

The Web: Not Dead After All

Back in 2010, Wired published a blunt cover story declaring “The Web Is Dead.” Six years later, the magazine published a partial mea culpa, citing Google’s recent work on progressive web apps as one reason that reports of the web’s demise were greatly exaggerated.

The difference between then and now is the emergence of some promising building blocks. Google’s AMP won’t save the mobile web on its own, but if we can get more companies like Flipboard and the Washington Post (and maybe a few associations) on board, it may not be long until we put mobile web apps back into serious discussion as a way forward online.

Because all those milliseconds add up.

(iStock/Thinkstock)