Surveying the Landscape

I made my first website around 20 years ago; it was perhaps 1997, I forget exactly. It was a Geocities site, which meant that all I needed to get an acceptable result was to combine animated GIFs with Comic Sans. After Geocities I moved to a self-hosted site, built using Microsoft Frontpage, and began to learn how to code. In 2000 I earned my first payment for building a website: I became a professional. The web has provided my livelihood for the 16 years since.

Which is why I was a little dismayed when Wired stated in 2010 that The Web Is Dead. Since then I’ve seen similar statements—either that the web is dying, or at the very least is no longer important—appear with increasing frequency from many in the Silicon Valley scene.

I don’t agree that the web is dead, or even that it’s dying. But now that I have shifted my career from web developer to technologist, I have to take a more broad view, beyond the web alone, to look for trends in technology and the way people use it. And what I see makes me worry, about what Chris Dixon identifies in his post, The Internet Economy:

The realistic danger isn’t that the web disappears, but that it gets marginalized.

The technology landscape today contains many threats to the web: not to its existence, but to its long-term health and prosperity as a first-choice platform for products and services.

These threats come in many forms. There’s the increasing commoditisation of tools, as companies like Squarespace make it easier than ever to get a good-looking site online quickly—and even easier in future as smarter tools like The Grid and Wix’ Artificial Design Intelligence minimise the effort required in launching on the web by using content-based automatic generation.

In the slightly longer term the commoditisation of AI could have an effect, automating common coding, design and content production roles and diminishing further the available work for professionals. Wired, ever the harbinger of doom, have already predicted ‘The End of Code’.

But the larger threats come in broader changes to the landscape: in user behaviour, as mobile eats the world, and publishing, as there’s a general consolidation around distribution platforms. Together, these trends indicate the possibility that the web—the open web, as I mean it—will be left behind.

From Apps to Platforms

In recent years, apps have provide the biggest competition to the web, with both Android and Apple continuing to prioritise on performance, creation and distribution tools, and publishers providing optimised experiences for app users over those on the web—if they provide any web interface at all. John Herrman’s article, Upload Complete (which I’ll be referring to extensively in this post) puts it like this:

In 2016, the still-vast and still-growing web is comparatively diminished. Slightly newer platforms, such as Instagram and Vine, hardly touch it; Snapchat doesn’t even acknowledge it.

Many products launch app-first, some launch app-only, and some formerly web-based products are being dropped in favour of apps—last year, two major Indian retailers, Flipkart and Myntra, chose an app-only mobile strategy, closing down their mobile web retail interfaces and leaving in their place a prompt to download their respective native apps.

Apple made an estimated $6bn from the App Store last year and seemingly de-prioritised their Safari browser, as it’s seen relatively few changes over the past few years. And in iOS9 they introduced universal links, a way to bypass the web by intercepting URLs and directing the user to an installed app rather than the browser.

Google have gone one further with the introduction of instant apps on Android; these also intercept URLs, but open them in app fragments streamed from a server, bypassing both the web and the requirement of having an app installed at all. They also recently announced that the previously web-only Chrome OS will run native Android apps.

Google’s business is advertising, and advertising is also a threat to the web. Most of the time advertising on the web is merely annoying; often, as with multiple app-download doorslams, it’s obstructive. But sometimes advertising—or, perhaps to be more precise, the tracking script that accompanies advertising—is onerous, or even downright dangerous.

A recent study showed that while ads take up only 9% of screen real estate (a generous average, IMO) they account for some 54% of load time. Another study (admittedly small-scale) suggested that between 18% and 79% of data downloaded on some sites was advertising/tracking. This is contributing to a chilling effect online:

Nearly one in two internet users say privacy and security concerns have stopped them from doing basic things online, such as posting to social networks, expressing opinions in forums or even buying things from websites.

Of course, ads can be blocked, and a recent report indicates that people are doing just that: of smartphone users, some 22% (that’s 419 million people) use ad-blocking software of some kind, whether browsers with built-in blocking (such as UC, or Opera) or browser extensions. And this may be generational: in the UK some 22% of adults aged 25-44 use ad-blocking software, but that number rises to almost half (47%) in those aged 18-24

(Some of these numbers are disputed, although the general trend is not).

The problem is, many sites rely on advertising for their income, and ad blockers reduced that income by $22 billion in 2014 (almost double the figure from the year before), and in an absence of alternative forms of monetisation, we’re likely to see a collapse coming soon. In the UK we’ve seen the internet hollow out the newspaper industry, and the advertising collapse is likely to see a similar story play out online: already this year publishers such as Buzzfeed, the Guardian, the Telegraph and Vice are reporting lower than expected revenue.

Publishers keen to turn a profit will have to partner with the platforms that make money: that is, Facebook and Google. In the first quarter of 2016 it was estimated that between them, these two took 85 cents of every new advertising dollar. Emily Bell, formerly of The Guardian, warned media publishers:

Having a legacy business configured around a website is now almost as much of a headache as the rumbling printing press, fuelled by paper and money.

Bell’s prediction that publishers might close their own websites and rely instead on platforms for distribution was borne out this year through a new UK newspaper called The New Day, which launched with no dedicated website, opting to distribute online entirely through social channels (it folded later in the same year, but that’s more related to physical sales).

In this climate, Evan Williams (formerly of Twitter and now of Medium) believes it’s inevitable that publishing—professional and personal—will move to centralised platforms:

The idea won’t be to start a website. That will be dead. 
The individual website won’t matter. The internet is not going to be about billions of people going to millions of websites. It will be about getting it from centralised websites.

Jon Lax, director of product design at Facebook, agrees:

A lot of designers are still focused on building websites and I’m not sure that’s a growth business any more. The trend of mobile has fundamentally changed the dynamics of how information is distributed. It turns out that people like feeds.

You may think, of course they would say that, they have vested interests. But publishers are moving their operations to Medium; and Lax closed his successful web design agency to work for Facebook.

In the earlier days of the web, from its inception to around 2000, content was generally discovered through portals such as Yahoo (I refuse to use the exclamation). From 2000 to a few years ago, search—or rather, Google—was the prominent paradigm. But today content is largely shared and discovered through feeds, on platforms including Medium, Twitter, Apple News, and of course Facebook, with its 1.6bn active users every month.

Earlier in this article I mentioned John Herrman’s piece, Upload Complete. It’s the final part of a series of articles called The Content Wars, published by the independent publisher The Awl, and explains very clearly what’s happening to the production and distribution of content online:

Facebook… has gradually replaced, within its platform, parts of the web its users shared most: outside videos; outside articles; outside personal sites; outside listings; outside streams; soon, maybe, outside audio streams. Twitter, slowly and with less leverage, has been doing the same.

The platforms that came of age on the web are leaving it behind.

Understood from the perspective of the web, the last five years have represented a sort of tragedy of the commons.

(The tragedy of the commons is the process wherein individual users acting independently and rationally according to their own self-interest behave contrary to the common good of all users by depleting that resource.)

The platforms grew big and strong. Websites and publishers catered to the needs of those platforms, vague worries about control and identity set aside for the necessary pursuit of audience in an unpredictable environment.

Shortly after the piece was published, The Awl closed their website and moved all their content to Medium.

The Next Billion

I’m going to digress briefly at this point to talk about “the next billion”. This is a phrase we hear a lot, and what it means is that in 2014 there were approximately three billion people online globally, and in 2018 the prediction is that four billion people will be online. So those are the next billion, and every major company wants to reach them as clients/customers/consumers.

In India alone, 105m people came online last year (854m are yet to come online). And the next billion are coming online in a very different environment than most of us: many—perhaps almost all—of them will only ever experience the online world through their mobiles; and most of them live in rural areas where networks are either poor quality or very expensive, or both.

A quick back-of-the-envelope calculation says that for the salary earned from one hour of minimum wage work in India, you can visit 15 websites. In Myanmar, where some 60% of the country have come online in the past two years alone, 1GB of data costs approx. 15% of the national average monthly income—and of course that percentage is likely to be much higher in the rural areas where income is lower.

In a great article, The Facebook-Loving Farmers of Myanmar, which reports on an ethnographic field study of the country opening up in the transition to democracy, Craig Mod explains:

The telcos operate on pre-paid systems. Nobody has a credit card. Everyone buys top-up from top-up shops… They feel each megabyte.

Mod observes that the result of this cost is that Facebook is incredibly popular, because:

Facebook has a compelling advantage over other news apps or even Twitter: The content of many posts and news items live inside Facebook itself. There are external links, but most of the article summaries and photos are self contained.

Facebook’s continued aggregation of content from outside of its platform, into its platform, offers a clear advantage to people for whom data cost and speed is a real issue. Indeed, so strong is Facebook’s presence across the emerging online world that for many people Facebook is the internet. Mod continues:

As Facebook continues to ramp up their Instant Articles the amount of content that lives in Facebook will only increase. For those who are data sensitive, this is a clear virtue.

For the unaware, Instant Articles are Facebook’s method of displaying content from publishers natively on their own platform rather than the publishers’ websites. Articles are stripped of navigation, tracking and advertising, to ensure faster loading. In the launch announcement of instant articles, Facebook said:

Stories [shared on Facebook] take an average of eight seconds to load, by far the slowest single content type. Instant Articles makes the reading experience as much as ten times faster than standard mobile web articles.

Facebook are not the only platform that consolidate stories in this way; Flipboard and Apple News, among others, also do the same. The AMP project, spearheaded by Google, offers a ‘web-friendly’ alternative that works best when cached on their own servers.

In addition to Instant Articles, Facebook have their own new advertising service, Canvas, to make distribution of content on their platform a more attractive proposition. The Canvas website also makes a point of playing up its superior performance compared to the web:

A fast-loading and seamless experience on Android and iOS. Canvas loads quickly, as much as ten times faster than the standard mobile web.

Messaging and Meta-Platforms

Other than Facebook, the other type of app that’s popular across the developing world—and, indeed, the rest of the world—is messaging. From WeChat in China to Line in Japan, from Kakao in Korea to Telegram in the Middle East to Kik in the USA, there are many different messaging apps across the different markets. Spanning every market is the biggest of them all, WhatsApp, with 1 billion monthly active users, although Facebook Messenger follows closely with 900 million.

Rather than being simply messaging apps, used for inter-personal text communication, most are actively building up the capabilities of their offerings to make them into fully-featured platforms. Anthony Green of Kik states:

Messaging apps are the new browsers.

The model they all want to follow is Tencent’s enormously popular and influential WeChat (Wēixìn in its native China). WeChat has around 700 million monthly active users—only slightly less than the estimated 720 million internet users in the country. The platform offers users services such as payments, banking, transport, entertainment and gaming (and client-friendly tools like tracking and analytics). Stephen Wang, a product manager at WeChat, put it like this:

At every time throughout the day, there is a touchpoint between WeChat and your normal life.

That’s not just wishful thinking; recent data shows that, in April 2016, some 35% of time online in China was spent in WeChat.

In his insightful article from 2015, Rise of the Meta-platforms and the New ‘Web Browsers’, Paul Kinlan of the Google Chrome team looked at messaging apps as meta-platforms; that is, platforms consumed on platforms where they are likely not a first class citizen. That definition also applies to the web browser, and Facebook. One of the things that was most clear to him was:

Many of the platforms on mobile are choosing to host content within their own app experience and this is in my opinion a direct and compelling threat to the web.

Facebook have Instant Articles; Apple, their News platform; WeChat, Official Accounts (brand pages within the platform). All are aimed at consolidating the user experience on their own platform, largely at the expense of other apps—and the web.

As people spend more time in messaging apps, it becomes inevitable that brands and services want to engage with them there. Ben Eidelson, formerly of Google Hangouts, attributes this to a simple insight:

If messaging is so useful and so popular…why do you only message with your friends and family?

Many of the messaging platforms have introduced agents of communication between users and services, through the messaging interface. This communication is generally referred to as Conversational UI, and the agents of communication are bots. The role of bots is suggested in the full quote from Anthony Green that I mentioned earlier:

Messaging apps are the new browsers, bots are the new web pages.

The messaging / communications platforms Facebook Messenger, Telegram, Skype, Kik, and Slack each have APIs that facilitate communication through bots. Some of Messenger’s earliest efforts have been somewhat overhyped and underwhelming, but this is a new area and likely to improve as people learn how to best take advantage of the opportunities of bots. Chris Messina offered this definition:

The way I see it, bots are just wrappers for APIs in conversational UIs. Apps are just wrappers for APIs on home screens.

Put in those terms, bots seem like the reasonable—even logical—step for messaging platforms. And as people are able to interact with bots in their platform of choice, they’re less likely to use other apps—including the web browser.

Conversational interaction also opens up the market for two other services that don’t require the web. The first is wearables, on which the most common actions are reading notifications, and interacting with them. The streamlining of services to suit bots also creates opportunity for interaction through simple, actionable alerts on a smart watch.

And a logical extension of conversational UI is into services controlled entirely by voice. Products like Amazon Echo, the recently announced Google Home, and Apple’s CarPlay featuring Siri, allow users to interact with online services in a way that requires limited visual interface—or none at all.

Finding Opportunities

A quick recap of the threats to the web as I see them: prioritisation of apps on mobile OSes, including URL interception; destructive online advertising and tracking; ad blocking contributing to a collapse of monetisation; publishers distributing through centralised platforms; emerging markets with no legacy attachment to the web and no incentive to use it; content consolidation onto platforms to suit the network demands of new markets; messaging platforms offering direct communication with services in a conversational interface; and new products with no visual UI at all.

Wow… that’s heavy.

But all is not lost. Things may look grim, but these are trends towards a possible future, not manifestations of the certain future. The web is far from done. While 1.6bn people use Facebook every month, everyone online —over 3bn people—has access to a web browser. In April, Google announced that one billion people access the web using Chrome on mobile alone.

The web still has two great advantages: distribution and reach. Jason Karaian, of Quartz, the online news publisher who have recently released a native, conversational chat client, explains why their main focus is still the web:

The best way to reach new consumers is through the open web.

We know the web model works because everyone wants to copy it. Apple’s universal links and Google’s instant apps both rely on URLs. But as Klint Finley put it in a recent Wired article:

If you want apps that work like the web, the web is still your best choice.

Wired recently backed down from their earlier assertion that the web is dead, with the article Wait! The Web Isn’t Dead After All. Google Made Sure of It. While they’re giving too much credit to Google alone, what they’re specifically referring to are Progressive Web Apps (PWA). These are sites that are built to a certain set of principles: for example that they be engaging, installable, shareable, and fast. They don’t mimic native apps, but are intended to match the experience and performance expectations set by native while keeping the web’s advantages.

I wrote earlier about Flipkart’s decision to drop their mobile site and go app-only, leaving their mobile web presence as simply a page with a push to download the app. But there was a sting in the tale: only 4% of visitors actually installed the app. Sales dropped, and Flipkart moved quickly to build a new mobile website, Flipkart Lite (mobile only), built to PWA principles. The site has seen big success in many metrics: for example, making it easier to add the site’s icon to the mobile homescreen, and therefore making it more visible, led to a 70% increase in conversions.

Much of the interest in PWA comes from the developing world, with early adopters especially common across Asia. This is not surprising: the world is changing as the next billion come online. And when they come online it’s on mobile, where they feel the cost of the data. Native apps come with a cost to download, and a cost to keep updated; offline-optimised web apps should offer the advantage of being lighter to load, with fewer and lighter updates.

As browsers become more capable and the web can offer experiences that compete with native, we may perhaps be seeing the beginning of a fightback against native apps. An early example are Patagonia, who scrapped their iOS app to go web only, with the message:

Now that our website is beautiful and easy to use on all mobile web browsers, we will no longer be supporting this app—you may delete it from your device.

Even in China, where WeChat substitutes the browser for many people, much of the interaction that happens with official accounts is done through mobile web pages viewed in an embedded browser rather than an external one. Alessio Basso, who works in Asia with platforms including WeChat, explains it thus:

WeChat is an incredible driver for adoption of mobile web development techniques. While in the West companies are still native-first (usually iOS-first), in China the most important presence one may have is a mobile-optimised website.

Asian messaging apps tend to be good at discovery, through their embedded QR code reader (QR codes are more common there, especially in China). This has created a category of web app which aren’t at all common outside of the region: the light app. These are rich, animated, one-hit web pages; a user might open one, read it, share it with a friend (hopefully), then never visit again. Kendra Schaefer describes them:

Light apps are often single-page, single-message and intended to be single-use.

This low-friction interaction is one of the strongest features the web has in its favour, and why I’m excited by the Physical Web, a method that involves a URL broadcast from a beacon and discovered by your browser (quietly, in the background), so that you can connect with web pages in your immediate vicinity. For example, in London (where I live) some buses transmit a URL so that passengers can connect to a web page with timetables, and route and travel information (and some advertising). No installation, no search, just information that’s directly appropriate to your location.

This localised discovery becomes even more powerful when combined with the Web Bluetooth API. The browser is enabled to directly connect to objects that broadcast a GATT service over Bluetooth Low Energy, allowing people to ‘walk up and use’ connected objects without any app intermediary. This puts the web at the heart of the ‘internet of things’.

And there are opportunities for the web in platforms we’d perhaps never consider. Virtual Reality, like mobile, suffers from cross-OS app incompatibility; Oculus Rift apps can’t run on Vive headsets, and vice versa. Any VR headset can run a browser, however; so the Web VR project aims to create a Javascript API that delivers content on any VR device.

All of this makes the web competitive, and offers some hope for the future. But for the web to really thrive as well as compete, perhaps what’s needed is a more fundamental rethink of the web and browser.

The Web in the Future

The role of the web has changed a lot in the 25 years or so since its inception. It was conceived as a hyperlinked document repository, but today it’s an integral part of our culture: we conduct social, financial, and governmental transactions on the web, we write our history there. So in order for the web to prosper into the future, it could be time to reconsider it. In his blog post Does the Web Matter Anymore? Allen Wirfs-Brock says:

We are still in the very early stages of the next digital era. If we are going to continue to use  “the web” as a label then it needs to represent a 20+ year vision that transcends web browsers.

This will involve changing our mental model of the web, from a series of linked pages or apps that are viewed in the browser, which is an app on native platforms, to something that’s everywhere—including, but not exclusive to, the browser. Dries Buytaert spoke about this:

I don’t define “the web” as websites alone, but rather as any user experience that’s delivered across multiple channels and devices.

And once we start thinking of it that way, we can start to consider web content less as a destination, and more in the way Jeff Jarvis describes:

A containerized, embeddable article that travels to any site with brand, revenue, analytics, and links attached.

As the tech landscape changes so that information is fragmented across different interfaces—including desktop, mobile, connected object, and conversation—then the web should change with it. Stephanie Rieger lays out an idea for this change in her article, Why Conversational Commerce May Be Our Best Chance To Re-imagine The Web:

If apps are to become mere services within an OS that delivers the best UI for the moment — why should the web not be a part of this change? The web could simply become yet another service. Or a core component. Or better yet— a defining model for how all other services should behave.

The web’s inventor, Tim Berners Lee, this week attended a conference aimed at discussing how to rebuild the web using an alternative, distributed model. Re-imagining the web provides the opportunity to also re-imagine its role: not as an app, but as a model. Stephanie Rieger again:

Once you accept the idea of the web as first-class citizen, it shifts the balance of power. The universally accessible, multi-device and multi-context web is there to enhance native infrastructure — not the other way around.

Why It Matters

The web doesn’t have to become marginalised by new technologies, but we who support it must fight for it. We must not get complacent and believe that the web has any innate right to be the platform of choice. I don’t think users care if they’re using the web or native apps; they just care about completing their task in the most convenient way. Yves Béhar said:

Technology does not have intrinsic value. It has to deliver value by enhancing the way we experience the world around us.

A web that enhances our experience of the world appeals to me more than the web as a diminished app on diminished native platforms. To achieve that will require hard work, consideration, and imagination. But I believe it’s worth it, because I agree with David Herman when he says:

The main reason I care about the web is because it’s the world’s biggest software platform that isn’t owned.

This article is based on Surveying The Landscape, a presentation I wrote and delivered in the first half of 2016 (also under the title I Am Grout (don’t ask)). It’s opinionated and incomplete, and I welcome feedback and criticism.