engagement’s battleground has shifted to groups

By jeremy | October 2, 2020

It’s always been about groups. Photo of Minitel by Nicolas Nova

It was ever thus: the internet was always about gathering and sharing. Even before the internet, the attraction of being connected was always less about access to services and information, and more for communing and community. The Chicago blizzard of 1978 helped propel bulletin boards, or BBS, into the mainstream. Usenet groups appeared a year later. The French makers of Minitel discovered around the same time that their test bed of users in Strasbourg preferred to hack the little desktop monitor and keyboard to chat with other users in the same building than find out when the post office was open. Two tools survived the internet’s shift to the web: email and usenet newsgroups.

So it’s no surprise that Facebook has woken up to the fact that if it’s stickiness you’re after, it’s better to build your business around people with shared interest than a shared history. (Of course the two can overlap.) As my friend Mar Pages has shown in her Facebook for groups – Everything you need to know page, this shift to groups goes back to (at least) 2017, when Mark Zuckerberg said that he wanted “develop the social infrastructure to give people the power to build a global community that works for all of us.” Of course, he would say that, but behind that was the obvious goal of trying to increase engagement. Groups had been around for at least a decade (turns out I am admin of at least four, though I can’t remember ever setting them up) but most (90%, according to Mar’s math) were not “meaningful communities” by whatever metric Facebook used, with a membership of only 100 million users. Facebook groups, in other words, were a wasteland.

So Facebook has been working on this, vowing in 2018 to boost membership to 1 billion members of meaningful communities and, in Mark’s (ironic, given what has happened since) words, “not only turn around the whole decline in community membership we’ve seen for decades, it will start to strengthen our social fabric and bring the world closer together.” Overnight it announced that there were 70 million administrators of what it called “active” groups — seemingly abandoning the term “meaningful” — and that there were 1.8 billion people using groups every month. This doesn’t really shed much light on how much progress they’re making, though it’s an advance on a 1.4 billion figure they released a year ago. It’s not unusual to move the goalposts but it suggests to me that Facebook realise that getting from 100 million to 1 billion in five years was a little too demanding, so now we’re getting something much more vague.

That said, there’s no denying the effort being put into Groups, and the boost it’s getting from Covid. But there are questions here that go deeper into the technopolistic world I wallow around in. What lengths will Facebook go to push people into Groups, how will it monetise this space, and when will it move to connect it to a much more active group-based app in its stable — WhatsApp? Finally, I’m wondering where the lines are going to be drawn when it comes to responsibility — will Facebook find it’s becoming more akin to a publisher, or will Groups allow it to offload that responsibility to admins and moderators, and if so, how does it reward them?

how far is Facebook going to go?

So, first question: how far will Facebook go to push us into groups? At the moment groups don’t contain in-line ads, so this is part of what Facebook has identified as an area which has seen increased engagement without monetization. Its announcement overnight was mostly focused on public groups — one you can see and browse without being a member — which will increase their visibility to non-members. It’s also allowing public group admins to generate revenue by allowing them to link up with companies to carrying branded content — sponsored content, to you and I. This does address part of the riddle of groups, which is why should admins and moderators dedicate so much of their time to these communities without any clear way to monetise it? Branded content goes some way to addressing that, though clearly Facebook is wary of allowing ads in private groups, since the subject matter may be sensitive and an ad unsuitable, ill-timed or worse, offensive.

To me the biggest challenge for Facebook is how to monetise WhatsApp which outside the U.S. has largely replaced SMS and is much more popular as a messaging app than Facebook Messenger or any other messaging app. I am a member of dozens of groups, often very specific to an event, a location, an interest or a relationship. It seems obvious that at some point Facebook, obsessed with monetizing ‘engagement’, has to try to lure this group of people into one of its money-making funnels. The Verge reported earlier this week reported that Facebook is now making it easier for people to post messages across Facebook, Instagram and Messenger, an early step in a plan announced last year to essentially combine the three with WhatsApp into a private services platform, not unlike something like WeChat. While Groups isn’t ostensibly part of this, in my mind there’s not much difference. To me there’s a much greater difference between a public Facebook and a private one, than with a private Facebook group and a WhatsApp group. The former is Twitter, essentially, where what you say is by default public, and you know it is (and behave or misbehave accordingly.) Private groups are where you know, or feel you know, the other people on that group, and behave quite differently. While WhatsApp groups are always going to be relatively small, they’re still a huge chunk of our “engagement time” and they’re an advertiser’s dream, as are the private Facebook groups, since they’re focused on a specific interest.

Responsibility

So finally, responsibility. Who is responsible for the content of Facebook groups? Mar tells me there’s a lot of automated culling and cuffing on groups, where even administrators find themselves suspended for activity that looks to Facebook’s bots as spammy. And of course Facebook has deleted thousands of perceived hate groups. So it’s hard for Facebook to claim it’s not responsible for content in these groups, even if most of the actual moderating is left to administrators and moderators, who are not Facebook employees, are not paid or compensated by Facebook, and who are doing the work — often quite time-consuming, not to mention wreaking an emotional toll. A sign of the tide may be an Australian court’s decision in June which holds media companies liable for defamatory posts on their Facebook page. While the lawsuit doesn’t include Facebook, the case led to Australia saying it might treat platforms like Twitter and Facebook as publishers and therefore liable for defamation for anything posted on their pages.

So where does this leave us, and where is this going? For all its fancy talk, clearly Facebook can see the writing on the wall: that the future of social networking lies in shared interests, just like the internet discovered itself. The shift to newsletters has shown that interests cannot be too niche, and that people love to talk to other people who have the same interest. But at the same time there’s a problem: Facebook has lasted this long, in my view, for two reasons: there are still a lot of people who haven’t signed up, and until everyone on the planet has joined Facebook the growth numbers will always look good. But the effort to keep people active on Facebook is like a hamster wheel: you have to keep feeding them compelling content, and there’s only so many foodie photos a user can take. This has led Facebook into rabbit hole not unlike the rabbit holes its algorithms such users into. If you have to rely on increasingly click-baity headlines and assertions to keep people engaged, at some point the snake starts to eat itself.

So there has to be an alternative. And that’s what groups are. Public groups are one thing, but that is not where the value is. That lies in the private groups, where passion, knowledge and value lie. But at the same time, who is in the driving seat for that? If the pull is strong enough for users, what’s to stop the owner of the group moving it to another platform, one which they are able to monetize and which offers more features? Think of a group as a walled garden within Facebook’s walled garden; if people already have to jump over one wall to access it, why not just move the inner walled garden somewhere else, since the friction — the effort required to reach the group — would be no different.

That said, I’m sure this won’t happen, at least immediately. But at some point the owner of a group of hundreds of thousands of passionate users is going to recognise that none of them are there because it’s Facebook, but because Facebook happens to be the platform. Somehow Facebook will have to find a way to lock in group owners with golden handcuffs, or win them over.

a solution to facebook: regulation, yes, but also a simpler tech

By jeremy | September 18, 2020

0 Comment

(Facebook HQ, 2024) Rust | Ampersand72 | Flickr)

The Social Dilemma (Netflix) does a great job explaining how Facebook and Google work, and, dare I say it, completes the picture I’ve been trying to paint in previous blog posts better than I could. It’s definitely worth a watch: essentially, the business model of both (and therefore nearly all digital-only business) relies on capturing, retaining, diverting and monetizing your attention. The sleight of hand is to make you feel that you’re connecting to your friends and interests. Whereas in reality you’re being sucked into interacting long enough to get your attention diverted to advertising.

So we knew that, we just didn’t necessarily know exactly how it worked, and how sophisticated and relentless it was. The documentary’s conclusion: It can’t be beat with technology. It needs regulation.

That’s where I’d disagree. Definitely there needs to be regulation. The Buzzfeed piece on the leaked memo of a fired Facebook employee (I don’t really see the need to share her name as she didn’t consent to it being used) illustrates just how little effort and resources Facebook has put into patrolling itself.

Indeed, the opposite: Bloomberg Businessweek this week shows how much effort Facebook has put into not having to patrol itself. Here’s the piece’s broad assertion:

Zuckerberg isn’t easily influenced by politics. But what he does care about—more than anything else perhaps—is Facebook’s ubiquity and its potential for growth. The result, critics say, has been an alliance of convenience between the world’s largest social network and the White House, in which Facebook looks the other way while Trump spreads misinformation about voting that could delegitimize the winner or even swing the election.

In other words, not only is the world’s largest social platform aware of the extent to which misinformation and misdirection can shape an individual’s beliefs — about themselves, about other people, about their country’s leadership — but that this can actually become a lever for political cover to support the company’s growth.

So it’s obvious, a no-brainer, that regulation has to happen. Regulation needs to be in place holding these platforms responsible, effectively, for every piece of content that is on there, in the same way any publisher or broadcaster is responsible. While some of this could be automated — searching for bots, tweaking algorithms so conspiracy theories don’t become memes that suck the vulnerable further down a lucrative but self-destructive wormhole — much of it is going to have be manual. As the leaked memo illustrates, a lot of this is just too hard to leave to computers; nor can it be done by outsourced drones. I don’t see any other way around it.

Sure, it’s going to hurt their margins, may even make parts of it unprofitable, but I believe the equation is quite a simple one. If Facebook alters in any way content that we post to it then it becomes a publisher, and so acquires the rights and responsibilities of a publisher. It does that all the time, prioritising some stuff over others, converting the links you add — checking first whether it is a banned link or not — to one that will then, when you click on it, warn you that you’re leaving Facebook. All of this is the hallmark of a publisher, so it’s obvious that eventually Facebook, and every digital platform like it, will be required to comply with the law affecting publications.

Not all of this will be good. In some countries restrictions on publications are politically onerous, but it’s inevitable. It will be a test of Facebook’s mettle about whether it will allow itself to become a quasi state broadcaster or publisher under such conditions, but that’s talk for another day.

For now I want to offer a simple alternative. We already have a solution to the problem of Facebook, Twitter etc, if all we’re looking for is a simple way to stay in touch with people, institutions and news. It’s called RSS.

RSS, or rich site syndication/really simple syndication, is a standard allowing updates to be fed from a source to a recipient without lots of complicated clutter in between (accounts, passwords, cookies etc, unless you want them). For a while it was the darling of the world. Certainly mine. It’s been around since the halcyon days of 2000/1, when there was little interest in monetizing content, or dumpster-diving for personal information. Indeed part of its beauty lay in the fact that you could subscribe to something without the producer of that content needing to even know your email address.

Indeed, RSS was the (technical) model that Facebook and Twitter drew on. Arguably it was the failure of RSS to adapt (and get itself out of a fratricidal hole) that opened the space up for Facebook and Twitter to move in. But essentially the principle has always remained the same: a river of content, selected by the user, pushed rather than pulled, which can include text and multimedia.

I haven’t ditched Facebook yet because it helps me understand the scale of the problem a little, but it’s largely a wasteland for my friends, most of whom are no longer on Facebook, or have disappeared from my feed because of poor algorithms (yes, the black box eventually disappears up its own ass, I suspect, as it tries to dig us every deeper rabbit holes).

Facebook, and to a certain extent Twitter, have left the door open to an alternative in part by fiddling with that river model, which leaves open an opportunity. There’s no reason RSS couldn’t rise to the occasion and replace it.

If I could easily subscribe to a personal RSS feeds of my friends in a simple, secure environment, free of sneaky attention-seeking widgets, I would be deliriously happy. We would be free of the neediness of desperately seeking approval, the worry that we may have missed something, the ads, the distractions.

Perhaps someone has done this already. And of course there would be issues. But we have to start somewhere. I’m not talking about a platform; I’m talking about a standard, that could be used by any app, even existing ones like email or chat, or a browser.

And I do agree with two other points made on The Social Dilemma:

that this technology is not going to go away. That the ability to influence us, to outsmart us, is not going to remain at this level. It’s going to improve. And we aren’t showing any signs of getting smarter, definitely not at the same rate;
that we shouldn’t allow those with technology to try to kid us into thinking they have the solution to the problem they have created.

But those points to me suggest another solution: an appropriate technology solution. Perhaps we don’t need something so complicated to stay in touch with our friends. I have resurrected an old (non-digital) birthday calendar and I try to make a point of sending birthday greetings to anyone on it. I feel that is cheating less than having a digital calendar, or Facebook, do it. So perhaps staying in touch with our friends, either on a daily basis or an annual one, doesn’t need to be overly complex either.

Anyone care to start the ball rolling?

the digitally ass-backwards age

By jeremy | August 31, 2020

0 Comment

We’re living in a digitally ass backwards age, where everything you do of commercial monetary value is retained, amazingly easy to recover and commercially mined at mind-boggling speed, but everything else — photos, an SMS, an email you sent to someone last week, the deed to your house — is lost in the muddle of that drive an inch or two from your finger tips.

Mike Bird’s tweet above is one illustration of this paradox. But perhaps the best one is this: I have a vast database of stuff I’ve collected over the years from the internet. But although I use the best tools out there to index that stuff, it’s still much easier to search for those documents on the internet via Google. Easier as in faster and more thorough (more documents, newer versions). It wasn’t always so, however. For one thing, the internet wasn’t always faster than your apps+your hard drive.

For another, Google used to have an excellent tool called Google Desktop which searched your computer for you while also searching the net. This was during the wonder years of 2004 through to about 2008 when software was Good, in both senses of the word, and the idea of indexing your hard drive and making things easy to find seemed like a worthy cause; some of you may recall Enfish and some other efforts in this field, which I suspect Google Desktop killed off before it took its own life, probably having completed its mission.) For a heavenly few years the two were united: what you had in your computer and what was online were one and the same.

Enfish: out of its hands

But Mammon won. Because the simple calculation was: there is no commercial value on your hard drive because privacy prevents Google from mining that data or learning intent from your searches so why would we waste money on that. Once Google had won the search engine wars in the 2000s (see my post on that here) which was what the Google Desktop war was about, Microsoft being their main competitor at the time), there was simply no commercial imperative in continuing to support Google Desktop. Besides, by then we had been persuaded to move most of our email onto the web, so why would we want to be searching our hard drives anyway? There really is no market for what I, foolishly, once believed was the future of the browser. To me then, and now, I still can’t quite figure out why would you NOT search the web and search your own computer at the same time?)

So in short: search is only valuable to the provider of that search if it is to look for things of commercial value, and things we have not yet purchased, and things we might intend to purchase, and have the means of purchasing. It’s not hard, when you look at it that way, to see how things have skewed, in the past 20 years from a paradigm of: “Your computer! It will be where you will create and save all your memories, your documents, your house plans, your photos, all in one place! Forever!” To: “Your devices! You can consume anything you like! Whenever you like! Instantly! Forget about the future! Live for now!”

OK, that’s obvious for the ‘order now from the gig economy’ world. But it started long ago. It started with iTunes, and continued with things like Kindle. Those digital books you ‘bought’? You don’t own them. You bought a license. You die, that book reverts to Amazon. You might be able to save those items somewhere, but you can’t really do anything to them as if they were yours, you can’t search them outside the Kindle app. They are imprisoned.

And yes, it’s true that your photos are saved from your camera relatively seamlessly to other devices, except it’s less than obvious how, evidenced by all the hours you spend trying to find a photo from that birthday party ten years ago that should have been saved to a computer but doesn’t seem to have been because that computer was upgraded a few years ago and doesn’t seem to have been moved across, or the WhatsApp message that was on an old phone or an email that was on an old account which you ditched when they started charging outrageous fees for attachments. Or automatic online and time machine backups which all go smoothly for years until you actually need to restore something in which case the one file you need is missing. And did you know that hard drives actually only last five years tops? Assuming you’ve been using hard drives to back up your treasured memories for 15 years (20 in mine) that means you should have replaced those hard drives at least twice.

In short, even if you have been doing everything right, you’ve been doing it wrong.

Our digital world is built to be commercially optimised for speed, not for ensuring our memories, our records, our heirlooms, belong to us, and are safe. This is not to say companies like Apple don’t try to sell us the whole shebang, make us feel that by selling us computer+services+software that we’re still in that 20-year-old paradigm of ‘your own computer where only good things happen’ but in fact it’s better to think of it as as a big shopping cart. Few of the things you do with it you actually create or produce, and while Apple does offer some tools to help you move stuff from one old device to another, the emphasis is not on safeguarding your ‘digital heritage’ as much as persuading you to keep subscribing to their digital services — to expand your plot in their walled garden.

Now you could argue, I suppose, that Facebook has taken a different approach, building a business on digital longevity, lovingly creating a website that thrives on us tending to our shared memories. And perhaps there’s some truth in that. But let’s explore that for a moment. Can I easily search my memories for something? Sort of. I suppose if I organised things correctly I could find a birthday party album. But the search bar on the timeline doesn’t seem to work, and the best way to navigate seems to be by year. That is a tedious option. Overall, the Past in Facebook world seems to be something for the company’s algorithms to dish up occasionally, rather than a place for us to explore. Like Google, Facebook presumably has less interest in it because it’s not reflective of our present or upcoming needs and desires.

It’s time for us to apply some pressure for the commercial world to stop thinking like this. We have given so much of our lives and data to these vast harvesting machines it’s time to wrest some of it back. And not just to demand the data, but to demand it in the format that we left it in. For example: You can now download your data from Facebook and Google but this is all — intentionally, in my view — in such a mess that it would take a year to put it some order. It is absurd, in my opinion, that the tools to curate our footprint online are impressive (in Facebook, for example, I can create an album called ‘Mum’ to remember my mother, where the photos are dated, geotagged and name-tagged, where they are displayed as thumb-nails and where I can choose who can view them and where those people can comment. This is all assumed as based features.

Now if I want to download that album to my computer, how do I do that? It took me one minute (not too bad, considering) to figure out that (arguably) probably my best option was to download all my photos. This of course may take some time. So while that was going on I decided to right click and download the photos individually. Now to be fair this is not as hard as it could be (in Safari, right click on the image, download as image to the default folder). The problem is that all that richness and effort is lost: The filenames are now meaningless — 30535610150348063416882213385n.jpg rather than 1959 Mum with baby Aldrich.jpg— the metadata is gone, the geotagging, the people tags, the comments, the albums. All that curating work you did on Facebook is for naught. It’s a stark reminder that the tagline about Facebook caring about you and your memories is a flagon of balderdash. (The download process is no better, though the albums are at least in separate folders.)

So that’s the first step. I will talk about future steps in future columns. Perhaps this Covid-19 year (and some of next, no doubt) will not be such a waste. We will realise the devices in front of us are not just oversized credit card swiping machines or pocket-sized cinemas, but actual treasure chests for organizing and storing our heritage, our heirlooms, things we can create and pass on. But first we need to reclaim them. This needn’t be a zero sum game. The Walled Garden machines can help us do this, and expand their business models to include better, more resilient home networks that include resilient storage, tagging and search (more of this in a future column.)

For now, start thinking: all this digital gardening you’re doing? Who is it for? And will it leave it a trace for anyone but those hordes of digital marketers?

one less word, one more brick in the garden wall

By jeremy | August 13, 2020

0 Comment

The change is this: last year Apple introduced something called Sign in With Apple, inevitably abbreviated to SiWA. You’ve probably seen it in action: it sits alongside other so-called social logins from Google, Facebook and others, allowing users to sign up for services and apps with their account at one of those services, meaning they don’t have to enter lots more information, and remember a new password.

Google and Facebook have been in this game since around 2011, and Apple is a little late to the party. They only introduced the option last September, and said that developers only needed to add the option to apps on their platform if the only existing choice was another social login.

So before we get in the weeds here, let’s just take a couple of steps back. What is the benefit to companies like Google, Facebook, Amazon, LinkedIn, Twitter and Apple to offer this log-in service? Surely it’s extra work for them to wire another company’s plumbing, so to speak?

Well, no. As you might have guessed, there’s a significant advantage for the likes of Facebook and Google — they get to track you to yet another corner of the web, while the host service saves money on setting up, managing and securing passwords etc. Yes, you might say, that is true of those companies that make money out of your data, but Apple don’t do that, right? Why would they be interested in doing this?

Here is where it’s a big murkier. Apple’s argument, as I understand it, is essentially this: We want to help keep our users’ data secure, and so if we offer this service — where users don’t need to set up a new password, they just use their existing Apple account — we help limit their exposure to crooks and bad guys. As if to prove that point, they offer an extra layer onto their SiWA: you can make yourself partly anonymous to the service you’re signing up for via a layer of throwaway email addresses, all of which Apple handles and which can’t be tracked back to you.

So no question: if you’re an Apple user and you want to use one of these services, then Apple is the way to go. So far, so good. Until March, that is. On March 4 Apple quietly removed a word — “exclusively” — from its terms for developers to post their apps on the App Store. Developers can’t get their apps onto your iPhone or iPad unless they get approval of the App Store police, and so this word turns out to be quite an important one.

Previously the sentence read: Apps must include the Apple SiWA option if they “exclusively use a third-party or social login service”. When that word disappeared, it meant that now if any app offered any third-party or social login service, they must offer the Apple version as well. Otherwise their app would not be approved by, or an existing app slung from, the App Store.

Baulking

So what in practice does this mean? In effect, Apple is saying that either apps remove all third-party sign-in options, or they include Apple’s as well. This might seem like no big deal, but some developers aren’t happy. One, a company called Pushbullet which (ironically) makes it easier to connect devices together, has pulled its iOS app in part because of the rule which they said would create a lot of extra work (See below). Another developer I spoke to said that after getting their app thrown out of the App Store until it complies they have decided to remove all third-party options and set up their own system, costing them $20,000 and likely some subscribers in the process. “It’s an example of Apple imposing their will as a condition for approving a minor change,” he says.

Some app developers have already baulked at the idea of social logins, realising they lose visibility into and access to their customers. While Facebook, say, will share some of the user’s profile details with the service when they use that sign-on option, many say that most people use for their social login the service they use the least: oftentimes the Facebook profile lacks a full name, or an email address. Others have found that they now have one more layer, rather than one fewer, when it comes to security. Indian top-up service Freecharge said this week it had recently removed social logins for Google and Facebook had been removed “for security reasons”. A spokesperson told me that “As part of our regular reviews, we decided to disable social login and streamline the login process with an OTP as it provides an easier user experience,” without elaborating.

And this is the thing. These social logins are popular among users, but probably for all the wrong reasons. We like them because they’re faster, less hassle and because we don’t have to remember our password. But that’s just it: we feel like we’re giving away less information, whereas we aren’t. We may use our least-used account for the sign-in, but that’s still data being passed onto two services — the app developer and the big technopoly. Some services still go ahead and create a separate account, so you’ve ended up worse off than better off.

Security

There are other reasons to think this isn’t a good deal either. Security is one: All these big players have been hacked at one point or another, so just because we have put all our sign-on eggs in one big basket doesn’t mean the eggs can’t break. And if someone does hack your Facebook account, say, you’ve just made it a lot easier for them to access all the other services you’ve provided your login details to.

Mohammad Ghasemisharif of the University of Illinois, who co-wrote a paper exploring the security risks of single sign-on (the umbrella term which includes social logins)¹, told me: “I do think there is a trade-off here. On one hand, big tech companies have enough resources and talents to allocate towards securing their products. However, they are still prone to mistakes like any other company, with a caveat that their mistakes could have a much larger impact.”

And is it a good thing, asks Troy Hunt, who runs the excellent haveibeenpwned, and has been critical of social logins in the past: “I worry about the conditioning of people to enter their social credentials into a context different to the social site itself,” he told me. At the same time, he sees advantages: “I love that it reduces the footprint of the total number of credentials repositories out there. I also like that it’s a lot harder to screw up from an implementation perspective.”

Never far from the tree

And Apple? Well Apple is right when it says it’s more in the privacy game rather than the data game. Friends I’ve spoken to say that having Apple offer social login means they don’t have to give their details to some random site, and yet still receive emails. They’d rather entrust their data to Apple than to someone mining it for data.

But my feeling is this: that shouldn’t really be what gives you pause. The real reason they (and Google, and Facebook, and Amazon) are doing this is because it adds another brick to the wall and a lock to do the door to keep you in their walled garden. For them, as I’ve said before, more important now is not selling you more hardware — that’s assured, even if you’re not maybe upgrading as much as you did before — but to get you locked into more services that are exclusive to Apple. An Apple SiWA fits the bill perfectly.

What happens, for example, if you swap your iPhone for and Android phone? How are you going to use your Apple SiWA now? (The answer is, you can, but it’s a kludge, and Apple has no incentive to make it any easier.)

Of course, this is also why Facebook, Google, LinkedIn (Microsoft), Twitter and Amazon are in this game. It’s a game of overlapping territories, a grand canvas upon which the great technopolistic game is played. No-one player dominates the whole map, but Apple has belatedly realised that this is a theatre they need to be playing in. Being a technopoly is always about controlling the gates, because as gatekeeper you can decide what goes through and what doesn’t, and what price has to be paid. Being a gatekeeper you can change the rules, as Apple did with no fanfare or explanation on March 4. They can shut things down.

The idea is always to make it easier to move around within the walled garden, so long as you’re not doing anything where you either go out of the walled garden, or your money does.

Does this all really matter? In itself, no. Though what I’m trying to show is how these small things end up becoming big things, but in a way that mostly flies under the radar. I couldn’t find anything in the Judiciary Committee documents about these social logins, nor in the recent UK’s Competition and Markets Authority’s report about the creeping battle to control the sign-up and authentication process. But it seems an obvious new battleground, and now that Apple is in it, expect it to get hotter, and sneakier, as the garden walls grow taller.

O Single Sign-Off, Where Art Thou? An Empirical
Analysis of Single Sign-On Account Hijacking
and Session Management on the Web, https://www.usenix.org/conference/usenixsecurity18/presentation/ghasemisharif ↩

were we ever free of walled gardens?

By jeremy | August 13, 2020

0 Comment

It’s hard to read some of the documents released by the House Judiciary Committee without some rising bile.

What is clear — among other things — is that the internet as a commercial entity has spent more of its life dominated by companies determined on building walled gardens than it has a genuinely free space.

First a bit of history The internet as a commercial environment only came into being through services like AOL and CompuServe in the 1980s, which allowed users to dial in to a quasi-internet service via modem.

CompuServe was launched in 1977, though it wasn’t until the mid to late 1980s that it became a broader service. By 1989 it had half a million subscribers in the U.S., and had services in a dozen countries. (I signed up for it in Hong Kong in 1991, by which time it had 620,000 subscribers and annual revenues above $200 million.)

87b5d2a307e2baad5a50be5bbbab20f83cc027f1a49cfb8e585098c938ce3f30

CompuServe app interface, mid 1990s (photo: Winworldpc.com)

CompuServe was a walled garden, curating the content you could view and extracting a heavy toll via its subscription fees. It did have access to the internet, but it was limited. I could also send and receive emails from my CompuServe account to those with ‘real’ internet addresses, and vice versa, but that was pretty much the extent of it in those years. By the mid 1990s it was facing competition from Internet Service Providers (ISPs) — who offered a suite of apps to give you access to the web itself, and a dial-in number — and AOL. By 1997 CompuServe was owned by its adversary, although it still had a user base of more than 2 million.
(Source: History of CompuServe Interactive Services, Inc. – FundingUniverse)

Indeed, by 1995 there was really no point in having a CompuServe account: just sign up with a local ISP and you had the web to yourself. Suddenly we were away from the walled gardens and we were free to wander the web. There was only one catch: it wasn’t easy to find stuff — something CompuServe excelled in via its directories and spaces — until search engines came along.

I don’t know about you but I still think of that decade, from about 1998 through to 2008, as the halcyon days of user choice, when it came to search engines. We had Altavista, Yahoo!, Ask Jeeves, InfoSeek, Lycos, Inktomi’s HotBot, AlltheWeb, and of course Google. But look more closely, as I did with data I cobbled together from various sources, and you’ll see that Google had stolen the show from as early as 2003:

Search engine market share. Research by loose wire, from various sources

By the end of the decade Google was the pre-eminent search engine with a market share of more than 90%. Yes, Google was a great search engine, but as we know now, this loss of choice carried with it a cost: our world became Google’s world, and Google’s world became ours. It was a walled garden much like CompuServe’s. Only this time it was being directed to Google’s services, or to Google’s growing data trove by being watched by Google’s cookies, essentially indirectly paying Google a tax for its dominance.

But they weren’t the only walled gardeners in town. Once again, we think of the arrival of Apple’s iPhone, iPad and App Store as this liberation from the yoke of Microsoft, boring software and hardware, and a lack of anything interesting to do with your phone. And that is true. What we tend to forget is that Steve Jobs had no interest in delivering an experience beyond the walls of his own garden. It’s quite explicit in the documents, and it’s a sober reminder to me that Apple, from the start of its ecosystem, had no interest in helping the user unless it could cream off the experience.

Here he is in Nov 2010 telling Amazon to use Apple’s payment mechanism or get the hell out:

And here he is in Feb 2011 making clear he wants only one bookstore in the App Store:

There’s nothing unclear about this. There’s no weasel words about protecting the user, or ensuring their experience is as good as possible. It’s old fashioned rent-seeking and wall-building. And of course it’s obvious to most of us. I suppose what is surprising to me is how far back it went, almost to the start.

And Facebook and Amazon are no different. All four appear insatiable, according to the documents, to use their market dominance, for their customers’ benefit, but the opposite: to make sure the rent all other players pay them is maximised, and where competition and choice arise, to snuff it out quickly. There’s nothing in the documents that I’ve read so far (and admittedly, I’m only some ways in) to suggest that any of these companies had any regard for their users in any of the decisions being made and negotiations being conducted. Of course, there’s no need for them to do so; there’s no oversight committee or PR angle that needs to be put front and center. When the TV arc lights are out and they think no one is listening, it’s all about building those walls ever higher.

So, it turns out, as a historian, that the internet has been under the influence or control of walled gardens longer than it’s been free: arguably the only time when we had a significant degree of freedom was from about 1995 to 2002. Certainly by the end of the 2000s we were doomed. I find it intriguing to note that this period was maybe when the web was its most creative, coming up with many of the tools and principles that these technopolies took as their own. But I’ll save that for another time.