Friday, May 5, 2023

The Next Web Reins In Big Tech — and Improves Everything Else

Solving the data problem restores competition and empowers users


Note: also available on Medium

In Part 1 we covered the mechanics of the next web. This post focuses on the ways it changes your relationship with Big Tech, and showcases the generational improvements the next web delivers throughout your digital existence.

To be clear-eyed about it, decades of dominance have given today’s giants the ability to implement advantages that will be tough to overcome. Think Amazon’s cutting edge logistics, Apple and Google’s ironclad duopoly in mobile operating systems, or social media services’ massive audiences.

However, the next web levels the playing field by introducing and enforcing competition across important dimensions where Big Tech clearly feels they don’t have to try very hard today—most notably privacy, but other control points as well, like the algorithms that dictate what you see. The next web frees your trapped data, so every existing or yet-to-be-born competitor gains the ability to deliver value in any way that the market rewards.

In the following use cases we’ll examine how this lets you take control of your experience on the web. Permanently.

Example 1: Amazon

Amazon has grown to be one of the biggest companies in the history of the world, with low prices, fast delivery, and easy returns. Innovations like Amazon Prime and the company’s cumulative investments in efficient logistics give Amazon an aura of unstoppability to customers, competitors, and platform sellers alike, who grudgingly accept Amazon’s market power… and its constantly creeping gross margins.

Less attention is given to the value of the user data that Amazon holds, and how it creates a self-reinforcing cycle of convenience that only serves to further entrap its customers. If you’re like many people, a purchase decision means opening up Amazon in a browser or app and selecting from among options you find there. Your purchase history makes re-buying a snap and “1-Click Ordering” takes advantage of the shipping and payment information you’ve entered previously.

This is what professional marketers call the “Will I?” vs. “Which one?” process. In a “Will I?” decision, the buyer decides to make a purchase and then proceeds to the “Which one?” stage. This is a natural and logical process and small differences in price or value can be determinant. However, marketers work hard to to eliminate the “Which one?” phase, traditionally through differentiated value which drives brand loyalty — and delivers significantly higher margins. Amazon has achieved much the same result through tactics like Prime 1-Day Shipping and 1-Click Ordering.

Again, the next web works by freeing your data, which restores competition. Let’s explore how that would work with Amazon.

As we discussed in Part 1, you own any data that pertains to you. The European Union’s General Data Protection Regulation (GDPR) and other modern privacy laws make this a legal fact. As a result, any company operating in one of these jurisdictions, including Amazon and all of Big Tech, have put in place methods for you to review, correct, and delete your data. There’s also a term called “data portability” that requires companies to let you download your complete data set in “machine readable format,” ostensibly so you can move to another service provider — but that merely puts you in the same locked-in position with some other walled garden vendor.

The next web makes data portability actually useful. You could download your Amazon history and import it to your pod. There you can access it via a non-Amazon app that lets you do things that Amazon would never design into its own apps. For example, clicking on an item and viewing, on one screen, all the times you bought it and how much you paid. Amazon doesn’t enable this type of easy analysis because it reveals pricing patterns and inflation over time.

Ok, that’s interesting but not transformative. Here’s what is.

Remember that Amazon is not the only merchant with the responsibility to give you your data. Virtually all medium-sized and large merchants have a history of everything you’ve bought, which is tracked via web activity, loyalty cards, or payment methods. That’s how modern hyper-targeted marketing works. So it includes your local supermarket, pharmacy, department store, home center, etc., etc. And all of them must provide access to your complete data set to you.

You can download your data from those other merchants and use the same app you use with Amazon, to construct a full view of your personal purchase history and compare identical items across different merchants. The app might even provide a comparative view of what an item would currently cost across all vendors, and at what delivery terms. Now you’re starting to talk transformative.

Take it a step further. Imagine a service that local merchants, even small ones, can use to see your purchase intent (anonymously) and automatically bid on it. If the offer works for you, pick it up on your way by. Free of shipping and delivery costs, these merchants can offer competitive pricing with fair margins when competing with Amazon and other mega-merchants. And you can keep your money in your community.

Now imagine that you don’t have to bother with any of the download/import stuff at all — the merchants adopt the next web’s interoperability standards (because customers demand it or due to GDPR-like laws) and you can just leave the data where it is. Everything just works. Ok, that’s transformative.

The point here is that, by freeing your data you force competition where little exists today. The next web “locks the web open” — for both you and the thousands of merchants that also feel trapped on the Amazon platform. Amazon will have to compete for the “Which one?” decision again, and that’s good for everyone.

Example 2: Apple IOS / Google Android

Your pod will become the default destination for a big chunk of your most important data. As part of a complete communication and sharing platform, that’s where email/calendar and contacts, files, photos and videos, and texts/IMs will all go. Also, important generated data (metadata) like history of calls, locations, and browsing.

For these apps you will go and import all your previous data from whatever vendors you’ve used. Once it’s all in your pod, you can view and search across the merged corpus, for example clicking on a contact and seeing all communications across all apps and services you’ve used together.

But if all that data now goes into your pod, what about the places it used to reside? If you’re like most people, the vast majority of it was stored by the vendor of your mobile operating system — Apple or Google. According to the next web’s primary thesis, these companies lose what you gain. And that’s exactly the case.

It doesn’t get enough attention, but Apple and Google have built one of the most powerful duopolies in the history of free enterprise. Apple has arguably been better about promoting user privacy than Google, whose entire business model is built on profiling you. But both companies, through the very nature of mobile operating systems, have exclusive access to some of your most important — and therefore most valuable — personal data.

It also positions the mobile OS vendor as a gatekeeper and toll-taker, enabling (or blocking) the connection between content creators and consumers. Both Apple and Google take up to 30% of every transaction, a cut arguably disproportionate to their role in delivering someone else’s creation. I think we all would rather that most of that money goes to the people creating the content we value.

The next web’s architecture creates an opportunity to evolve past this duopoly state. While your pod is in the cloud, it also has the capability to sync with your mobile device. Data can be selectively replicated for speedy local operation, with changes propagating back through the cloud copy and onward.

Containerized versions of this next web client software — basically a partition on your device, much like the mobile device management (MDM) software that isolates your work content from your personal content — could abstract much of what the mobile operating system handles today. The container would have high-level permissions over the OS and hardware, and manage those rights granularly for its own data and apps. You could still run any OS-native apps outside the next web container, but anything inside would be basically invisible to the OS.

This reduces the importance of both the mobile OS and the hardware itself to a more appropriate role in the value stack. It also should make it much easier to migrate between the two platforms, as well as future operating systems should they emerge.

Example 3: Social Apps

I’ve always used Facebook as the subject of this use case, as people have wanted to leave Facebook forever. But here in May 2023, the exit energy clearly belongs to Twitter, so we’ll use that.

There are many problems for someone who wants to leave Twitter, but the biggest one is that you have to start over — which many people can’t get past.

Moving to an alternative service means abandoning all your Twitter content and interactions, including posts, comments, follows, followers, blocks, likes, and re-shares. Even the algorithms need to start over, learning the things you like and want to see.

If you accept the need to start over, you still encounter the problem that others who have left Twitter are scattered among multiple isolated services. You could use more than one of course, but that multiplies the work of monitoring and posting. And even then, finding people to follow and gaining followers require a lot of time and effort.

Here’s how that would work on the next web.

You’ll open a page that lists all Twitter competitors, each with a checkbox next to it. You can add one, if it’s not listed, with a URL. You’ll check the boxes on the ones that you want to use and click a button that says, “Federate”.

Immediately all your Twitter data since the beginning of time will be propagated across all those services. You can thereafter use any one of them that you want, or even switch back and forth between them. Any data you create or generate will flow to all of them automatically. You can terminate participation in any one at any time and your data there will be deleted.

Blue checkmarks or other verification schemes become nonsensical because your identity ties back to your DID, not each service’s “account.” If you use your real name, it’s the same identity that you will use everywhere else on the web. If you prefer an alias, that’s fine too, and you can still view your data within the context of your pod. Getting banned or throttled on one service doesn’t affect any of the others, as they each control their own policies, so someone can always find you somewhere.

That equality is a critical point. Because they all have access to the same data, these Twitter competitors are free to compete in virtually any dimension, including privacy and choice. They could have varying moderation policies, ranging from strict to unfettered. They can have different terms of use (e.g., no data sharing or user profiling), or economic models (e.g, ad-supported or paid). They might let you tune your own algorithms, or even — the holy grail of control — support third party clients.

The point is, this scenario breaks the data locks that make it so hard to leave Twitter, and true competition is the result. Most people regard competition as good for markets.

Note that this doesn’t just work for Twitter, but for all social media — specifically, any service that collects data from users and shares it back to its membership. So Facebook, LinkedIn, Pinterest, Reddit, Instagram, NextDoor, Youtube, TikTok, eBay, etc., etc., etc... All of them will have to share your data with existing or new competitors, and thereafter be forced to compete for your patronage continuously.

Note also that this theme isn’t limited to Tim Berners-Lee’s Solid Project or the larger decentralized web movement. Twitter co-founder Jack Dorsey has advocated for just this type of interoperability, and supports Twitter alternatives like Mastodon, Nostr and his open Twitter competitor, Bluesky. Those systems use Twitter-style data federation, but the Solid approach is designed to work everywhere. It’s engineered to be as universal as Berners-Lee’s original World Wide Web. When everyone has their own pod and identity, the infrastructure is in place for data freedom and a once-in-a-generation transformation in how the web works.

The Bottom Line

The next web will happen because end users and developers will demand it. It’s true that Big Tech hates and fears it; data freedom disrupts their high-margin business models. But a combination of user demand and a perhaps a smidge of regulation can make it impossible for Big Tech to avoid it.

A practical course would be for some forward thinking body to create a law — the EU has traditionally been the leader in these things — that states that companies must support data interoperability within some prescribed time frame, say two or three years. Whether that’s Solid, or some modified version, or something else altogether, it will set our best computer scientists and entrepreneurs on the path to creating it.

And then the web will work the way it ought to: for the benefit of people and society, not just corporations and their shareholders.

No comments:

Post a Comment

Play nice please. Mean stuff will be deleted.