Breaking Digital Shackles 3

A pedantic approach

Catching Up

In our first piece we provided a background on enshittification, reviewed some proposed recommendations, found them lacking, and decided to plot our own course to a solution based on the principle of promoting interoperability.

In our second piece we outlined the goals and tenets of our approach, dove into the individual use cases for interoperability, and began to sketch the outlines of our solution starting with the use case for individual access and the benefits to consumers for our definitions of interoperability and model substitution.

Breaking Digital Shackles 1

Breaking Digital Shackles 2

Now, we’ll flip the script and work through how our definition of interoperability will impact platforms. We’ll speculate on what business models may emerge. Ensure our framework meets our goals of improving customer experience and the competitive landscape without throwing the baby out with the bathwater. We’ll close by connecting all the pieces and proposing our API Mandate for clients and platforms.

Individual Access: Impact to Platforms

Reminder: throughout the following sections I use the term “Clients” to refer to 3rd parties who build software and products on top of a “Platform’s” data.

So far we’ve focused on the benefits to users and clients of the potential new business models our approach to ending enshittification would unlock. It’s worth pausing here to look at the counterparty, and see how platforms benefit from our proposed model for interoperability. Or at the very least, how it merely shaves platforms rather than slaughters them.

• Consumers are lazy: There is always friction to switching and the costs in time or effort are non-zero. It’s 2024 and most users still don’t use ad blockers. Google pays billions of dollars to be the default search engine on iOS devices. Even when there is an ocean of alternatives, defaults matter. Most users will continue to use the platforms’ default UIs to access their content. Even with automation, generating an API key or following an OAuth flow will be too much for some users. And many users will simply never bother to explore alternatives at all.

• Entrenches Platform Centrality: Would Bluesky, Mastodon, Threads et al. have become serious platforms if it was possible to just make X more like Twitter classic or better customize it to users’ personal tastes? On the one hand, platforms are giving up their hold as exclusive gatekeepers to their content. However, they are compensated for this loss by becoming even more entrenched on the basis of the software infrastructure they have built and the data they hold. This gives their offerings an inescapable gravitational pull. By allowing users more palatable and novel experiences of their content they are heading off the opportunities for potential competitors to emerge. Our proposal makes the platforms stickier and more resilient, even as it may chip at their margins and remove their UX moat.

• Deep Pockets help with Compliance: Similar to the previous item, existing incumbents, and the largest incumbents at that, benefit most from regulatory requirements. Like GDPR and other regulatory regimes large companies can more easily bear the cost than smaller ones. This advantage won’t go away.

• Outsourced Innovation: Many of the symptoms of enshittification are due to ossifying internal cultures. These platforms have developed profitable money funnels but by nature of their core business are locked into trajectories that lead them ever further towards eventual senescence— looking at you FaceBook. Freed from the need to provide a monolithic experience, there’s room for clients to innovate on a platform’s offerings and for the platform to incorporate successful experiments.

These items, taken together, are enough incentive to partially blunt the grumbling and refusals of these platforms who, inevitably, will look unkindly on any attempt to breach their well-fortified positions. What these companies fear more than anything is missing the next big thing and their massive and sustained spending on acquisitions and R&D confirm this. The trade-off is a familiar one, platforms receive safety and security through the centrality of their data and compliance capabilities. And platforms avoid missing the next big thing by having a constant stream of outsourced innovation injected into their ecosystem by clients. This also generates a virtuous feedback loop where platforms are incentivized to provide quality APIs and feature parity to encourage 3rd-party developers to develop for them rather than their competition. Imagine, ByteDance and FaceBook competing to build quality public APIs to drive developers to innovate on their offerings.

To be clear, this is a delicate dance. We are still making the platforms trade known money-spinner for a potentially uncertain future. And replacing bilateral great power competition for a far more multilateral world. How can we do this in a delicate manner that doesn’t smother quality or innovation with regulation? And avoid ultimately destroying or degrading these platforms with our reformist zeal.

Client Requirements

As we proposed requirements on the platforms to offer APIs to client developers to build new customer experiences, we are, correspondingly, proposing certain guidelines client developers must follow for implementing platform integrations:

• Immutable client identifiers: Platforms should know what clients and surfaces (apps, UXs) their users are using to view their content or if they are doing so programmatically. This gives the platforms the opportunity to see how, and in what manner, their data is being used and accessed. From this information they can choose to cherry-pick learnings, copy success, or innovate.

• Interaction metrics: Platforms should know what content and ads users are interacting with even when not on their platform. This extends to advertising data as well which we’ll discuss further. A view, a like, an ad interaction, or any other metric should be just as visible on the platform’s UI as on a client’s UI.

• Revenue sharing: Platforms must be able to make money from experiences others build around their data. We don’t want two-side markets where all the benefit accrues to the market makers and we don’t want to reduce the platforms to mere utilities freed from capitalism “red in tooth and claw”.

The idea here again is that interoperability should not be a dagger in the heart of the platforms but that there should be the space to innovate, aggregate, and develop novel products and incentives.

Value of Individual Data

A slight digression here but I think it’s worth heading off the avenue of conversation that inevitably develops around users being remunerated directly for their data.

One of the great public misunderstandings is that your individual social data has incredible intrinsic worth. Your data is valuable primarily because these platforms have aggregated it at a massive, global scale and built a panoply of valuable products on top of that aggregated data.

In isolation, your individual data is not worth much. Without aggregation in order to sell products or make predictions, it becomes much less valuable. Thus it is quite reasonable much of that value of your data should accrue to the service provider, not you.

Sitting on oil shale has been essentially worthless for eons. On the demand side, it is only a valuable commodity because we exist in highly industrialized society where there are tremendous and variegated uses for hydrocarbons. And, on the supply side, because of tremendous technological and capital investment into extracting that shale oil via fracking. Your digital data is similar without the platforms (miners and refiners) and marketing and advertising demand (industrial consumers) your commodity oil (data) is worthless.

Furthermore, the value of your data corresponds largely to extrinsic value you provide. If you are an influencer, wealthy, live in a wealthy country, and have high content and advertising engagement, your data has a lot of value. You are the casino whale. Individuals with lower extrinsic value have lower value data.

Revenue Models

One of the major impacts of what we are proposing for platforms is a dramatic change to how they make their money. Our goal here is to make a bigger pie to be divided by more parties.

I think there’s a wide spectrum of potential revenue models here with a number of different levers.

1. Ad-free, no-tracking, subscription

2. Ad-free, aggregation and tracking, subscription

3. Fixed purchase, ad-support

4. Fixed purchase, annual

5. Customized ads

6. Auctioned ads

7. Platform ads

At one extreme, if you prefer not to be the product, you have to pay for the product. Especially if you are in a high-desirability demographic group or region. It is potentially reasonable that if you really wanted a totally ad-free experience across multiple aggregated social networks with no data tracking at all, you’d end up paying streaming-esque prices of a say $15-$50 a month depending on how private your data is. For many wealthy, western consumers this would be a valuable trade-off and while the platforms would gnash their teeth at losing some of their whales and more importantly the valuable demographics they represent they would not lose their shirt. This experiment is already getting under way in Europe.

The next tier down might be something like the Kindle tier where you mix fixed costs with limited advertising and tracking.

Amazon is able to offer some device models at a subsidized price by including personalized advertising on the device.

Kindle e-reader display ads as a screensaver when your device is in sleep mode and on your home screen marked as "Sponsored." On Fire tablet, ads display on the lock screen. Sponsored content tiles in the Continue and Discover row on the Fire tablet home screen are an addition to our featured content. Tap on an advertisement for more details. Connect your device to the internet to receive the latest offers.

If you originally purchased a device with ads, you can unsubscribe by making a one-time payment to cover the difference between the device price with and without lockscreen ads. You can find more details on how to unsubscribe from lockscreen ads on Help page: Remove Ads from Your Lockscreen.

Customized Ads

I think the next offering of customized ads is potentially the most interesting and lucrative. Each platform knows something about you and they know a lot about aggregates but they know less about you than the aggregation of your activity across multiple platforms, possibly augmented with additional interactions or data you provide. There is a potential synergy here where users, advertisers, and platforms all have an incentive to play ball.

Users, if they have to look at ads and promoted content, would prefer it be maximally relevant to them. While some digital purists will already be shaking with rage, consumers on the whole seem quite happy with an ad-based model that gives them access to an incredible array of offerings, with no out-of-pocket costs. And consumers are generally willing to share plenty of data about themselves if there is a suitable carrot.

Advertisers don’t want to waste money paying for ads that aren’t well-targeted. And, platforms want to know as much as possible about their users to maximize their value to advertisers.

There’s a lot of interesting wrinkles here. An aggregator client could learn to select among the ads different platform selects for you, they could feed their user information into the platform, or they could provide their own novel recommendations and click information which could then be fed back into the platforms.

There are countless options here:

• Client Feed-In: When requesting your content and ads the client could pass user information to the platform to better tailor the offering in exchange for a percentage of click or ad display revenue.

• Platform Eyeball Auction: Create a revenue model where the platforms virtually compete for what ads get shown to you. Say Pinterest is very certain they have a high value ad to show you, they could attach a bounty bid to their advertisement of $0.005 while Twitter who is less confident of their targeting of you might only pay $0.001 for your eyeballs.

• Client Picks, Platform Winner-takes-all: The client selects an ad from all connected platforms and shares the click revenue and information only with the winning platform. Or shares the revenue only with the winner but the data with all connected platforms.

• Client Algorithm: A client could say we can choose ads better than all the platforms because we have more tailored data on our users and make that ad revenue and data sharing their value proposition to platforms.

I think the space for innovation here expansive and poorly explored. There’s gold in them there hills with real significant benefits for advertisers, platforms, and other non-user stakeholders who can get a number of significant benefits. Again, with all of this being in service of the goal of making compromises that can provide something to all parties.

Connecting the Pieces

So how do we make this all work? It would be quite a shameful about-face if I suggested a regulatory framework as a stick for all this… but alas, that is what I’m doing, although I hope an extremely parsimonious dash of regulation and process will be sufficient.

Financial disclosure and constant modification are the order of the day. Clients and platforms would be in a perpetual, and public, state of revenue and data sharing agreement and negotiation. There needs to be some strong forcing factor that keeps either the platforms or the new client entrants from acting in bad faith. At a high level there are behaviors we want to encourage and discourage:

• Encourage Innovation: We want clients to innovate and provide superior solutions to both sides of the market, users and data consumers.

• Discourage Preferential Treatment: We want to discourage platforms from offering preferential treatment to implementations that are more amenable to their existing state or business model. We want to provide an extremely flexible framework to encourage novelty.

• Encourage Collective Action: We don’t want any single platform to be able to hold out and force ghettoized client implementations that cannot offer either full feature parity (depth) or build the desired experience across all relevant platforms (breadth).

• Discourage Slaughter: We don’t want new entrants to be able to bleed out the platforms and limit their ability to thrive and evolve.

• Discourage Litigation: Our goal should be to create a process that is heavy on flexibility, negotiation and mediation rather than one based on litigation. We need flexible adaptation over time to change how the pie is split by different offerings as the competitive landscape evolves.

A client that simply made a unpaid, ad-free product would be rejected outright, as would a platform that simply rejected a client’s business model or made so financially infeasible an offer as to be a de facto rejection— we will see how Meta’s ad-free version in Europe works out¹. This information should be completely public. Every client should negotiate it’s own terms but the deals should be broadly similar across platforms and clients. A welcome benefit here is it would expose the often murky conversations about how much different data and defaults are worth and how that value is captured e.g. ad clicks, ad spend, tracking, 1st party aggregation, 3rd party aggregation, resale, OS default apps etc.

Initially, a baseline that platforms must retain 80% (or some other number) of their existing revenue feels reasonable, that’s still a massive pot of money for those wishing to build extensions on top of the platform, especially as they can do so much more cheaply than the cost of building the platform itself and the accompanying compliance and plumbing that includes. To be a broken record, our mandate is to shear sheep not kill them.

Given the number of parties involved, an eventual solution might look like a B2B marketplace like Google Ads with clients bidding in a real-time auction for a license to pursue a particular payment or revenue sharing model. “I can provide a custom ad-supported model that shares user data with the platforms for 15% of the click revenue.”

Alternatively, you might need some sort of independent committee to evaluate and set rates for different business models and sign off on revenue sharing models and periodically adjust rates and fees.

API Mandate

This brings us finally to our goal of defining what our API mandate is for clients and platforms.

1. All platforms will henceforth expose their data and functionality through service interfaces using widely accepted standards.

2. These platform service interfaces MUST provide FULL parity in terms of priority, reliability, uptime, and performance to clients as data served by the platforms themselves. And the platforms must provide public service health and status data to ensure compliance.

3. All platforms must allow model substitution when determining what content and advertisements to show to customers.

4. All clients must transparently develop and provide a data and revenue sharing model with platforms and advertisers.

5. All clients must provide all agreed data and revenue to platforms using consistent formats for immutable client and user identifiers.

Conclusion

I hope you have enjoyed this journey. Writing and thinking about this topic has been incredibly rewarding. While delving into the details was challenging at times, the fundamental insight—that equitable API-driven interoperability can drive better outcomes and provide a path forward on one of the major issues of our age—kept me motivated.

There is a better future ahead for both platforms and users—a future that hinges not just on the will of citizens and governments but on building a strong technical foundation for a more equitable model.

Our approach may lack the fire and bombast of a more adversarial strategy and may not satisfy those who prefer more drastic measures. However, I hope it offers a firm, technically feasible, and achievable path toward halting and reversing the unfolding process of enshittification.

Now is the time for action. As technologists, citizens, and policymakers, we have the opportunity to shape the innovative and competitive landscape in which platforms and users will continue to evolve. By collaborating on solutions rooted in interoperability and mutual benefit, we can foster a digital ecosystem that reflects our highest aspirations.

Let us seize the opportunity to create a better world and embrace the technical challenges required to get there.

1. https://about.fb.com/news/2023/10/facebook-and-instagram-to-offer-subscription-for-no-ads-in-europe/