The extract below, "FairPay concepts -- economic exchange in the digital era" provides an introduction to FairPay, a family of innovative win-win revenue models that is broadly relevant to social media and AI. (This was "Part I" from a 2021 post that introduced FairPay for the Web Monetization community.) Additional background on FairPay on this blog and many business and scholarly publications is listed under the Selected Items tab, above. [Revised 12/16/24]
(I had developed this wide-ranging family of innovative FairPay strategies beginning in 2010 -- and evangelized those ideas with hundreds of businesses and researchers -- until shifting my focus to broader issues of social media, and now AI, in my other blog. I hope to provided a more fully updated presentation of how these concepts apply now, when I can.)
The FairPay framework mixes and matches elements that range from simple and familiar, to new and sophisticated. While many of the elements are already in common use, the new advanced forms generated great interest, but limited uptake. A key gating issue for the most advanced forms was the novelty and cognitive load on users of a fully adaptive and participative pricing process. But now, the expected rise of Personal AI agents (PAIs) promises to largely eliminate that problem, as a customer PAI agent negotiates with each vendor's AI agent -- "Have your AI call my AI." Importantly, that negotiation is not over price, but over how value is understood and how the value surplus can be shared fairly by both parties. Use of an AI agent can greatly reduce the cognitive burden on the user, and the vendor, thus simplifying development and adoption. Please read this prior explanation with the idea of gradual introduction of such Personal AI agents in mind.
To get directly to the core elements, from simple to complex, with less background, see The Elements of Next-Gen Relationships and Pricing -- A Unifying Framework.
...or continue with this broader conceptual overview...
FairPay concepts -- economic exchange in the digital era
A thought experiment about value -- Reisman's Demon
Imagine a supernatural demon that might power a system of commerce. This demon has a "god's-eye" view, a perfect ability to observe activity and read the minds of buyers and sellers to determine individualized "value-in-use:"
● The demon knows how each buyer uses a product or service, how much they like it, what value it provides them, and how that relates to their larger objectives and willingness/ability to pay. It understands that the value of a given item or unit of service depends on when and how it is experienced. It is also aware of broader/external/social value impacts.
● This demon can determine the economic value surplus of the offering -- how much value it generates beyond the cost to produce and deliver it.
● The demon can go even farther, to arbitrate how the economic value surplus can be shared fairly between the producer and the customer. How much of the surplus should go to the customer, as a value gain over the price paid, and how much should go the producer, as a profit over the cost of production and delivery, to sustain their ability to continue those activities.
Even if we lack such a demon, we can internalize it as an ideal, and design relationships and pricing methods that seek to approximate what it knows. This demon would apply all of the elements described below. Advanced forms of FairPay apply all or most of them. Keep this demon and its sense of value and fairness in mind as you think about which elements of FairPay you might apply now, and which you might add in over time.
What? Yield my privacy?
Obviously, the demon’s “god’s eye” view sees through any cloak of privacy. Thus the challenge in applying the insights of the demon will depend on which actors can see through which privacy constraints to what extent. As noted below, this can depend on which actors are trusted for what purpose, and one approach to facilitating that with maximum privacy is to interpose trusted intermediaries that are legally obliged to act as fiduciaries on the user’s behalf to safeguard their data. Otherwise, if users demand absolute privacy, that has a cost. If you are a privacy absolutist, this may seem a major turnoff, but if you think through the economics outlined here, you may see why there might be a case to seek a more balanced solution -- to get both a level of privacy and the benefits of the Relationship Economy.
Note also that in many contexts we have not just a single buyer and seller, but intermediaries in a value chain, including potentially aggregators/bundlers (including the special role of WM “providers”). My demon would be able to arbitrate a sharing of the value surplus all the way down the chain (to the extent that privacy constraints do not preclude that).
Digital changes everything
The digital era brings two interacting sea-changes to value exchange. They have been increasingly understood separately, but how they combine has been largely ignored. The failure to consider that is why we have been frantically automating our old logic, making it more efficient at faulty economics, and wondering why things are getting worse. My demon sheds light on how they fit together, and how we can do much better.
- The Invisible Hand fails for digital services. We are steeped in the pricing model of classical economics: the market uses prices to ration scarce supply against demand. But the invisible hand flails at digital services because there is no scarcity of supply. Suppliers have turned to "artificial scarcity" to maintain prices (using paywalls and Digital Rights Management) but users rebel at that as an obvious artifice that seems hostile. "Information wants to be free." FairPay shows how the solution to this new problem is tied to the other change.
- Commerce has been moving from one-shot games of transactions to repeated games of loyalty and cooperation, to put it in game-theory terms, Traditional mass-marketing has centered on one-time transactions targeted to an endless universe of consumers -- lose one, find another. But we are increasingly turning to the superior economics of repeat business and "loyalty loops," especially with the emergence of "The Subscription Economy," "The Membership Economy," and the “Creator” or “Passion” Economy. I call this "The Relationship Economy" because even in the age of "1:1 marketing" and "mass-customization" we have barely begun to realize how improvements in computer-mediated relationships will empower mass-customized, 1-1 value propositions.
These two come together in FairPay, because we need a new social contract to sustain creation of value. The only way to justify a fair price to creators of digital value is to understand that we are not paying for current value, but to sustain the continuing creation of future value.
● The only way to sustain a one-shot game for valuing an item of digital service is artificial. That makes it a zero-sum game in which the price will be based on pricing power, not win-win cooperation on co-creating value.
● By shifting to a repeating game of relationship, we can create an "Invisible Handshake," a win-win process for seeking to find a fair price for each customer that can sustain and incent ongoing creation of the services they want to have available in the future.
FairPay provides methods for adaptively seeking actual fair values in a win-win way -- for each consumer at each stage of this repeated game. That suggests ways to blend the best features of micropayments, subscriptions, memberships, and tipping/donation models to suit each business-customer context. [Lack of support for ongoing relationships, at least at higher levels, would limit the use of Web Monetization and payments protocols as a base for such options.]
FairPay is a user-centered strategy framework and architecture, not a product
FairPay is not a product, platform, or protocol. It is a framework and architecture that can be embodied in products, SaaS services/platforms, and/or protocols. My blog has extensive background (tabs list key items from the blog and other publications (including a book, Harvard Business Review and two scholarly journals, Inc., and Techonomy) and conferences. The concepts of FairPay have been evolving since 2010 in discussions with hundreds of businesses (and non-profits) of all sizes, in a range of digital content and service businesses to much praise but are still not widely understood. FairPay combines a set of well-proven elements, drawing on recent development in marketing theory, behavioral economics, and game theory. The most advanced combinations still need testing to prove and refine, but the foundational elements that have already been proven offer clear lessons applicable to most conventional commerce.
Many view economic exchange through a political lens. Many businesspeople have hard-headed zero-sum attitudes, but many are more enlightened about win-win consumer and social value. Many consumers, creators, and technologists have an anti-business perspective. FairPay seeks to find a balance that is fair and win-win for all (as my demon would see it). My prime motivation in developing FairPay has been to transcend the apparent conflict between market capitalism and social values. FairPay seeks a logic for economics in the digital era that harnesses the genius of markets in a way that moves beyond the invisible hand to the invisible handshake that can restore human values to economics. I hope you will see that shine through whatever perspective you start from. I think this change in perspective can change the world in a way that all parties can embrace.
My mission is to evangelize the concepts, and to advise those who seek to implement and test advanced forms of FairPay as I can. (More background on how I came to this is in the endnote below.)
Micropayments, subscriptions, and pricing risk
The consumer risk is very different for micropayments versus subscriptions. Stefan Thomas' 2018 article notes the issue of both models being built as closed systems, and other Coil and Web Monetization documents refer to issues of friction and privacy, but I suggest it is the issue of consumer pricing risk that is paramount to user acceptance of these models.
Starting with micropayment pricing risk, many are familiar with Clay Shirky's 2000 classic "The Case Against Micropayments." (Not so many know Andrew Odlyzko's more scholarly paper with the same title, based on his work on telecommunication economics, where micropayments have a century-old history, such as for minutes of long distance usage.) Key issues are:
● Most pressing is the problem of "the ticking clock," the constantly incrementing meter, which brings the risk of "bill shock" when micropayments add up to macropayments. That has long been known (as has the resulting consumer preference and higher willingness to pay for flat rate plans, even though they cost the typical consumer more). This problem can be somewhat reduced by providing for volume discounts and for price caps and other variabilities, but simple micropayment models rarely allow for that at all, or do it with just a few usage tiers (as for mobile phone minutes or data gigabytes). But volume discounts and caps require tracking usage over a billing period, and that requires a persistent identity.
● There are also problems of paying for items that were not satisfactory -- and of scanning many items lightly but having to pay full price for all of them. Some services enable ways to adjust for that, but again, that may require an identity.
Those risks to the user can be countered by aggregators that offer consumers flat rate plans (like Netflix, Spotify, and Coil). These might be considered hybrids that charge consumers on a subscription basis but pay their suppliers on a micropayment basis. That shifts consumer pricing risk to the issues with subscriptions (next). It also shifts an unfair level of pricing risk to the suppliers. Instead of set unit rates per item, the suppliers get some share of the flat rate, so heavy users generate rapidly declining unit rates of payment that quickly diminish to zero. The whales who should be generating the most revenue instead get an unsustainably high discount. Aggregators struggle, and creators struggle even more.
(Keep in mind that the “micro” part of micropayments can occur at the actual payment level, or just at a metering level. Most traditionally micropayment pricing models are not actually paid as individual micropayments, but metered and then totalled into the monthly bill. That reduces friction and transaction costs, but leaves the risks.)
Subscriptions involve a rather different set of pricing risks to the user:
● Most B2C subscriptions are flat-rate, all-you-can-eat (AYCE) plans. That eliminates micropayment usage-based "bill shock," but brings the new risk of not using enough in any given period to justify the price. It also continues the risk of not being happy with what you used.
● As subscription models proliferate, the new problem of "subscription hell" has become a major issue. If every provider of video or news or magazine content demands $5 or $10/month (even if you may only want one item per month), you need to spend a fortune to have access to all the content you want. That leads to the wasteful pattern of subscribe-binge-cancel.
● Bundling of services, as with cable TV channel bundles offered by aggregators, can provide a level of discounting across multiple suppliers, but then consumers face the problem of "bundle hell." How do I choose in advance among all these bundles? How do I know if I will want to watch HBO or Showtime or Cinemax or any bundle of premium channels in any given month?
No wonder so many consumers, publishers, and creators are frustrated and angry. That is why we need to step back, reexamine where we are, and look for a new logic.
Keep in mind these fundamental questions of pricing risk:
● Who takes the pricing risk? Buyer or seller or both? Remember that both value and risk are a function of the price, the experience or outcome of an exchange, and the time the price is set.
● That breaks down into two questions: who decides the price, and when do they decide it? Only when the answers are right will the value exchange be efficient and fair.
The Relationship Economy
Relationships are even more central to the subscription and membership models that are increasingly dominating commerce. Such businesses are very focused on Customer Lifetime Value, to offset the fact that customer acquisition costs are high and churn is costly.
Currently, much of this focus on relationships is one-way. They want to know how to target you and sell you on their formulation of a value proposition. But the most enlightened businesses want to not just talk at you but listen to you. Computer-mediated communications are making it increasingly easy and essential to build real customer relationships that seek to understand and center on value as perceived by each customer.
(For much more on this theme, see The Relationship Economy -- It's All About Valuing Customer Experiences.)
The history of the price tag
We tend to forget that the price tag was only invented in the mid-1800s. We just assume that sellers set a price and customers take it or leave it. It has mostly been that way through our entire lifetimes.
● But for most of human history, prices in village markets were customized. Prices (in money or barter) emerged from individual negotiations in personal contexts, depending on needs, bargaining powers, and relationships. They generally reflected win-win "communal norms" including caring, fairness, and even generosity.
● The price tag was invented by John Wanamaker and others when they first built large department stores (see this amusing video). Sellers became institutional, targeting a mass market of "consumers." They had to be scalable and efficient, and thus to limit the discretion of sales clerks. That changed everything: the “take it or leave it” offer led many to leave it -- leading to bargain-hunting and feelings of exploitation and alienation that have become endemic and still worsening.
Now we are in an age of mass-customization and 1:1 marketing -- why not for price? The question is how to do it fairly, effectively, and efficiently at scale.
Value-based pricing
The challenge is that being value-based is complex. Being usage-based is a start. Units of usage may be items, minutes, miles, etc. and more units usually correlates to more value. But not always, and usually not in a linear way. Value is a function of many dimensions, each possibly involving different units of usage -- and still there is much more to value.
Beyond usage, outcomes or performance are increasingly recognized as a truer measure of value. Increasingly, businesses are realizing that they are not selling products or items, they are selling desired outcomes, often in the form of an experience. One of the best business books of 2020 is The Ends Game (coauthored by Marco Bertini, the marketing scholar who was my coauthor on an HBR article and a journal article on FairPay).
Truly value/outcomes-based pricing is still limited in consumer markets, but it is increasingly considered best practice in B2B markets. This is especially true for big-ticket items like industrial machinery, because determining value is complex and costly. It is still a challenge to do it well at scale for small transactions. But subscription services company Zuora has shown that even basic usage-based pricing can be powerful as an element of a total strategy. Their tracking of data on over 900 companies that they serve found that the fastest growth was with some non-zero level of usage based pricing, but less than 50% usage-based.
But usage-based pricing takes us back to the per minute or per gigabyte micropayment models that "consumers hate" as Shirky said. FairPay suggests a smarter way that consumers may come to love. But, before we get to that, how does the new challenge of digital change things?
Digital changes everything about pricing and customer relationships
We all know the classic dilemma of digital pricing:
● "Information wants to be free" because it can be infinitely replicated at essentially zero cost. It is a world of abundance. Consumers have become used to free, and freemium, and ask “why should we pay anything at all?” But...
● "Information also wants to be expensive" because it often has very high value and is usually costly to create. Creators need to earn a living and invest in creating more information.
To resolve this dilemma, we must re-think the core assumptions of our value exchange process to find a new logic. We are no longer allocating scarce resources with the invisible hand, but we need to sustain creation of future services. How can we do that in a way that balances value, ability to pay, cost, and a fair profit that creators need to live on?
Hint: most consumers are willing to pay even when they don’t have to, if they feel you deserve it. But before explaining how FairPay addresses that, a few more key ideas.
Experience goods and the long tail of customer demand
Conventional set prices lop off the long tail by refusing to make sales to those unwilling to pay the set price. This eliminates a potentially significant market, out of fear that selling to those buyers will cause the other buyers to demand lower prices. Conventional set prices also lop off the top of the fat head, since the seller gets only the set price, even from those who might be willing to pay more. So, revenue is only the green box, even though there is a red surplus at the top of the head, and a long amber tail to the right. This shows the huge opportunity that FairPay opens up. No matter what price you pick, it will be right only for a small fraction of your potential buyers. If my demon could set individualized prices, we could get revenue from all interested customers.
This long tail problem is especially challenging because digital products and services are not discrete scarce “products,” but actually services that are “experience goods.” We are buying access, entitlements, and usage – and the broader outcomes they enable. The nature of that value is very personal and depends on how many units of service are consumed, over what time, with what intensity, and with what results.
On the plus side, digital services are highly measurable. There is rich data on what was consumed and how. That has a cost to privacy, but the gain is that it can enable providers to understand at least key components of the value provided to each user. That data on individual consumption can be used to help customize prices, and that customization can be very good for the user when done fairly.
Free and freemium to reduce pricing risk
The marketing value of free trials has long been known, and digital has led to a plethora of variations on free, in the forms of freemium, pay what you want (PWYW), crowdfunding, tipjars, and free trials. Clearly smart use of free services can reduce pricing risk, but what is the right level for a given user? The problem with freemium and other free offers is that it is still a pre-set price – what is the line between free and paid? However you may set it, it will be wrong for many users, much of the time. And because we are dealing with experience goods, what is right for a given user at one time, will be wrong at another time. How can we embrace this dynamic variability and manage pricing risk?
Post-pricing – separate the sale from the price
The realization that sparked my original conception of FairPay was that I would be happy to pay for a service that I valued after I had experienced that value. After seeing the long history of digital pricing challenges and watching trials of “freeware” software and PWYW offers (such as Radiohead’s widely noted PWYW album offer in 2007), I was reflecting on some services that surprised me at how much I valued them. It struck me that I never would have been willing to agree to pay up-front, but I would be happy to pay in hindsight.
By pricing the experience after it is known, you remove the customer’s risk discount. Up-front pricing requires the customer to discount for possible disappointment, and that risk often leads them to not buy at all. And even with PWYW offers, how do we know what price we want to offer? Studies show that many buyers balk, rather than deal with that uncertainty. Strangely, very few PWYW offers let users set the price after the experience (but recent studies now show that those are far more effective).
Tipping models are the primary exception to setting prices up front. But even so, the challenge is to get the customer to actually pay after they have consumed the service.
Managing pricing risk
Who takes the pricing risk? This comes down to two key issues:
- Who decides the price? We are conditioned to think it is the seller, but with PWYW and similar offers, it is the buyer. It can be joint, in the case of auctions, and was traditionally joint in village markets. Clearly joint determination has the potential to best manage risk to each party, and to apply the fullest information from both parties on the nature of the experience and its costs.
- When do they decide it? Before the selection, at the time of selection, or after the experience? If I buy a cable TV bundle, do I know which premium channels I will want when I subscribe? At the start of each month? When I select each program? Do I know what price seems fair before the end of the month’s viewing?
FairPay leads to multiple levels of answers. It suggests the best answers are when the decision is joint, and after the experience. That gets closest to what the demon knows. However, even if the seller unilaterally decides the price, they can do that better after the experience.
Remember, for digital services, the provider risks nothing …except the opportunity to take money in exchange for no value. That will be less and less tolerated.
How FairPay changes the game
FairPay centers on value throughout the course of an economic relationship by creating a repeated game that seeks fairness and cooperation by empowering customers, engaging in dialog with them, and tracking their reputation for fairness.
· The conventional repeated game is a one-sided game of customer loyalty: "Here is our monthly price, take it or leave it. We hope you will take the risk--and be satisfied enough that you will continue this game."
· The FairPay repeated game is a cooperative game of joint fairness: "We will remove your pricing risk by letting you pay what you think is fair for you after each month's use--but we will continue this game (beyond a few trial cycles) only if we agree that you are being reasonably fair."
1. The seller sets the basic rules up front, explaining how this new model works, gives the buyer access, and then at the end of the period reminds the user what they used and suggests a price they think fair.
2. The buyer has access for the period, reviews the results and suggested price, and is free to adjust that up or down as they think fair, and is invited to give reasons for any adjustment (using multiple choice selections).
3. The seller decides whether to repeat the game by tracking the price and any reasons given, assessing its fairness, and considering fairness over prior cycles.
The seller can nudge the buyer toward being fairer and more generous. They would be especially lenient for an initial learning period, treating that much like a free trial.
- It learns to find the value sweet spot for each customer, and to dynamically segment customers based on what they value and their fairness reputation.
- Fairness can be enforced as strictly or leniently as the seller desires for any given customer or segment.
- Alternatively, there can be no enforcement (making payments purely voluntary), but still set after the experience, and still in a process that can individually nudge toward generosity.
- It can be combined with conventional models and offered as a privilege to the customers who will be most delighted and fair.
- Instead of using occasional sampling and focus groups to discover the right value proposition in an artificial setting, this can constantly test and review value propositions for each customer on each real transaction cycle.
Aligning price with value in the broadest sense – in both directions
Because the FairPay process sets value in dialog with each user, it can factor in whatever aspects of value the two parties agree are relevant. That value can include aspects of value that are generally ignored in pricing, and aspect of value the go from customer to the “provider.”
- From provider to consumer: not just value in use relating to specific experiences and outcomes, but other “soft” value, including service and support; participation, listening, and responsiveness, and social values to the community, environment, etc.
- From consumer to provider -- a “reverse meter:” Beyond monetary payments, the dialogs on value can incorporate other currencies, such as negotiated levels of attention to advertising or use of personal data; credit for user-generated content or other co-creations such as participatory journalism, and virality, leads, and volume/loyalty discounts.
Again, the value proposition can consider whatever factors both parties agree to be relevant. This can include ability to pay in far more nuanced ways than now common with student and senior discounts. This can include any aspects of Corporate Social Responsibility (CSR), Triple Bottom Line, or Environment Social and General (ESG) – and brings them in to the main financial bottom line. It can make explicit the now implicit price premium expected by businesses that gain customer approval as good corporate citizens.
This framework can also extend through the ecosystem value chain. With the “reverse meter” negotiated attention to ads makes the user become the customer so that ads are more relevant and non-intrusive. With aggregators, customers can designate a value share to specific favorite creators, such as to the artists most listened to and appreciated on a music service like Spotify. Such contributions can be a voluntary layer on top of any standard pricing -- effectively a tipping layer on top of set pricing for the basic service).
Dynamic value discrimination not price discrimination
Many see that FairPay is a form of dynamic pricing and ask about that. Price discrimination rightfully concerns consumers because as currently practiced, it is usually done in stealth, as a way to extract as much of the consumer’s value surplus as possible. But with FairPay, this is transparent, and the customer opts in to the dynamic price, actually being the one to set it.
- “Discrimination” can be a negative word, but with FairPay it become “self-discrimination.” That is why I refer to this as “value discrimination” rather than “price discrimination.” I argue that “price discrimination” can be good when it is “value discrimination.”
- FairPay engages consumers in a rewarding process based on jointly customized value propositions that lead to fair segmentation in all dimensions of value: context, usage, time, number of users, devices, and ability to pay.
Value discrimination leads to optimal co-creation of value with optimal sharing of the value surplus.
Key evidence and enablers – not as crazy as it may seem
While full forms of FairPay have not yet been proven in practice, the elements behind it are well established.
Modern behavioral economics sheds light on how this strategy leverages human nature. People are not Homo economicus, purely rational profit maximizers who will never pay any more than they must. Thousands of PWYW success stories and dozens of research studies prove that people are Homo reciprocans, driven to reciprocate fairness with fairness (and even altruism). This applies in two ways:
- Traits: Individuals vary in how inclined they are to fairness, reciprocity, altruism and related traits that affect how generously they are willing to pay. This argues for segmenting users based on their fairness traits.
- Situations: How these traits apply depends on the nature of the relationship. Economic/exchange norms are coldly business-like, favoring hard-nosed quid pro quo behavior, while social/communal norms are more friendly and human, favoring more flexibility and generosity. Even business relationships can enjoy more social/communal norms when both sides have positive feelings and trust toward the other. This argues for building the kind of relationship that shifts customers toward those favorable behavioral norms.
Game theory shows that when a repeated game is well designed, players will invest in a positive fairness reputation in order to gain a continuing privilege. If they see that FairPay is a privilege that benefits them, they will invest in fairness so they can continue to enjoy that privilege. (My resource guide links to many studies, including some showing that, as well as some showing that post-pricing enhances the results of PWYW offers.)
Computer-mediated dialog is rapidly improving to enable sophisticate dialogs about value to be handled at scale with only limited need for human intervention. Simple decision rules can track fairness to control nudging and offer decisions. As artificial intelligence, predictive analytics, machine learning, and natural language understanding are applied, the process can gain sophistication and nuance. A wealth of usage data will enable validation of whether customer’s reasons for paying less that suggested are honest.
This background bears on the most common concerns about FairPay:
The first is: Will people really pay if they do not have to? The behavioral economics makes it clear that most people are happy to pay when they think it is only fair that they do so. Hopefully, the above discussion and the evidence in my resource guide shows how FairPay achieves that, for most people, and allows the free riders who will not play fairly to be sorted out.
The second is: Isn’t this much too complicated, putting too much cognitive load on the customer? That may be the greatest challenge, but there are ways to limit that. Most importantly, this is a learning experience, and most of the learning will happen in the first few cycles. Once each customer has gone through a few cycles, the business will learn what each customer values and why, and how fair and generous they are, and the customer will see that. The pricing process can gradually go on autopilot, subject to correction whenever needed. The suggested prices can be charged, with the understanding that the customer can come back within a reasonable time to request an adjustment if they so desire. An even better solution might emerge in the form of user agent bots that can largely offload that cognitive load from users, as described in Part 2.
[[And, especially in the Web Monetization community, there is a privacy concern.]]
Also, consider that there is a well-established model that works much like FairPay – tipping at a restaurant. We need not tip anything, but most people do, especially at a restaurant that one frequents (and especially in the US, where tipping is the norm). We may have a default model of 20% or whatever, but after each meal we do a complex multivariate assessment that considers many factors, such as courtesy, helpfulness, efficiency, and broader values. We can do that in a fuzzy way, with great nuance, usually in no more than a few seconds.
Of course tipping has a cognitive load, and different people and cultures have different levels of comfort and openness to the norms of fairness in tipping. As noted, cognitive load can be reduced by learning to predict what the user will consider fair so the user can opt-in to autopilot mode (with options for retroactive adjustments) when those predictions are converging well, and by nudging toward communal norms that motivate fairness. And unlike FairPay, tipping is done without any provision for transparent dialog on the perceived value received, the reasons for the amount of the tip given, and the server’s feedback on fairness. Uses of FairPay would seek to frame norms for such dialog.
[Update 12/16/24: Again, while this may seem complex, keep in mind that AI agents will increasingly hide the complexity -- a user pricing agent will interact with the vendor pricing agent to handle these dialogs with increasingly limited need for human oversight by either user or vendor.]
FairPay works because it does not have to be right all the time. It is enough that it is approximately right most of the time, the errors tend to average out, and it converges toward increasing accuracy as we continue to learn. And because we are dealing with low marginal cost services, the seller can err in favor of the customer whenever in doubt.
FairPay shows how digital can enable a return to human values
It may now be apparent that FairPay seeks a return to traditional communal approaches to value exchange that have largely been lost and almost forgotten in our modern world – but seeks it in a new way.
- My value demon is not new -- it is just a simple formulation of the model we use when human peers exchange value -- as we have done for millennia in village markets.
- Negotiating customized prices with the joint participation of the buyer and seller is not new – FairPay just gives it a new twist to deal with digital abundance, where the scarcity is sustaining creator resources and share of customer wallet, not of current supply.
- Centering prices on rich, multidimensional considerations of value, and with respect to fairness, and communal norms is not new – that was how humans exchanged value through most of history.
- It is the alienated zero-sum game of mass marketing that is relatively new, an artifact of the need to scale with inadequate technology -- and that is a problem we can now transcend.
- Our ability to transcend that is still limited and unfamiliar, but as we learn, and improve human-centered technology, we can use automation, AI and ML to enable businesses to act more like humans that have a real relationship with each customer – and we can create agent services that protect customer fiduciary interests.
Getting to FairPay -- Deconstructing the Elements of FairPay
The full form of the FairPay repeated game described above is obviously a significant change in perspective for both businesses and consumers. Some people I talk to get the idea immediately and love it, some are stuck on how to get there and whether it can really work. Passion economy creators and service providers are among the biggest fans of FairPay concepts, but most lack the resources to implement the software. That creates an opportunity for entrepreneurs to facilitate that with SaaS offerings.
FairPay will be most applicable in the near term to business that can create passion and loyalty in their customers [[which seems the case for many candidates for Web Monetization]] – but most businesses can do that at least for selected customer segments. Many posts on my blog address strategies for determining which kinds of businesses, which services -- and which customer segments to tackle first. Moves toward FairPay can be stepwise, beginning with baby steps, many of which are in wide use and well-proven, as outlined in many of those posts.
My post on The Elements of FairPay deconstructs FairPay into a framework of synergistic elements, as shown in two tables. These elements can be applied in whatever combination fits any business context, to move it toward fairer, more effective, and more efficient relationships, whether in isolated baby steps that are largely conventional, or in fuller and more novel combinations. In that sense, FairPay is an innovation architecture for transitioning any business to become better centered on customer-relationship-value. That post provides a helpful framework for exploring what lessons FairPay offers [[regarding Web Monetization and payments]]. Here we outline the most relevant aspects.
 |
The first view of this chart (above) suggests a “ladder of value,” beginning with elements that become more value centered. The elements are listed in the rows, starting with the most foundational elements (the lowest rungs) and then elements that amplify the power. The columns are suggestive of which elements are most relevant to for-profit and non-profit use cases, with sub-cases for what is common now, and what is most applicable in low-trust versus high-trust environments. The second view (below) defines some important combinations of elements relevant to different stages and use cases.
Full “Gated” FairPay includes enforcement of fairness to continue playing the game, making FairPay a revocable privilege. Voluntary FairPay relaxes that to serve as an enhanced form of PWYW or tipping that still includes key features of post-pricing and reputation tracking to enable individualized nudging toward fairness. [[Risk-free and FairMicroPay are variations that may have special relevance to Web Monetization.]]
Risk-free subscriptions
As I said above, for digital services, the provider risks nothing …except the opportunity to take money in exchange for no value. That will be less and less tolerated.
While the best way to manage risk is to set prices with customer participation, in the full FairPay repeated game, I propose this “risk-free” model as a way to approximate that while maintaining full seller control of pricing. Many businesses are hesitant to yield control until FairPay is more proven, so this is a way for the seller to predict what the user would do, to set prices based on the seller’s best guess of what my demon would work out.
Think of it as a cable TV bundle that lets the customer view whatever they want each month, then creates a bundle price as if they had picked a bundle that would give them just that. That price can factor in standard versus premium programs and how much was viewed. It can have a cap on price to avoid risk of “bill shock.” But I can also start at zero if nothing was viewed that month, and ramp up at a non-linear rate that can includes a volume discount. A full description is in my post "Risk-Free" Subscriptions to The Celestial Jukebox?
FairMicroPay
This simplified form of FairPay was developed in discussion with businesses seeking to monetize content using cryptocurrencies [[in a way that seems to have parallels with Web Monetization and payments]], to add a relationship value-based adjustment layer. The idea is to add a FairPay layer that adds this:
- ·Let the user adjust the standard base price within limits, as permitted by a smart contract:
- o Downward as a volume discount, or as a refund/discount for lack of desired value, or
- o Upward as a value-based bonus or sustaining contribution.
- · Identify each user and track their fairness reputation, and nudge and alter price adjustment limits accordingly.
[[That might achieve much of the functionality of FairPay in a lightweight way, and that is what I suggest be considered as minimum functionality that can be layered on top of Web Monetization and payments protocols. This is explored further in PART 2.]]
Aggregation and value-based pricing – no more “subscription hell” or “bundle hell”
FairPay can be applied by individual creators/service providers or by aggregators. It provides a new way to harmonize both direct and aggregated models because the value-based prices it seeks are similarly aligned in either case. With full FairPay, or even the more limited “risk-free” model, what you pay relates to the value of what you use, regardless of whether the relationship is direct or with an aggregator.
· If you want to access a broad array of services with no fuss, use an aggregator and pay commensurate with the value received.
· If you have an affinity for a specific service provider, subscribe and have a direct relationship, and again pay commensurate with the value received.
· The price with an aggregator may be a bit higher to reflect that service, or not --their share of the value surplus may be paid by the service provider, as a marketing cost. As suggested above the balance here may vary with the context.
· Either way you avoid “subscription hell” because you are not paying $5 or $10/month for all you can eat for each service, you only pay for what you do eat. You can subscribe directly to as many low-volume publishers as you like, because you do not pay for all you can eat, only for what you do eat.
· Either way you avoid “bundle hell” because you automatically get a fair bundle price, computed after the fact based on what you finally chose to use.
· Either way you can be allowed to make adjustments if items were disappointing, or to pay bonuses to specific creators or to all that you used, to the extent you feel that was warranted.
This can behave far better over a wide range of usage patterns than either conventional micropayments (pay per item) or subscriptions (all you can eat for a flat rate). Whether you use the full form of FairPay with balanced control by both parties, or just the simpler “risk-free” model where the seller unilaterally estimates what my demon would do, the result is similar. It all comes down to the shape of the volume discount curve – how the price changes with volume:
· At low volume, the unit price can start small. It can even start at zero for new users who are “sampling,” but established users may not start at that low a unit price, since they know more, and are at low volume.
· As volume increases, the price can increment at a moderate unit rate that gradually declines at higher volumes.
· As volumes get high, the unit rate can become very small. To eliminate risk of bill shock, there may be a price cap (after which additional units do not increment the price at all).
· To match a “risk-free” subscription with a price to a conventional flat-rate subscription at $10/month, the cap might be higher (maybe $12-15/month), or not, as noted below. Many users will not be charged even $10, but the whales might be charged a bit more to produce that same average revenue per user.
Contrast this to a flat rate, all you can eat subscription:
· At low volume, the unit rate is very high – the full flat rate for one item, or even for no items at all.
· At higher volumes, the unit rate declines asymptotically to zero. That is fair up to a point, but very heavy users get a bargain – an infinite volume discount -- which drives the price up for more typical users.
· And, remember that we can expect many more users to subscribe to the risk free plan, since they have no risk of having wasted their money if they have low usage. That means the cap might not need to be higher than for flat rate AYCE and might even be lower.
And compare it to pay-per-item micropayments:
· At low volume, the unit price is moderate, but high enough to deter many users from sampling. (For example, Blendle charges are typically $.25-.49 for a news article.)
· At increasing volume, the unit price remains at that “moderate” level, but the meter keeps incrementing rapidly at the same unit rate, with no volume discount at all.
· At higher volumes the total price becomes far higher than a flat-rate subscription, sometimes by orders of magnitude – “bill shock.”
Platform and Database opportunities
FairPay seeks to leverage relationships to make commerce more win-win. That makes customer relationship databases a vital tool, both to individual creators/service providers and to aggregators. FairPay’s reliance on adaptive relationships that center on learning about value and fairness in a scalable environment supported by automation entails a non-trivial software requirement to manage these value proposition decision processes in realtime.
While simple steps up the ladder of value using some of the Elements may be easy, full forms with fairness enforced by selective warnings and revocation of FairPay privileges take code. I have described an example of simple rules-based algorithms for that. Estimates that have been supported by third parties suggest such an implementation might require about three person-months each of programmer time and of business analyst time. Eventually, advanced versions with AI/ML might go well beyond that. [Update 12/14/24: Emphasis added, as increasingly relevant -- the customer AI will be able to interact with the vendor AI to make this easy for users.]
For the many small service providers who see FairPay as appealing, the need for such software argues for a SaaS service. That could be added on by existing SaaS providers like Patreon, Substack, Medium, and the like, or could come from new entrants. I believe this presents a huge entrepreneurial opportunity, given the economies of scale and potential network effects in such an offering.
There is also a network effect in the FairPay reputation database that can apply in an aggregation context (subject to suitable privacy controls). The FairPay learning process will develop valuable data on what each customer values and how fair their pricing is – that data is central to deciding how to play the fairness game. And ultimately the converse -- fairness data on which providers are fair in their dealings with customers could also be applied to the consumers’ benefit.
In an aggregator context, this consumer fairness reputation score data is much like a credit rating score. A provider could use fairness scores from a consumer’s prior relationships to decide whether to make FairPay offers to a consumer they do not know, just as businesses use credit scores to determine what credit to offer. Think of FairPay as a process of extending FairPay credit – how much value to provide on credit before seeing how fair the customer will be in paying for that value. Providers with high value services might limit offers to only consumers with high fairness ratings, while a provider seeking wide market distribution might cast a much wider net.
This could be done without divulging these fairness scores to those businesses, by having the aggregator (or some other intermediary) make the determination of who to send offers to -- and to send them on behalf of the offering business -- based on a fairness threshold set by that business. The business would simply hear back from customers who accepted that offer, and only know that their fairness score at least met their threshold. Similar consumer data intermediary models have been proven in practice. (The RxRemedy/HealthScout.com business that I worked for in the late 1990s did this successfully for sensitive personal health data.) [I plan to expand on how such an Offering Interest Agent might work in a future post.]
Value versus privacy?
Consumers are rightly enraged at the abuses of what Shoshana Zuboff has called Surveillance Capitalism. However, there is growing awareness that there are complex issues of shared data as a public good, and whether the important issue is not the collection of data per se, but the control of how it is used.
Here, what matters is that FairPay presents a model of commerce based on cooperation to co-create value. That leads to the idea that, when properly managed, win-win trust and transparency in dialogs about value can be more beneficial to consumers than absolute privacy. It is not a zero-sum question of FairPay dialogs on value versus privacy, but one of negotiating a win-win balance of effective FairPay dialogs with agreed limits on what fairness data is tracked, how it can or cannot be used, and how it can or cannot be shared. [[More on this as it relates to Web Monetization and payments continues in PART 2.]]
