The New Case for Micropayments

A Micropayments Platform Based On Clay Shirky's Ideas,
Among Others

Marc Glasberg - September 20, 2009.

What special conditions are necessary in order to successfully sell content over the Internet? How can a micropayment system reduce mental transaction costs?

"So far micropayments people have been mostly ignoring this crucial user interface / mental transaction cost problem and (because they've been ignoring it, and because it's usually a very hard problem) generally have not solved that problem. Thus micropayments have failed."

Nick Szabo, 2007 (See "Micropayments redux", Comments)

A quick recap

In 1998 Jakob Nielsen wrote a classical paper called "A Case For Micropayments". Two years later Clay Shirky wrote a response called "A Case Against Micropayments" arguing that micropayments would never work. Shirky's main idea was the following:

"Users are the ones with the money, and micropayments do not take user preferences into account. (...) There is a certain amount of anxiety involved in any decision to buy, no matter how small, and it derives not from the interface used or the time required, but from the very act of deciding." source

In other words, Shirky argued that the very act of deciding about paying involves some mental effort, no matter how small the payment is, and that this effort is essentially a cognitive cost users pay in addition to money. This idea was first developed by computer scientist Nick Szabo, who coined it "Mental Transaction Costs", or MTC.

Shirky used the MTC idea to completely dismiss the practicality of micropayments. A heated debate continued over the next decade, and the unfortunate outcome is that until now there wasn't a single micropayment company that addressed the mental transaction costs problem seriously, as Nick Szabo himself put it (see the citation that opens this article). While I agree with Shirky that there is no way of completely avoiding mental transaction costs when making payments, we have found many ways to reduce it. For micropayments to become practical it is not at all necessary to completely eliminate the MTC, if you can reduce it and apply some of the techniques I will list below.

Mental Transaction Costs are relative, not absolute

Usually the MTC of some payment transaction is not evaluated by the user in absolute terms, but against the possible alternatives. In other words, it's an opportunity cost!

Suppose users go to an online newspaper to read an article, and suddenly are asked to pay or to fill out a form. There is a MTC involved in paying or filling out a form, but there is also a MTC involved in searching the same news somewhere else. If what users are trying to read is general news, they know it's not difficult to find it somewhere else, so the alternative MTC is low. On the contrary, if what users are trying to read is more specialized, like a very specific book review, then the prospect of having to search it somewhere else may not present a low MTC. Note, however, that presented with no low MTC option users can simply decide to forgo both options, and simply give up reading about that subject and search for something else, but take into consideration that there is also MTC involved in giving up something you want. Giving up is still a choice, albeit unconscious, between the different MTC incurred.

Note we come up with an interesting conclusion here, that instead of simply reducing the MTC for your paid content you may also make sure that all other alternatives have higher MTC. This way, you can make the MTC work for you, not against you. You can achieve this by building a Virtual Perimeter around your online business.

How Apple has created a Virtual Perimeter

After Apple's huge success with iTunes micropayments for music, Clay Shirky's opponents were quick to point out that he was wrong about micropayments being impossible. Shirky should have known, of course, that it is always a risky proposition saying that something is impossible, since opponents need only to find a single working case, and they did in fact find two: iTunes and Internet gaming. I'll talk about games soon, but first let me concentrate on iTunes.

Skirky's response to iTunes was that, ok, iTunes is a good micropayment example, but it only works because iTunes is a monopoly created by music labels, that could not be replicated by others:

"Fans of the iTunes model are right to point out that people use it because they find it more convenient, but they overlook the legal and regulatory hurdles put in place precisely to make other models less convenient (especially for law-abiding citizens.) (...) Newspapers, even if every single one of them acted in collusion, cannot establish a monopoly on news." source

It is indeed important to analyze why iTunes is more convenient than downloading music through file-sharing software, but it's not at all for the reason Shirky points out. Basically Shirky is saying that users are buying from iTunes because they are forced to, they have no other option. This is simply not true. Monopoly by one definition is the market condition that exists when there is only one seller. I can hardly see how a company can be called a monopolist if you can get the same or almost the same product for free from your home computer. In addition, I have talked to dozens of friends who regularly buy from iTunes and not a single one told me they use the service because they think they would not be law-abiding citizens if they downloaded it for free.

If the overwhelming majority of Internet users don't think it is wrong downloading music for free, how is iTunes selling content hand over fist? The answer is that Apple has created a virtual perimeter around its business. This perimeter tells us there is an inside and there is an outside, and that inside is better. What does that mean in iTunes context? Users already have iPods and iPhones. iPods connect directly with iTunes music player. iTunes music player is also itself the browser for iTunes Music Store, and for the App Store that sells iPhone applications. iPhone only lets users upload applications from App Store, and so on. The result is that once you are inside Apple's virtual perimeter, it's not at all the same to get music from some other music store.

Shirky is right when he says that people find iTunes more convenient because there are hurdles put in place precisely to make other models less convenient, but the hurdle is not legal or regulatory. The hurdle is that Apple has created a significant MTC for you to acquire music outside of Apple's virtual perimeter, once you have an iPod.

Let's talk about online gaming to see how they have also created a virtual perimeter. I'll let Shirky explain this to you:

"(...) Cyworld, a wildly popular online forum in Korea, is able to collect small payments for digital items, denominated in a currency called Dotori, because once a user is in Cyworld, SK Telecom, the corporate parent, controls all the distribution options. A Cyworld user who wants a certain kind of digital decoration for their online presence has to buy it through Cyworld if they want it; the monopoly within the environment is enough to prevent competition for pricing of digital goods."

Micropayments for games don't work only in Cyworld, but in gaming websites all over the Internet. Shirky talks about the "monopoly within the environment" and this is almost what I'm referring to as the virtual perimeter. This is not the same as a real monopoly because there are thousands of gaming websites you can play in, but once you are addicted to a specific game, it's not at all the same to get your digital items from another gaming website. The result is that, once you are hooked, the site has created a significant MTC for you to go play a similar game somewhere else.

Bear with me for 2 more examples. Facebook and Evernote.

Suppose you are a Facebook user and you want to send Facebook virtual gifts to your friends, but you are required to pay for the gifts. It doesn't matter if MySpace decides to make the same kind of virtual gifts available for free on their network, because once you are a Facebook user it's not at all the same to send gifts from MySpace or other social networks. There is a virtual perimeter around Facebook, so if you "live" in Facebook you cannot simply switch to MySpace to be able to use the free gifts without incurring high mental transaction costs: Your friends are not there, you haven't invested in building your profile on MySpace, and so on.

In the same way, if you are an application developer inside of Facebook, then MySpace applications will have a hard time competing with you. Of course, you will still compete with other applications inside Facebook, so your application should also try to create its own virtual perimeter; a virtual perimeter inside another.

Of course, all this works the other way around if you are a MySpace user.

Now about the Evernote company. Go read this.

Evernote studied the behavior of the earliest adopters and found that the longer customers used the service, the more likely they were to start paying for premium services, because the data is more valuable to the customer as it becomes larger. In other words, for each customer a virtual perimeter is slowly created until it's not at all the same to use the same premium functionality from a similar website. Once this occurs Evernote can really start charging.

How to create a virtual perimeter

To create a virtual perimeter you must do 2 things:

Rule 1 - Create a website that users prefer, all things considered.

Why: This is, of course, simple and intuitive. If users don't prefer your website they will simply switch to the competitor's. And when I say "all things considered" I mean price, so having lots of amazing free content is (almost always) necessary to achieve this, as I will discuss later.

Rule 2 - Make sure that the premium features you charge for are not independent from the rest of your website content.

Why: As all competing sites are just one click away, if your website offers some very interesting but independent feature, users can easily get this feature somewhere else, and still continue to use your website for the rest. Thus, you need to make sure that the paid premium features improve the use of the free features on your site. And, if possible, also make sure that paid premium features cannot be replicated by your competitors unless they also replicate the free features of your site. These types of premium features are the only ones you may really charge for.

If you go back to the previous examples - iTunes, online gaming, Facebook and Evernote - you will see that each one of them applies these two rules, and thus creates a virtual perimeter.

Those who try selling content without creating a virtual perimeter will have a hard time. Consider, for instance, that online newspapers have a lot of content but, as Nicholas Carr pointed out, each piece of content stands on its own. Suppose a Newspaper A offers the best political content for free and charges only for the sports section. Users can still read all free articles and get the sports section from other online newspapers. The news service own nature is that it is possible to read about sports without reading about politics, and worse, reading about sports doesn't improve user experience when reading about politics. In other words, we have a news website here, but no virtual perimeter around it.

After reading some Nicholas Carr and David Carr articles about the online newspapers business I came up with interesting ideas about how to create a virtual perimeter for online newspapers, and I will soon detail them in a follow up article which will include a live demo.

What a Virtual Perimeter is not

A Virtual Perimeter is not a walled garden, is not a payment barrier, and is not a closed system where interconnection with the rest of the Internet is prohibited. Think about the previous examples and this will become clear. Facebook even has an API to let others create applications that work inside of Facebook.

A Virtual Perimeter is just a very good way to protect some of your website features from the profit margin destruction created by the Internets' wonderful efficiencies of scale and distribution. A Virtual Perimeter is an Internet "hack".

Freemium is necessary to create Virtual Perimeters

The Freemium business model was first described by venture capitalist Fred Wilson. It's very simple: you offer a lot of content for free, and charge only for advanced, premium features. I will now discuss why Freemium is necessary for you to achieve Rule 1 of creating a virtual perimeter for your website.

Suppose you are the CEO of a gaming website that offers a game composed of 2 features, all for free:


Let's also suppose your competitor offers almost the same, but their game is not as good as yours:

Since your game is better than your competitor's, you will gain a user base more easily than they will. Despite what seems like an impenetrable quality barrier, if you start charging for your game you will quickly realize you lost most of your users to your competitor. This is because, as Josh Kopelman puts it, there is a market for free and there is a total different market for paid. When prices reach zero you lower the mental transaction costs and demand soars. See the graphic below and note how the number of users jumps when you reach zero price:

Remember Rule 1 for creating virtual perimeters: it says that you must make sure users prefer your website over the competitors, all things considered, including price. When users considered both features A and B and price was free for both websites, they preferred yours. When users considered both features A and B but your game is paid most preferred the competitor's.

Let's see how Freemium changes that. Instead of charging for features A and B, give A for free and completely remove feature B. Now users will compare your great feature A with the not so great features A and B of your competitor. They will likely prefer your website as feature A is what really counts and you will regain all users. Now, if you start charging for the optional feature B you will not lose your customers. You will maintain your user base, and will be able to show ads to them and monetize with Google AdWords or the like, in combination with charging for feature B.

But what about feature B itself? Even if your whole user base won't decrease anymore, will the user base of paid feature B still decrease like shown in the graphic above? Sure, but the decrease won't be as sharp. See below:

Since a game's digital items (feature B) cannot be used separately from the game itself (feature A), users cannot get your competitor's feature B for free unless they stop using your game. But, as discussed, they won't do this because they prefer your game. Thus, your feature B is not competing with your competitor's feature B. Your feature B is now only competing with the alternative of not using feature B at all. In other words, once users have decided to play your game and not another, it's not at all the same to get digital items from some other game that won't work in yours.

Please note, this game example above makes sure we are creating a virtual perimeter not only by following the Freemium business model with amazing content (Rule 1), but also by making sure that the premium features are not independent from the rest of your website content (Rule 2). See how easy it is to destroy the virtual perimeter by breaking Rule 2: Suppose feature B was not the digital items to use within the game but, for example, the ability to read about tricks for winning action games. Users can read about these tricks on other websites for free, while still continue playing your game just the same. This new feature B, although still interesting to users, is now independent from Feature A. The virtual perimeter is then gone, although the Freemium is still present.


The previous discussion about Virtual Perimeters explains how your website can overcome the MTC created by selling paid content, by making sure your users would have an even bigger MTC if they switched to a competing website. This doesn't mean you shouldn't also try to reduce your own MTC. In other words, there are 2 strategies you must implement: you must decrease your own MTC and at the same time increase the MTC of the alternatives. Both are imperative.

Whether you succeed in implementing a Virtual Perimeter or not, depends solely on your website's strategy, and is not dependent on the payment platform you use. However, reducing your own MTC is very dependent on the payment platform you use and other important details. To start reducing your MTC you first need to understand aggregation.

Contrary to micropayments (ranging from 1 cent to 5 dollars), regular payments are used for traditional subscription models where users pay for a long period of time, say a whole year, and get access to the entire content of a website. For example, Flickr charges you $24.95 for 1 year of "pro account". Traditional subscriptions have a single advantage over micropayments: websites get a big upfront payment. This is very nice, of course, but traditional subscriptions also suffer from some serious drawbacks. As Jakob Nielsen put it in 1998:

"The main problem with subscription fees is that they provide a single choice: between paying nothing (thus getting nothing) and paying a large fee (thus getting everything). Faced with this decision, most users will chose to pay nothing and will go to other sites. It is rare that you will know in advance that you will use a site enough to justify a large fee and the time to register. Thus, most people will only subscribe to very few sites." source

Thus, selling content on the Internet has been hard because both paid models - micropayments and traditional subscriptions - have problems. Both create MTC for different reasons. While for micropayments the payment decision itself creates a MTC that is big in comparison to the content's price, for traditional subscription the bigger financial investment, time to register and other considerations will also create a great MCT.

However, consider that this division between micropayments and traditional subscriptions is somewhat arbitrary. When you come to think of it, there is in fact a continuum between them, where micropayments and traditional subscriptions are just the extreme cases. Wouldn't it be great if we could find something in the middle that encapsulated the advantages of both micropayments and traditional subscriptions without their disadvantages? Yes, it would, and here it is: micro-subscriptions.

Consider, for instance, regular newspaper distribution model. Clay Shirky makes a convincing argument that evaluating the correct price of individual articles in a newspaper may be more difficult than evaluating the whole newspaper:

"Users have no trouble deciding whether a $1 newspaper is worthwhile - did it interest you, did it keep you from getting bored, did reading it let you sound up to date - but how could you decide whether each part of the newspaper is worth a penny? Was each of 100 individual stories in the newspaper worth a penny, even though you didn't read all of them? Was each of the 25 stories you read worth 4 cents apiece? If you read a story halfway through, was it worth half what a full story was worth? And so on. When you disaggregate a newspaper, it becomes harder to value, not easier. By accepting that different people will find different things interesting, and by rolling all of those things together, a newspaper achieves what micropayments cannot: clarity in pricing." source

However, when Shirky compares a single newspaper article with a single newspaper he is not comparing a micropayment with a traditional subscription. He is comparing a micropayment with a micro-subscription. The newspaper cost is still in the micropayments range, below 5 dollars, and readers buy them a la carte. A traditional subscription, on the contrary, is when readers subscribe to an entire newspaper for a whole year. See the table below:

Micropayment Micro-subscription Traditional Subscription
1 newspaper article 1 newspaper 365 newspapers


Notice how the newspaper micro-subscription option here enjoys the advantages of both other options, and none of their disadvantages. On the contrary to traditional newspaper subscription, readers buy their newspaper micro-subscription when they feel like it, without having to fill out any forms, and they don't need to know in advance if they want to read this particular newspaper enough to justify a large fee and the time to register. And on the contrary to evaluating one single newspaper article, readers have no problem evaluating one single newspaper. They are used to it, and just expect it to be more or less what they usually get. MTC is low.

Of course, this definition of a newspaper micro-subscription is still arbitrary. We could define a micro-subscription to be the first 3 pages of the sports section, or a month's worth of newspapers, but these aggregations probably don't make much sense, commercially speaking. We can therefore avoid the pointless discussion about how big or how small an aggregation constitutes a micro-subscription by defining it in the following way:

"A micro-subscription is an aggregation big enough to make sense to users, while still in the micropayments range of price"


The micro-subscription advantages can be replicated for the Internet. For example, users can be given a 10 hour access to the entire New York Times Online for 20 cents. Users would do this with a single click in a micro-subscription button like this one: without having to create any logins or passwords, without having to provide any credit cards, and without having to type anything to sign in. Since 10 hours is much more than users need to read the newspapers for now, they wouldn't have to think if this amount of time is enough. But since this is not a long period of time, users wouldn't have to remember or care about this purchase any more than here and now. Still, the system should unobtrusively display the time the user has left and visually display that the user has access granted. Of course, a virtual perimeter should also be created for the New York Times website as well, if this is to work (as I said I will present a live demo for a virtual perimeter for newspapers in my next post). In other words, users would be able to pay as they browse, without disrupting their user experience.


I have detailed under which special conditions content can be sold over the Internet:

To that goal we have created the Payment Platform so that websites can unleash the power of existing payment systems like PayPal and others, according to the principles here discussed. As to my knowledge, we've been the only company so far to have taken these issues seriously and, for this reason, developed a payment platform unique in many respects. Please, check our live demos, and see how it works.

The ideas here presented, and for most part itself, could not have existed without the writings and of Nick Szabo, Clay Shirky, Jakob Nielsen, Fred Wilson, Josh Kopelman, Nicholas Carr and David Carr, and whose applicable essays have been linked from the iCents discussion page.

blog comments powered by Disqus