Lately, Scott McCloud (comic artist, smart guy, advocate of online content) has been extolling the virtues of micropayments (1 and 2).
He's gotten into some hot water for it.
Partly, I think it's because he's preaching to the converted. Telling artists why they should want it frustrates them, since they can't have it.
Scott's a blue-sky optimist.

I'm a cynic.

So I'm going to describe some problems with micropayments.
Don't get me wrong. I believe in paying for content. I own 300 CDs; the only mp3 I ever downloaded from Napster was a song from a CD I'd just ordered online.
I believe artists should get paid for their work. It's sad I've spent $3000 on music, yet only at best $300 has reached the artists. They deserve more.
For example, my favorite musician. If I could get it to her, I'd send Lisa Germano $100 this second--on top of the $80 I've already spent on her CDs.
But I just want to make it clear that I'm not somehow innately biased against the goals of micropayments.
Of course, that's the exact kind of situation that donation systems and the Street Performer Protocol were absolutely made for.
I'd send her $1000 if it meant she'd make a new album (e.g. if 20 others did too).

But I'd only do that once in my life, unless I became a lot more wealthy.
I make a living creating retail commercial software that is notoriously pirated if it's not copy-protected. I want to get paid for the content I create, too, so believe me, I'm sympathetic.
And you know what? I hate ads. I don't want the online world to be funded by ads.
So, I'm all for a working micropayment system. I think it would be great.
I worry that there are a lot of obstacles to it happening. I want to tell you the ones I've heard that seem worth discussing, whether I believe them or not.

I'm going to try to restrict my attention to reasonably sane concerns about micropayments. I'm ignoring any issues with micropayments that I think Scott has offered effective arguments against in his articles and in Reinventing Comics.

I'm going to mention barriers to micropayments, but I'm going to try to offer insights into why and how micropayments could overcome these barriers.

By the way, micropayments is an awfully unwieldy term--four syllables, twelve letters--especially when compared to its real-world equivalents, like cash.

Fortunately, it's easy to abbreviate.

  • micro becomes its Greek equivalent, µ
  • substitute pay for payment

Put it together and you have "µpay".

A long-standing net tradition suggests using the most similar roman character to µ which leads us to the term


This gets at the heart of the matter: if you're a user, you pay; and if you're a content creator, the Users pay.

(Hopefully that's a significantly negative enough connotation that no company runs off and trademarks uPay on me.)

And, by some strange coincidence, users are a significant factor in many of the barriers to upay. Without further ado:

1. Users don't want to pay for content

2. Users don't want to deal with upay decision-making

3. If users pay for content, they want to keep that content

4. Getting upay into browsers is hard

5. Critical mass chicken & egg problem

6. Technological improvements could remove the success tax before upay happens

7. upay requires site authors become a business

8. upay introduces middlemen

9. Middlemen could get monopolies

I think most of these barriers have responses in favor of upay (although some are things we can't know either way); but there's enough things that can go wrong here that even if you'd like to see upay happen (and I would), unless you're quite the optimist, it's hard to see how it ever could.

Let's look at them in detail.

1. Users don't want to pay for content

As I've described above, I'm very happy to pay for content. The question is whether everyone else is.

Let's get one thing straight first. People do pay for content. Some people claim that people don't generally pay for content, that advertising is the norm.

That claim is easily invalidated, by examining all the communications media and distribution channels. (I'll omit online distribution, since we don't know how to make those work financially, except email and Usenet.)

Here are the communications media and content that are distributed purely based on advertising:


  • TV shows on TV
  • music on radio
Some distribution systems rely on both advertising and user payment; the details vary, but let's assume that advertising really is the dominant income source:


  • magazines
  • newspapers
And some distribution media don't make use of advertisements at all, except to the degree that advertisements are creeping into everything:


  • movies in theaters and DVD/videotape and premium cable
  • music on CD, tape, vinyl
  • live music
  • books
  • TV shows made for cable on premium cable
  • TV shows on DVD/videotape
  • movies made for cable or direct-to-video on premium cable
  • computer games

[I've kept comic books off the list because I don't know how significant the advertising in them is for their income. Perhaps I should break down the kinds of content that appears in newspapers, so that comic strips will appear directly. I've omitted public television, since I'm not sure how much money they get from each of their sources.]

But note that the music industry doesn't make any money from advertisement-supported radio; in fact, record companies have to pay money to get their music on the air, so the source of money for the industry is not radio, and is not advertising. (Musicians themselves do get money for airplay, but only through a statistical system that ignores infrequently-played artists--a system that people don't seem very happy with.)

If you accept that argument, then music, movies, and computer games are never funded by advertisement. The only really non-text content that is advertising-funded would be syndicated comic strips and TV shows.

So people will pay for content, especially non-least-common-demoninator content.

Another argument made for "users won't pay" is that users won't pay an Internet access fee and also pay for content. Certainly users will resist; but most people buy cars, pay for gas, and pay taxes to have their roads maintained, and pay a toll to use certain roads and bridges. Personally, I pay a surcharge for local calls outside a certain radius.

People will pay if they have to.

That's what it comes down to.


It's not true for me, but it might well be true: if there is free content of sufficient quality, people may not be willing to even upay for slightly higher-quality content.

[Personally, I'd rather pay a buck a month for new Leisuretown content rather than read Dilbert for free--but there are only a few artists who I'd pay that much for, since a buck-a-month adds up quickly. (Two cents a month is way more tolerable.) Then again, I'm not as big a comics fan as I am a music fan.]

If that's true, then there's a problem: we need to get rid of all of the free content at the same time we install a upay system.

That doesn't sound easy to me.

If you're an optimist, you may figure people won't care if one thing costs a tenth of a cent, and another is free, assuming they're both equally easy to get to. And maybe you're right.

Or maybe that will be the catalyst for the success of upay: when all the free content dies off due to "the success tax". [The success tax is the increased server bills an author must pay if her content is successful. The closest I've ever come to paying it was a $50 bill one month when my site was slashdotted and 2500 people viewed a 150K "web comic".]


Would you gamble tens of millions of dollars on it, though? 'Cause some venture capitalist may have to, to get it going.

2. Users don't want to deal with upay decision-making

Some people have advanced the argument that dealing with upay will be too hard for users. This comes in two forms: user interface and cognitive workload. It's clear we have to make the user interface good; problems with how we get there are considered in point #4.

The cognitive workload argument basically says that making people decide whether they want to pay their money for a given piece of content (especially an unknown piece of content) is going to require mental effort that will scare people away.

This might be true on the surface. Personally, I value my time a lot and already find myself making that same kind of judgement call all the time, especially when I find myself writing essay-length articles like this one on some obscure web forum somewhere (or worse yet, my personal site and its ten uniques). But most people do seem to value their money more than their time.

I think this isn't really going to be a problem if upay is done right: as a tool to avoid the success tax, but not as the only or primary payment tool.

For example, consider this scenario (and this is only one possibility):

  • some content is free or advertising supported (say,
  • things I know I like, I have a monthly or yearly subscription to
  • sites I'm not subscribed to have a minimum of free content, which is updated rarely if at all (so I can't keep going back and getting more content for free)
  • all content on a subscription site is available using micropayments

I think it can work, if we can get the user interface right. But. Well. I'll get to that.

3. If users pay for content, they want to keep that content

If you buy a book, you can keep that book until it falls apart. You can even sell it later.

If you buy a videotape, you can keep that videotape until it falls apart. You can even sell it later.

If you buy a CD, you can keep that CD until it falls apart. You can even sell it later.

If you buy some online content...

A counter-example that comes to mind is movies. If you buy a ticket to a movie, two hours later, all you have left is your memory of the movie and maybe a pressing need for a toothpick.

But the movie theater is more than just the movie.

If online content I upay for gives me THX surround sound from my laptop speakers and air conditions my apartment, maybe I wouldn't worry about this barrier.

Until that happens, I think we have to expect users will resist paying for content they can't keep. Heck, they even get to keep TV shows they videotape, despite not even paying for those. (Courts ruled time-shifting legal; I dunno the legal status of watching a show multiple times, but clearly at this point it's happening--it's what users are used to.)

I see three possible solutions to this problem.

I. If you want repeated access, you the user should subscribe... or suck up the upay if you decide to go re-read that one particular comic again.

I don't like this much. Subscription is ok if I'm already subscribed, but there are real problems with having to either subscribe or make micropayments for something I'm not currently subscribed to. Suppose I want to send a link to a particular comic to someone, but I didn't save the URL. How do I skim through looking for that one particular comic without paying for all the others?

It's probably worth adding this to our scenario, though:

  • subscriptions to a site are automatically discounted by however much you've spent on micropayments there

II. upay services track what's been paid for and make repeated accesses free.

I think that's going to be way too much data to track and log for a long period of time (e.g. a year). The web server could do it instead, spreading the workload. But even with Moore's law and the rapid increase in hard drive storage capabilities, this sounds unpleasant to me. upay services are going to have narrow enough margins as it is to pay the price of customer support this would surely lead to.

We'd also have to be careful about moving pages around. If I access the front page of a comic strip site and read today's comic, I shouldn't keep getting free access to that page over and over as each new day's comic is added. I should get free access to that one archived comic.

III. Browsers permanently download content that's been paid for.

I think this is the best possible approach; the storage is paid for by the user, the content is right there with no latency for the user, and the content is kept permanently.

But if you buy the argument that this is a barrier, and the other two approaches above aren't acceptable--if this really is necessary--then this adds yet another feature we need browsers to have before upay can succeed.

  • browsers permanently store content that was paid for with micropayments

[It also potentially introduces the rather odd experience that the upay content is permanently downloaded, but the subscription content is not, if subscription content continues to work with the existing mechanism.]

4. Getting upay into browsers is hard

And personally, I think that's the biggest barrier of all to upay. I think upay would work well with a really tricked-out browser; but I don't see how we're ever going to get there. It's not a programming problem--I'm a programmer, I know these things can be programmed--it's a problem of getting it into people's hands.

Let's look at a slightly expanded version of the proposal in Scott McCloud's Reinventing Comics:

  • links to content of cost over $0.0X bring up a dialogue asking if I really want to do this
  • links under $0.0X are specially hilighted so you know clicking on them will charge you money; possibly color-coded into two or three ranges
  • there's a running tally shown so I can see how much I've been spending on micropayments today, this week, this month

Unfortunately, this level of support is not something we can easily get with a plugin. Remember, we want all net-accessible content to be this way. I suppose if, as a start, all upay content had to appear inside, say, a Flash window, we could survive, but it likely makes the browsing experience less pleasant--which is one of the barriers we're trying to overcome in the first place. [Nor can we use something like Java. If it could be done in Java, a site author could then write Java code that makes you pay for things unintentionally, too.]

If you take my word on that--I'm a programmer, so I think I'm right, but I'm not a browser or plug-in programmer, so I suppose I could be wrong--then we have a pretty hard problem.

We need to get people using a new browser.

That's a pretty huge barrier to overcome. I have trouble seeing it happening, unless it just came bundled in with some new browser everybody was getting anyway. The only case where that seems plausible frightens the heck out of me.

If it's a new browser, somebody needs to write this new browser.

That's a significant barrier, too. Somebody could write a new browser, but the creeping featurism of the HTML specification makes writing a new, fully-functional browser challenging. Somebody could take the Netscape source code, Mozilla, and add this feature, but Mozilla is a big, complex beast, and it would take a while; and widespread opinion is that Netscape 6 is inferior to Internet Explorer.

Or somebody could modify IE. But only one company can do that. See barrier #9.

5. Critical mass chicken & egg problem

upay suffers from severe chicken & egg problems. Until there is enough content out there for users to upay for, companies trying to bundle together user transactions to reduce overhead are stuck. Until they get enough users with the technology, content creators are stuck.

The former problem... well, I guess that's what venture capital is for.

We could try to overcome the latter problem by making sure that users who aren't upay-enabled can still access the site:

  • all content that can be micropaid for is also available with ads or some other per-access support fee (not subscriptions)

But that begs the question of how that content is paid for, since ad revenue is failing. Basically, that requires some solution other than upay which works; and if we could find that, why bother with upay?

About the only thing we can realistically do is leave that content available through low-cost subscriptions, possibly through aggregated sites.

  • all content that can be micropaid for is also available through low-cost subscriptions

But I think that's a kiss of death since nobody can get at the non-free content cheaply. So instead we need to somehow magically get upay-capable browsers into the hands of all users, and then authors can switch over to upaid content.

And I guess that could happen. Eventually we'll get to #9, too.

6. Technological improvements could remove the success tax before upay happens

One theory advanced by Scott and by's Phillip is that upay is probably going to happen eventually. Enough banking is moving online; more bandwidth is becoming available; costs are going down; so upay is just an eventual plateau along that evolutionary curve.

Maybe; I'm not sure how the browser deployment would happen evolutionarily. But someone made a plausible argument to me that if upay is dependent on certain technological innovations, then it may never be needed, because those technological innovations directly solve the problems.

For example, one concern with upay is that the need to communicate with a third-party to handle the payment introduces a time delay. Increasing network bandwidth at servers and the major net pipelines may reduce this latency, but that same increased bandwidth will make bandwidth cheaper, possibly moving the "success tax" down to a level where it's not a problem.

Another thing advocates of upay discuss is the success of peer-to-peer distribution models. But those distribution models obviate the success tax!

Napster and Gnutella are famous file-sharing systems which often end up with heavily-desired data on multiple "servers" (the peers who are file-sharing), so that a file which is popular will be downloaded from many different sites, and no one site gets a heavy server bill.

An industrial-strength version of this is under development, called Freenet; with Freenet, you "inject" a document onto the Freenet, even anonymously, and the document is automatically redistributed across the Freenet in relation to how many people request it. Freenet, however, is being written by and for users, as opposed to by and for ISPs (Internet Service Providers). As a result, Freenet does not leverage the net efficiently, and will probably always be slow and risky (you may not be able to find a document on the Freenet even if it's there).

Instead, imagine something like this being done at the ISP level.

A webhost might be connected to two different "upstream providers". Any requests for a file will come through one of those two providers. If those providers stored a copy of the file after the first time it was requested (putting it in a "cache"), the server would only need to distribute the file two times, instead of, say, twenty thousand times, and the costs of bandwidth from the server would go down.

Now, each of the upstream providers would still see ten thousand requests each, and they'd have to serve them themselves. That might even be more expensive for each of the providers than just passing the data through--but if you imagine that each of the four machines they connect to do the same thing, suddenly everyone's costs go down, except they need to sit around storing a lot of data. But hard drives are a lot cheaper than T3 lines.

The protocol used on the web is designed to allow caches, and to handle things like "what happens if somebody updates their page, but a server somewhere is keeping the old copy". Unfortunately, all of this stuff is optional, and it's hard to get all ISPs everywhere to support it. One approach would be to write a new protocol--which is the Gnutella/Freenet approach.

Better still would be to get a new protocol where this distribution is mandatory and is supported by ISPs. I don't think anybody is trying to go this route, but it's a proven successful model: Usenet uses a distribution model a lot like the above--but all information is always distributed to all sites (which requires a lot of hard drive space and tends to result in data disappearing within weeks after being added when the hard drive fills). Before the web was a success, Usenet was the closest thing to the web; it was successful, users communicated with other users, and some people had audiences of thousands or even tens of thousands of readers--at no extra cost to the widely-read writer.

Any of the above models could remove the success tax. Now, the micropayment model is good for more than just the success tax, but I think a lot of people just want to see the net get to the point where content creators can get stuff up at no cost to themselves. And with no success tax, any money from ads or donations is profit.

But people like Scott are arguing that overcoming the success tax isn't the only goal of micropayments--indeed, the last I Can't Stop Thinking used a $0.25/month example for a situation with an apparent success tax of $0.02/month/user.

While I think a "no success tax plus donations" model could be reasonable, I have no problem with the idea that some people would like to make a living posting content to the net, and, indeed, I'd be happy to see anyone I've linked to here succeed (and have already contributed my share), so I'm willing to grant that overcoming the success tax isn't enough. Whether all users feel that way... well, see barrier #1.

7. upay requires site authors become a business

This is an interesting point raised to me by an acquaintance, and I'm not sure if it's true or not. If you just want to use upay to avoid the success tax, and not actually turn a profit--you just want to run your website as a hobby and avoid huge server bills if it becomes unexpectedly popular--you're still collecting money for providing a service, and you probably have legal obligations to operate as a business--taxes, etc.

I can see a few ways around it. As I understand it, if you earn royalties (say, from a book), that income is taxable, but it doesn't obligate you to act like a business. So it's possible that we could structure the arrangement so that this sort of income isn't viewed as "businesslike".

For example, one could use an aggregating partner, a business which hosts a bunch of web content and manages all the micropayments. They charge enough to turn a small profit, and provide authors with free webhosting. But this introduces a middleman; see the next barrier.

My guess is that web-provider micropayment processing is the way to go; if they can give "royalties" to content providers, great, if not, then a successful author might need to become a business if she wants to start turning a profit.

Let's add that to the scenario.

  • free webhosts use micropayments to offset hosting costs

That sounds more negative than I mean it to be; free webhosts these days conjures up a sort of gutter mentality, because they have so many intrusive banner ads. You have to imagine the free webhost in this scenario as a class act. (Assuming users don't consider micropayments to be equivalent to an intrusive banner ad, since the money is going to the hosting site.)

8. upay introduces middlemen

One of things that excited Scott McCloud in Reinventing Comics about digital delivery was that it removed the middlemen. (See pages 182-183.) For comic books, the old system makes content flow through this path:

  • author
  • publisher
  • printer
  • distributer
  • retailer
  • consumer

Meanwhile, the money flows in the opposite direction, and gets significantly reduced at each step.

With the alternative, digital delivery, the content flows through this path:

  • author
  • net connenction
  • web host
  • Internet
  • consumer's ISP
  • consumer

but the money goes direct in the other direction.

This is literally true--the money can go straight through, and none of the intermediaries knows about it or can touch it (since it's encrypted).

But the intermediaries all get paid.

A lot of it happens behind the scenes, but clearly the consumer pays for his ISP connection, and the author pays for her ISP connection and for her webhosting.

The good thing is that those costs are not in any way in proportion to the amount of money coming through from consumer to author--the author could turn a big profit.

The bad thing is that those costs are not in any way in proportion to the amount of money coming through from consumer to author--the author can lose big. That's the success tax. There's flat fees for the net connections, but the big issue is that the web hosting charges are in proportion to the amount of content accessed for the web hosting, instead of the money coming to the author.

So upay offers a way to make sure that the author gets back money in proportion to the amount of content accessed through the web hosting.

But unfortunately, that's not much different than having the money flow back through directly.

Indeed, the vision described in this essay posits at least two middlemen: the micropayment services (third parties along the lines of PayPal or VISA) and the web hosts. In my worst case scenario throw in the makers of the browser.

Each of those parties knows not just about the access to the content, but they have access to the dollar amount of the transaction, and are in a business position to demand a cut.

Forunately, their costs are lower, so the cuts should be lower, but those are smaller cuts from a much smaller initial pie--since we're assuming users aren't going to be willing to spend that much.

There's even some danger of editorial influence if things go the webhosting route; a web host could dictate what amount the author could charge for a page. Ideally, though, a web host would be an aggregator who just lets anyone do whatever they want, much like the way cafepress lets sellers choose their profits.

9. Middlemen could get monopolies

There's some real danger of monopolies here. To be successful, these things have to be widespread. To get them widespread, somebody has to move first.

The biggest monopoly danger comes from the browser, because that's hard to get into people's hands. There's not much anyone can do to prevent new hosts from aggregating content and making micropayments; there's not much anyone can do to keep a new micropayment third party from trying to enter the show. Or rather, not much anyone other than the browser vendor can do.

Given that µSoft has made it clear in the past that they'd love to be an online toll-collector, making money off of every transaction, putting a high-quality upay user interface into Internet Explorer frightens me.

Additionally, a company with a monopoly on the upay browser could leverage that into monopolies on micropayment services and web hosting, by favoring or discounting micropayments to particular providers and services. They probably couldn't do it legally, but that doesn't mean it wouldn't happen. If it happened, charges would be inflated, and nobody would make any money.

Here's another way to get a monopoly.

Let's imagine an alternative to micropayments. Suppose we try just assessing a flat fee to all users and distributing that money to content providers based on how often their site is accessed. To reduce the measurement cost, we could use random sampling. Sites with only ten or twenty accesses would get missed and miss their share entirely, but hey. Users who never access content that gets sampled--users who are only fans of obscure work--have their flat fee distributed to the big success instead, but hey, the obscure work doesn't have a significant success tax problem.

The above is exactly how ASCAP and BMI already work right now, handling performance copyright for music. And, oh my no, they're not abusive with their monopoly powers or anything, goodness no. Clubs that host local bands that play all-original music love paying money that all goes into the hands of the music industry and top-40 artists. Yessir.

About the only monopolies I'm comfortable with are the Library of Congress and the Post Office. I was never comfortable with Deja News archiving Usenet for profit; it's still unclear how Google is going to be successful at it; if they can make it work without ruining the service, more power to them.

No wait. Not everybody sits that far on the left.

That's a political barrier nobody's going to overcome anytime soon.
Maybe the thing to make this all work isn't to leave it in the hands of market forces.
Maybe the scenario I've described is achievable.
  • some content is free or advertising supported
  • I can get a monthly or yearly subscription to sites
  • sites have a minimum of free content which is updated rarely
  • all content on a subscription site is available using micropayments
  • subscriptions to a site are discounted by however much you've spent on micropayments there
  • browsers permanently store content that was paid for with micropayments
  • links to content of cost over $0.0X bring up a dialogue asking if I really want to do this
  • links under $0.0X are specially hilighted so you know clicking on them will charge you money; possibly color-coded into two or three ranges
  • there's a running tally shown so I can see how much I've been spending on micropayments today, this week, this month
  • free webhosts use micropayments to offset hosting costs
When people propose things like this, they're supposed to say "is that so much to ask for?"
But it seems to me it is a lot to ask for.
And I have trouble seeing how it won't be at least that complex to overcome the barriers...
...because users might not pay for content if there's free content...
...because users won't want to deal with µmanaging their µpayment decisions...
...because users will only want to pay once for each thing they access...
...because getting a good interface into users' hands will be hard...
...because a lot of people have to participate before it will work, and it could be to a user's financial advantage not to participate...
...because the technological improvements needed might make micropayments irrelevant anyway...
...because micropayments might require hobbyists become businesses...
...because the micropayment middlemen might not be that big an improvement over old-style publishing...
...and those middlemen might become monopolies, if they aren't already.
We need some way to get money into the hands of the people who create.
But we need something that will work. And I doubt that upay will. But will I be happy if micropayments succeed?
Is the sky blue?

written 2001-06-30
sean at nothings dot org