Error 402: Would You Pay A Tenth Of A Penny For This Article?

In our continuing Error 402 series on the monetization of web content, we've been talking a lot about things that haven't worked and a few things that have (kinda?) worked, but not in a particularly appealing way (ads, mainly). We will eventually get to more examples of things that are working - along with some areas that might hopefully work better in the future - but today we have another story of a regularly hyped up idea... that never quite seems to work: micropayments.

Thinking about micropayments for digital content predates the modern internet. Ted Nelson, who envisioned a very different kind of digital hypertext system in the 1960s called Xanadu, had assumed that it would make use of micropayments. But instead we got the World Wide Web and its vision of hypertext and related concepts.

The lack of an actual implementation of the Error 402 Payment Required functionality is often chalked up to the lack of an infrastructure that could handle micropayments. In the early years, this was often blamed on a variety of different technical issues, such as how to manage transaction costs when payments were tiny (envisioned to be less than a penny). With credit cards charging a percentage plus a small fixed fee for each transaction, that just doesn't work for true micropayments.

There were, of course, some attempts to get around this technically, usually through bundling and tokens, with the idea being that even if the transactions for something that costs $0.001 would be too big, if you bundle together a million such transactions by using tokens in the interim, then settle up across the larger bundle, the transaction costs turn out to be lower. Historically, though, those systems were cumbersome and complicated and didn't get much usage.

There have been other attempts to deal with the transaction cost problem, including (notably, and with disclaimers) the Interledger Protocol (the Foundation behind which is sponsoring this series), but there remain questions about whether or not this was really a technological problem blocking adoption or if there were aspects that were even more fundamental.

For example there have been many discussions on the issue of mental transaction costs. Nick Szabo, who has been involved in the digital currency world longer than most people, wrote extensively about the mental transaction cost issue back in the 1990s.

The short version, though, is if you have to pay something for every article you read, you have to have a mental debate over whether or not it's worth paying, and that mental process of considering whether or not something is worth a tenth of a penny (or less or whatever) when you don't know what it is before you've even read it is taxing. And if you have to do that for everything you read, it becomes very taxing, such that people just won't bother.

That doesn't mean it's impossible. As lots of people point out, for most people electric bills effectively work on a kind of micropayment system, where turning on your lights does cost a little bit, but your bill is based on adding up all the little uses of electricity and then presenting you with a bundled up bill. Not that anyone particularly enjoys getting their electricity bills, but at least you can understand how it came about.

That said, electricity is different. For one, you know what you're getting (mostly) when you utilize electricity (your light turns on! your computer boots up! etc). One reason micropayments for content are trickier is you have no idea if what you're going to read will even be worth it.

I've heard of a proposed solution to this: make micropayments reversible for some period of time, such that if you open an article and regret it, you can get your money back with one click. Of course, that would also be open to gaming.

About a decade ago we saw a flurry of attempts to do micropayments in a manner that minimized the mental transaction cost by using a flat-rate system. While there were a number of companies that tried this, the one I was most familiar with (and tried here on Techdirt) was Flattr. The idea with Flattr was that you designated some monthly flat rate that you would pay ($5? $10? whatever) and then as you browsed the web, if you encountered websites that used Flattr and had Flattr buttons on them, you could upvote" that content, and thereby allocate some percentage of your monthly fees to that site.

So, for example, if you visit 100 sites with Flattr enabled, and click the button on each of them, then your $5 would be split with $0.05 going to each site (there were also transaction fees and Flattr's own cut, but you get the idea). The upvoting also (in theory) created a Reddit-like directory of content people were enjoying so much they were giving money.

So this model could reach the level of micropayments, but without the mental transaction cost of having to think through will this particular site be worth any money," because no matter what you're paying your $5/month. The only question is how that $5 gets divvied up.

But, also, it wasn't required, so there was a different kind of mental cost: the cost of remembering to click the Flattr buttons, and the decision about whether you wanted to share some of your monthly donation with that particular site. Again, there are possible ways of dealing with some of this, including not needing to click but just designating the money based on visits, but that then risks the possibility of giving money to a site you hate read."

Of course, there's still a larger and more fundamental problem with micropayments: most of the web is still pretty much free. We can go back to our information wants to be expensive; information wants to be free" discussion. When you have a ton of supply, much of it available for free, it's always going to be difficult (though far from impossible) to convince people to step up and pay. This is why advertising still remains the top method of monetization: the end user doesn't have to pay (other than in annoyance).

And, thus, even if you can solve the technical transaction problems, and avoid the mental transaction problems, is there any way to still make it work? Hopefully, the answer is yes... and we'll keep exploring some of what is working and where it might go as the series continues.