Stripe, Balanced, WePay for Payments: Not big enough to #fail

When Amazon Payments first told us they wouldn't allow our crowd-funding business to continue using their system, we thought it was a phishing attack. That's because a robot sent it, and what does a robot know? But it turned out they were serious, they had to cut us off, because what if we destroyed civilization as we know it? Would they be on the hook for it?

Having dealt with Amazon Payments and Paypal, I dreaded what we'd have to go through. But now, after learning more than I ever wanted to know about the online payments industry, I feel that getting booted by Amazon was the best thing that could have happened to unglue.it. There's a new generation of payments companies out there, and when the big payments companies have their Lehmann-Bear-Stearns-Brothers moment, you'll be happy if you escaped in time.

To understand what makes payments companies tick, you have to understand something about the risks they take on, and the ways they try to mitigate risk. For a merchant, taking a credit card means you don't have to worry about whether someone's credit is good. For a consumer, paying with a credit card means you have some recourse if a merchant tries to cheat you. Credit card fraud is probably older than credit cards, and credit card companies have decades of experience detecting fraud and reducing its cost.

Internet commerce introduced entirely new categories of risks, and conventional payments providers were not eager to change. To address the needs of a new era, a number of payments companies came into existence, most notably, Paypal. As Paypal took off, companies like Amazon, Apple, Ebay and Google did not stand idly by.

But anyone who's used Paypal, Amazon Payments, or Google Wallet recently can't help but notice that these services have calcified. Even worse is that they've metastasized as impersonal, unresponsive machines. In the payments biz, calcification has advantages. Consumers have learned to trust these companies and their brands, and the companies have in turn done their best to turn into rocks of calcite stability.

But if you want to change the world with your fantastic start-up, you're screwed. If you're doing something new, the big guys can't evaluate your risk profile. Because they're machines. For changing the world, you need people, or at least cyborgs. You need high-powered big-data risk analysis supplementing the judgments of seriously smart people. And over the last month, I've found some of this at three different companies.

First, let me tell you about some companies we didn't pursue. Dwolla is really interesting, but we thought it would be hard to sell Dwolla and its requirement for direct bank account access to our core demographic. If our users were high-involvement, high frequency users, we might feel differently. Similarly, the marketing barriers to using things like bitcoin are prohibitive at this time. Mobile payments solutions such as Square aren't so relevant for our application.

Another company we were interested in was Braintree Payments. We reached out to them, but they told us that three weeks prior to our request, their banking networks had told them to stop supporting crowdfunding businesses.

So we looked at WePay, Balanced, and Stripe. Three breaths of fresh air.

Of these three the company that most closely replicates the Amazon/Paypal model is WePay. Like Amazon/Paypal, WePay allows users to create a WePay account. As WePay becomes more and more accepted, consumers who might be hesitant to give credit card info to an unfamiliar service like Unglue.it will recognize the WePay brand and grant access to their WePay account, if they already have one, via OAuth. If the user doesn't have a WePay account, the the marketplace can create one on their behalf. When everything works, friction in the transaction drops dramatically.

Wepay's API (which was not available when we were looking for a payments provider a year ago) allows all the key functions Gluejar would need to build a crowdfunding application. These include conditional and delayed payments. For example, WePay is the payments provider for GoFundMe. The biggest downside for us is that Wepay is quite specific about only allowing payees who are US citizens. Still, we were able to identify a work-around for our non-US rights-holders. Balanced and Stripe had similar issues and work arounds.

Balanced, the smallest of the three companies we talked to, is in many ways the most exciting. They're focused on providing solutions specifically tailored to markets. They've baked the concepts of multiple buyers and multiple sellers into their well-documented API, and offer features such as the ability to escrow payments into the marketplace account. If you so much as browse their website you'll be given an API key and be offered the chance to do a test transaction.

Markets present rather different risk profiles to a payment provider than do simple merchants. The diversity of participants and flow of payments makes it harder to detect fraud. But if you have smart, experienced people involved, the risks can be assessed and mitigated. Fraud detection algorithms can be optimized for markets. The possibilities of markets enabled by the internet are unfathomable. For example, last week I learned of a company, VoiceBunny, that has built a marketplace for voice talents. Crowdvoicing! You feed some text into an API and soon, someone with a great voice will speak it for you. They're using Paypal now, but they would be smart to look at Balanced.

GitTip IS using balanced, after a fiasco with FeeFighters and finding their business model incompatible with Stripe's underwriters. Gittip's founder, Chad Whitacre, had to "understand and mitigate the risks of running a marketplace", and Balanced is helping him do that.

Stripe was the most "boring" payments company we looked at. I use the word "boring" with much admiration, because they're focusing on the relatively boring old problem of enabling internet merchants to take credit cards and just doing a better job of it. On the implementation side, they've succeeded brilliantly, with an API and developer support that just makes you want to throw your Paypal code into the trash and drink some good bubbly to celebrate its demise.

The folks at Stripe took the time to understand the issues of payment risk inherent in our business model, and what we're doing to address it. To be able to use Stripe for our marketplace, we would need to take some responsibility for delivering the Creative-Commons Licensed eBooks that our crowd is funding. It means that we can't work like Kickstarter, which disclaims any and all responsibility for completion and delivery of the projects on its website. Since that quality control and delivery assurance is a big part of the extra value we provide over Kickstarter, Stripe is a viable option for us. As are Balanced and WePay.

In deciding among these three great companies, the questions we asked ourselves were more about what we want to be than anything else. Do we require the scaling characteristics of a marketplace or are we just a merchant that wants to take creditcards? Do we want a payment network or can we stand on our own? As the world of internet payments develops, which company will develop solutions aligned with our particular needs?

I'm not going to say which company we've settled on until we make sure that everything works, but I've very sure that whether it's that one or one of the others, we'll be in a much happier space than we were with Amazon or Paypal. If you are considering a switch from a payments company that is too big to care about your success, or if you are deciding on a provider for a new application, you should really look at the new companies. A healthy internet economy needs a diversity of payments providers who employ more than just algorithms, who support businesses rather than accounts.

The "I Used This" Button

I frequent-flyered off to San Francisco this weekend to surprise my Ph. D. Advisor, Jim Harris, for his 70th Birthday. I was the first of his students to graduate, and he's up to 105. On thing I learned helping to start his group was the immense value of being thrown together with a group of smart people with a variety of experience. I met members of Jim's current group, which includes a student from Gunn High School, visiting scholars from around the world, and Ph. D. students bursting with ideas.

I manufactured some business-related meetings for the trip, some of which I'll relate in a to-be-written post, but I also lucked into a hackathon for Open-Access hosted by PLoS. I spent the day with a group of smart people with a wide range of experience, including software developers, product managers, film-makers and a librarian or three.

The group I ended up working with included Greg Grossmeier from Creative Commons, Cameron Neylon from PLoS, and Ana Nelson, the developer-entrepreneur behind dexy. We were interested in counting open-access things. Counting things can be harder than you think, because you have to define the things and identify them; you need to be able to tell whether a thing is the same thing as another thing, or perhaps it's three things. Counting bananas is one thing, but have you ever tried counting ideas?

Creative Commons (CC) is interested in knowing how much its licenses are used. When an Unglue.it ebook edition is released (of course under Creative Commons!), how often is it used? Does a single license apply to the entire book, or can we apply different licenses to the different resources inside the book? For example, an author may want to use a CC-BY license for the text of a book, which might contain figures that are used under CC BY-ND. And the metadata should be CC0. How should these licenses be expressed?

After some discussion, we settled down to work on some specific projects. My project turned out not to be code at all, but rather a description of a scheme for measuring Creative Commons usage, i.e. the rest of this blog post.

Creative Commons has thought about ways to measure the usage of its licenses. For example, it can track the display of its license "badges", such as the one right here. Web browsers will send a referrer header that tells the image server the web page and user IP address. But there are problems. Many web sites use their own copy of the badge. In an ebook, the badge would be embedded in the ebook file. If the page is served over a secure socket, the referrer won't be set. And do you really want to tell Creative Commons about everything you're reading?

Speaking of which, have you clicked on a Facebook "Like" button this week? Was it good for you too?

Suppose there was a button on Creative Commons licensed documents that allowed the user to express their delight at the creator's enlightened choice of license. Would you click it? I call it the "I Used This" (IUT) button, but maybe you can think of a better name.

• The IUT button would send a signal to a Creative Commons server about usage of the resource. These signals would be compiled and reported.
• IUT button would also send attribution url.
• Pressing the button would display an amusing animation to the user. Perhaps every button would have a different animation to avoid button fatigue.
• The button would be at the center of an advocacy campaign for open licenses.
• Unlike the Facebook Like button, the IUT button would respect a user's privacy. A signal would be sent only when initiated by the user, and would be optional.
• An IUT button packaged as a javascript would work in epub, html, etc.
• IUT signals would be evidence of the resource's status as a CC licensed work. A licensor attempting to revoke a CC license (you can't do that!) would have to overcome a verifiable usage trail.
• Users could create accounts at CC to provide a retrospective record of the user's Use.
• Clicking the IUT Button would put the attribution url on clipboard to ease correct citations.
• Usage information for each resource would be public- creators could easily track the usage signals for their works.
• We might need anti-ballot-stuffing measures if CC usage rankings become commercially important.

If efforts like Unglue.it are to succeed, people who appreciate the benefits of Creative Commons licensing need to stand up and be counted. We need to make it a mass movement in the minds of every lover of books, everywhere.

Sometimes you need to do more than just consume. Sometimes you need to do some SHOUTING.

Why I'm not mad at Amazon

Amazon seems moderately competent to me. That puts them in the top decile of large American corporations. So when Sarah Houghton, the "Librarian in Black" tweeted me "I hope you won't mind if I'm still mad at them ;)" I realized that I'm not mad at Amazon at all. A big company acting like a big company is not the sort of thing I get mad at. I've worked for one of the biggest companies in the world (at least it used to be!)  So when Amazon forces my small company to suspend crowdfunding because of its big-company traits, I find it hard to get too worked up.

But let me tell you the whole story.

Twelve years ago, my company, which was named "Openly Informatics" at the time, launched a book-linking service called "LinkBaton". We offered users the ability to link to the bookstore or library of their choice. Our value proposition for websites was that by letting users pick their favorite bookstore- Amazon or BN or whatever, they'd get more affiliate commissions. Meanwhile, we'd make money by getting good search optimization for our links. The problem with the business was that Amazon was so good at converting traffic to sales that any diversion of traffic to a non-Amazon site would reduce rather than increase a website's revenue. Still, our traffic built up, and at our peak, we were driving 200 transactions per day and getting nice sales commissions through Amazon Associates. Our peak day was September 10, 2001.

But that's another story entirely.

So we "pivoted" into library software, linking to journal articles instead of books. That turned out to be successful, and I ended up selling the business and the name, but not the company, to a big non-profit in the library space. I worked there a little over three years, met some wonderful people, and learned a whole lot.

When I left that job, I reactivated the company (now called "Gluejar") and did some odds and ends, consulting, selling some books through Amazon Marketplace, and blogging about ebooks and libraries. Amazon was happy to give Gluejar a business account.

Last year, I decided that creating a public sector for ebooks was an important thing to work on, and that crowdfunding was the way to do it. Unglue.it was born. The main technical uncertainty was the payment system. I hired some engineers and we compared various providers and chose PayPal, mostly because they offer better international coverage than Amazon Payments. Also we were hesitant to use Amazon because book publishers hate Amazon- they're jealous of Amazon's prowess in  e-commerce.

But Amazon Web Services is amazing. We have all our servers in the Amazon cloud.

And then we waited for PayPal to turn us on. That was January. We're still waiting. It seems PayPal has been reevaluating their stance toward crowdfunding businesses.

In April, we felt we'd waited long enough and decided to look at Amazon again. Kickstarter was seeing more and more success with Amazon Payments, so we figured they really had their act together. We already had a business account, we just had to get approved for Flexible Payment System (FPS) access. We verified our bank account, and 2 weeks later, FPS was turned on. With real money. Amazon is so awesome!

After all sorts of testing, we launched the website on May 17. One month later, our first campaign succeeded. We had supporting tweets from CoryDoctorow. And TimOreilly. And Weezer, goddammit! We carefully ran 259 payment authorizations, and real money appeared in our account. An ebook, Oral Literature in Africa, will be free to the world! Everything just worked.

Two weeks ago, I got this very phishy email:

Greetings from Amazon Payments,
Thank you for registering with Amazon Payments. We appreciate your interest in our product.
Unfortunately, at this time, we are not able to approve your request for an Amazon Payments business account based on our review of your intended use of our payments service.
As stated in our Acceptable Use Policy, the following product or services are prohibited from using Amazon Payments:
• Donations and Charitable Solicitations - includes charities and non-profit organizations without a valid 501(c)(3) tax exempt status, charitable solicitations, commercial fundraisers (including commercial co-venturers), or any activity associated with the solicitation of donations.
A member of our business team will be in contact with you to further discuss your business model and use of Amazon Payments.
---------------------
We have temporarily activated your account. Please initiate a new transaction to withdraw funds into your bank account.
Sincerely,
Account Specialist
Amazon Payments
https://payments.amazon.com

We thought it was a phishing attack because

1. They didn't know we'd had a business account for 3 years
2. They said nothing about FPS
3. The action requested was to make a withdrawal. Huh?
4. Donations? If what we're doing is asking for donations, when we commit to deliver ebook licenses for contributions, then what is Kickstarter, which promises squat, doing??

So the next morning, I called Amazon Payments customer service. The representative confirmed that nothing had changed in our account. But there was a note from the "verification team". And it was impossible, even for her, to talk to the verification team. So she said we should reply to the phishy email, and wait for the "member of the business team" to contact us.

So I replied, and the passive-agressive robot at the other end seemed to either want a human to intervene or to nuke our account. That scared us, because the account was linked to AWS and all sorts of good things. We felt confident that when we could speak with an actual human, all the misunderstandings would melt away. And so, after a week of not hearing more, I went away for a week at the New Jersey shore.

On Tuesday afternoon, I finally got an email from a human at Amazon Payments who knew what was going on, a Senior Account Manager. We scheduled a call for Wednesday noon. (we sleep late here at the beach) He was really nice and treated me with utmost respect. I explained my company's long history with Amazon and our crowdfunding achievements to date. He explained his much shorter history with Amazon and expressed admiration for what we were doing. He then explained some of the regulatory and contractual burdens that Amazon Payments (an entity separate from Amazon) has to meet, and how everything was more complicated with crowdfunding accounts. As a trivial example, they have to verify that their payments are not going towards alcohol or firearms. He did not mention nuclear weapons or drug money laundering, but I extrapolated. And it seems that Amazon has to spend money to do all this verification. They discovered all this crap because they're doing payments for Kickstarter. And until they figure out how to make sure that Kickstarter is not funding alcohol, drugs, nuclear weapons, prostitution and credit-default swaps, they've decided not to accept any more crowdfunding accounts.

I make it sound a bit crazy, but that's the world we live in. Think about it: how would you make sure the thousand business partners of your thousand business partners aren't dealing cocaine-tipped Stinger missile futures contracts on the side? And the questions we've gotten from the Paypal people were heading in the same direction, so I'm inclined to believe Amazon Payments Guy. I think it's rather unlikely that Amazon is singling us out because we're funding free ebooks, or, as has was suggested on Twitter and Hacker News, that we threaten Amazon's Kindle business. In my dreams. (Which will come true!)

So we had to suspend our ungluing campaigns. Striking a blow against arms dealing and drug money launderers everywhere.

It would have been nice if, back in April, we had gotten a clear signal from Amazon that they weren't doing new crowdfunding businesses. Most likely, we would have gotten that signal in April if not for our 3 year old business account. I can easily imagine a meeting inside Amazon where some participants wanted to go full blast on crowdfunding and others wanted to ban it, resulting in a compromise position of no new crowdfunding "for now", but no announcement either. The crack we fell into was just collateral damage. Small companies that get shoved aside are not Amazon's problem. But no, I'm not mad about it. Did I mention how amazing AWS is?

I'm curious though. If it was me running payments at Amazon, I'd see this mess as an opportunity. If Amazon had tighter, front-to-back control of crowdfunding, it could do it better, more efficiently. If it was me running payments at Amazon, I'd set up a better-designed crowdfunding marketplace. No wait! If it was me running payments at Amazon, I would buy Kickstarter!

And based on sheer speculation, it's entirely possible that Amazon is right now trying to buy Kickstarter. In October 2009, Kickstarter said it was working to support international projects   and as of 2012, nothing has happened. Why dyaspose?

As for our relaunch, we'll be looking at companies like Wepay, Dwolla, Stripe, Balanced, Braintree and a few others. Maybe the regulatory crap is easier to deal with in a smaller company that can know its customers. I'd love to hear suggestions. We're also working on a way to decouple our market from payment providers entirely. Who knows, maybe Paypal will come to our rescue. whatever happens, we'll be back sooner than you'd think.

Until then, if you like our model for ungluing ebooks, please register on the site. Or start a nuclear reactor project on Kickstarter.

Update: Here's the followup post.

Running a Start-Up

I'm a runner. Not a fast runner- my "fast" pace is 9:20 per mile. Some days I have to push myself, particularly when it gets above 90°F. My muscles are old and stiff, and they'd rather just prop themselves in front of my armchair. Some days I don't have the will to keep going the full 5 kilometers and I stop and walk a bit. Some days I'm riled up about something and I run faster without realizing it.

I like to think while running. The blood starts flowing to my groggy brain and synapses start firing. Before that happens, I get over the complaints from my legs and feet and chest. The music from my iPod Shuffle helps a lot. Thank you for that, Steve.

And then I get to the end of my run. I drip with sweat and cool down by sweeping the oak pollen or bark or acorns off the back deck. Even after my shower, I'm still soaked.

What makes me do it? It's not like I'm training for Olympic glory, there's no coach urging me on.

Doing a start-up is like running. I often enjoy it, but just as often I have to push myself to do things I'd rather skip. Once in a while it takes your breath away, like running on a perfect day, getting to the top of a hill into a cloudless sky. Mostly though, it's fighting to keep up a trot when walking would be a lot easier.

I enjoy building things, and I like writing code. So that part of a tech start-up is not a problem. Calling people on the phone, on the other hand, is torture for me, but I do some of that because I have to. Multi-tasking is not my strong suit. Managing creative and smart people can be simultaneously a joy and a bundle of stress.

And then there's the money. Every founder has the ever-present worry about worry about running out of money, which is tempered only by the knowledge that if you don't run out of money, you'll have succeeded.

So, about money. I'm sometimes asked why Unglue.it is a for-profit business rather than a not-for-profit. The main reason is what I just said. I like building things and technology is fun. Calling people on the phone to ask for grants- that would be a nightmare for me.

But I think it's important for a business like Unglue.it to be transparent about finances. So I've written a post on the Unglue.it blog reporting on its finances. Authors and publishers will need to be become more open about the money they expect to receive for ungluing a work, and in many cases this will be uncomfortable for them. So unglue.it needs to set an example.

It's a fair amount of money I'm putting into Unglue.it. Why do I take the risk? I consider myself to be extremely lucky to be able to make the investment, because it's an important idea that needs to be explored. But it's also like my running. What good are legs if you don't stretch them, make them work?

I'm a worrier, and there's plenty to worry about. My recent worries have surrounded things like insurance (Unglue.it's "presence" in 4 states makes that not fun). And payment providers. Amazon could decide tomorrow that we're a threat to their book business and turn off our payments account. Google could decide on Wednesday that books aren't worth the trouble and shut down their book search API. On Thursday, hackers could take control of our servers and trash our database.

But you just have to keep at it. As of the end of June, the Unglue.it team had logged 2,222 Git commits and finished 333 Pivotal stories. We've talked to hundreds of people. We've written support email, press releases, blog posts and heartfelt opinion pieces. We're keeping at it, so there's less time to worry.

The wonderful response to Unglue.it has kept the entire team going strong. Yesterday, Unglue.it signed up its 1,800th ungluer. Most likely, we'll fall just a bit short of our goal to grow that number by 40% every month. That's a very healthy growth pace, but it's still a tiny number. To give our campaigns the exposure needed to match the earnings potential of conventional book selling, we'll need to have 10 times or even a hundred times the number of ungluers. Even at 40% per month, we'll need till the end of the year to reach 10,000 ungluers. We'll get there one way or another, but it's hard to know how long it will take. It's going to be a marathon, and I've never run one.

Every now and then I'll need to slow to a walk. If you offer me some water, I won't say no.

Secret Desert Meeting Report: It Was Hot

Don't think that a relative absence of blog posts means I haven't been working hard.

Unglue.it was a first-time exhibitor at the American Library Association Annual Meeting. We talked to a lot of people. We gave out stickers. I gave a talk. I met a mermaid. Here are my observations:

• I'll bet we talked to 200 people. And of those 200, there were about 5 who weren't excited to learn about what Unglue.it is doing. That's a pretty good batting average.
• Of the 200, half said they'd heard about us or read about us somewhere. About 5 of those had actually supported one of the campaigns. And those 5 were bringing Andromeda cookies to keep her from expiring. That's hitting a lot of singles but not scoring a lot of runs. But I'm happy to report that the unglue.it team survived.
• The total attendance at ALA was around 20,000. That means we barely reached 1% of attendees. But they were a very good 1%. The number of people who have signed up to be ungluers passed 1,400. Ungluers pledged over $1500 to the campaign for Joe Nassise's Riverwatch. But Joe will make much more than that by leaving the ebook on Amazon, so it won't be our next book to unglue. We need to multiply our numbers by 10. And we will. Our 195 converts will talk to 195 more and so on. We have enough stickers.
• The mermaid was also excited about Unglue.it.
• We had some discussions that will lead to amazing things. 

After the meeting was over, The unglue.it team converged on an undisclosed desert location to figure out what to do next. We decided to get some sleep.

It gets hot in the desert. 100°F by noontime. But that didn't stop us from our mission. We had electronics. We had internet.  We had expense reimbursement policy to discuss. The words "Gold Lamé" were mentioned. World domination was considered inevitable, given enough time.

And speaking of a good batting average, Carlos Ruiz of the Philadelphia Phillies was named to the National League All-Star team today, and the honor has never been more deserved. Major-league catchers crouch and stand about 200 times a game, catching 90 mile-per-hour fast balls and curves that make the air hum. Or not, and then they have to block the ball with their bodies. They are routine smash targets for runners trying to score through their bodies. On top of this, they're expected to throw out runners, manage their pitchers,  and know the weaknesses of the opposing hitters. Oh, and hit occasionally. This year, Ruiz is leading the league in batting average, and to use a techniccal term, mashing.

There was a time when Ruiz was considered too old to be a prospect. He just needed some time.

Awesomeness takes its own time.